Vol 2 No 1 - June 2013

International Journal of Computer Science
and Business Informatics

(IJCSBI.ORG)
ISSN: 1694-2507 (Print)

VOL 2, NO 1
JUNE 2013
IJCSBI.ORG
Table of Contents VOL 2, NO 1 JUNE 2013
Analytical Hierarchy Process Based Framework for Modelling Preferences and Priorities in
Requirements .................................................................................................................................... 1
R.Subha and S.Palaniswami
Unstructured Data Integration through Automata-Driven Information Extraction ............................... 1

Maroun Abi Assaf, Kablan Barbar, Youakim Badr and Mahmoud Rammal
Wireless Solution for Water Saving In Agriculture Using Embedded System ......................................... 1
Venkata Narayana Eluri, K. Madhusudhana Rao and A. Srinag
Upcoming Trends of Virtual Experiments for Laboratories................................................................... 1

Bhaskar Y. Kathane, Pradeep B. Dahikar and Satish J. Sharma
Negotiation Based Resource Allocation for Distributed Environments ................................................. 1

Krishnamoorthy M. and Senthil Murugan B.
Wireless Solution for Water Saving In Agriculture Using Embedded System ........................................ 1
Venkata Narayana Eluri, K. Madhusudhana Rao and A. Srinag
Digital Image Tamper Detection Techniques - A Comprehensive Study ............................................... 1

Minati Mishra and Flt. Lt. Dr. M. C. Adhikary
A Multi-parametric based W-PAC Mechanism in Ad Hoc Network using IPv6 and IPv4 Address ............ 1
S. Thirumurugan and Dr. E. George Dharma Prakash Raj
A New Online XML Document Clustering Based on XCLS++ .................................................................. 1

Ahmad Khodayar E Qaramaleki and Ahmad Khodayar E Qaramaleki
Secured and Energy Based QoS Routing in MANETs ............................................................................ 1
S. Sridhar and R. Baskaran
International Journal of Computer Science and Business Informatics
IJCSBI.ORG
Analytical Hierarchy Process Based

Framework for Modelling Preferences
and Priorities in Requirements
R.Subha
Assistant Professor,
Department of Computer Science and Engineering
Sri Krishna College of Technology, Kovaipudur, Coimbatore-641 042,
Tamil Nadu, India
S.Palaniswami
Principal,
Government College of Engineering,
Bodinayakanur, Tamil Nadu, India
ABSTRACT
Requirements engineering is a software engineering process which covers all the activities
involved in discovering, documenting and maintaining a set of requirements for a
computer-based system. The priorities that stakeholders associate with requirements may
vary from stakeholder to stakeholder. Different priorities imply different design decisions
for the system. So there must be a model to support the representation of preference in
requirements. In this paper we develop a framework to model alternative solutions for
mandatory goals and preferred goals based on priorities. A framework is created for
specifying preferences and priorities among requirements. The priorities among the
preferences are analysed by Analytical Hierarchy Process method (AHP). AHPs pair wise
comparison method is used to assess the relative value of the candidate requirements. The
preferences are analysed based on the prioritization of each task and a definite plan is
generated to view all those tasks according to priority.
Keywords
Requirements Engineering, Preference Goals, Priorities, Goal oriented frame work,
Analytic Hierarchy Process.
1. INTRODUCTION
Software engineering is a modelling activity in which, software engineers
deal with complexity through modelling by focusing at anytime, only the
relevant details and ignoring everything else. Software Engineering
encompasses all the life cycle activities of a project associated with
understanding a products necessary capabilities and attributes.
Requirements Engineering (RE) is the process of establishing the services
that the customer requires from a system and the constraints under which it
ISSN: 1694-2108 | Vol. 2, No. 1. JUNE 2013 1

IJCSBI.ORG
operates and is developed. It may range from a high-level abstract statement
of a service or a system constraint to a detailed mathematical functional
specification. Requirements engineering acts as a bridge between the real-
world needs of users, customers and other constituencies affected by the
software system and the capabilities and opportunities afforded by software-
intensive technologies.
Preference goals are the goals that are chosen by the stakeholder, but it will
not always be in the same degree of desirability [1]. Sometimes preference
goals are considered as those goals which help other goals benefit from it.
For instance, in the book shop details example, we may opt to accomplish
the book order through a payment approach which is undesirable for the
customere.g., money orderbut still we favour the solution because, an
alternate goal may be suitable, say to maintain robust legal documents.
Such goals are represented as preferences or preference goals.
Prioritizing requirements is one of the tasks of the Requirements Phase. It

determines the implementation order of the requirements in an incremental
and iterative development cycle. Preferences are not always longed for or
essential [2]. For a distinct situation or context or for a distinct stakeholder,
a subset of preferences seems to be more applicable than any other
requirement. Stakeholders satisfaction must be considered to be the
unblemished goal and a navigating theme by the prioritizing process.
Analytical Hierarchy Process (AHP) is used to estimate the priorities among

the requirements. AHPs pair wise comparison technique is used to assess
the relative value of the candidate requirements.AHP shows pair wise
comparisons and subsequent approximation of the Eigen vector of the
resulting comparison matrix in order to acquire greater accuracy of the
numeric prioritization result. This also instills confidence in us that the
model has been elicited and constructed appropriately, that the domain is
well established and that AHP yields an accurate prioritization profile (e.g.,
with a good consistency ratio).Requirement prioritization is utilized in the
software product management to investigate which candidate requirement in
a software product should be comprehended in a certain release.
Requirements are also prioritized to minimize risk during development so
that the most essential or the requirements that are extremely risky are
implemented first. It is always better to force the stakeholders to address all
the requirements and not just their own Prioritized requirements or the ones
with the determined order of importance to some stakeholder or class of
stakeholders pertaining to the requirements along one or more dimensions
ISSN: 1694-2108 | Vol. 2, No. 1. JUNE 2013 2

IJCSBI.ORG
(e.g., personal preference, business value, cost of implementation, and risk).
Based on the priorities we develop a goal model to reveal better solutions of
goals that are mandatory and goals that are preferred. The priorities and
preferences among requirements are specified by creating a framework. A
Goal Model is a key aspect of Requirements Engineering that may be used
in Stakeholder analysis, Context analysis, and Scenario more evidently.
Through acute cooperation of actors in the predestined software a system
should acquire the targets that are goals. The tradeoffs between cost,
performance, flexibility, security and other goals can be located and quickly
resolved by utilizing goal modelling. Aberrant interests between
stakeholders are also revealed using goal modelling. Goal modelling can
spot conflicts when interference occurs among the goals. Goal modelling
empowers requirement completeness to be measured. If all the goals in the
goal model are attained the requirements can be considered to be complete.
A goal model is effective in concisely capturing large number of alternative

sets of low-level tasks, operations, and configurations that can fulfill high-
level stakeholder requirements. The high-level stakeholders requirements
include specifying preferences and priorities among various requirements
and select specifications that fulfill mandatory requirements. Based on the
priorities among discrete tasks alternative solutions are detected. This
enhances the segregation between mandatory and preference goals, and
paves way to the destined goal. This also establishes the relative
significance of all the requirements.
The rest of the paper is organized as follows. In Sect. 2, we present the goal
and the preference goal and assign certain priorities to those goals by AHP
method. Then, in Sect. 3, we take a look at related literature. In Sect. 4, we
have various methodologies to support the goal oriented framework In Sect.
5, we discuss the performance of the proposed reasoning task and conclude
it in Sect. 6.
2. RELATED WORKS
Earlier, in modeling and analysis only the preferences of the goals were
considered for analysis and the priorities for the preferences were expelled.
Technes perspective showcased the RE process which was involved in
describing requirements problems and comparing the candidate solutions at
an early stage. Object-orientation, predicates, temporal constraints, and task
sequencing are missing since they contribute to the detailed solution of the
chosen candidate solution [3]. Techne precedes and complements RMLs for
detailed specification which does include such features. In the course of
ISSN: 1694-2108 | Vol. 2, No. 1. JUNE 2013 3

IJCSBI.ORG
these phases, the requirements problem for the system-to-be is being
structured and its candidate solutions are described and compared, based on
their desirability to the stakeholders.
In modeling mandatory requirements, only tasks and goals are considered,

whereas, the alternative solutions are not specified. A modelling and
reasoning toolset for exploring and evaluating the alternative solutions to
requirements problems, is adopted [4]. The temporal constraints are
expressed by extending the common goal modelling formalism and the
alternative plans for fulfilling stakeholder goals are represented using partial
goal satisfaction. The users can specify their priorities over optional, nice-
to-have properties of potential solution by using a preference specification
language.
In customization technique the preferences are not taken into account and
the software is developed according to the customers individual
requirements. Software needs to support customization at a much finer-rate.
Software assists people to live their lives despite disabilities, constitute a
significant class of applications where greater degree of customization is
essential [3]. This paper proposes a framework that supports the design of
highly customizable software by adopting goal-oriented analysis techniques.
The working of this framework can be demonstrated with a simple example
involving the design of a customizable email system.
Different design decisions for the system crop up due to varying priorities.
So we need to introduce a framework to specify both the preference
requirements and priorities in order to fulfil the mandatory requirements.
This enhances the extension of a traditional goal modelling notation to
support the representation of optional and preference requirements [5]. In
our extension, optional goals are marked off from the mandatory ones.
Later, quantitative prioritizations of the former are constructed and used as a
criterion for evaluating alternative ways to achieve the latter. In
Requirements Engineering (RE), goal-oriented techniques assist as to what
the stakeholders wants (goals) and the means (actions/tasks/plans) by which
these goals can be achieved, thereby gaining noticeable attention.
The basic concerns that are involved in the stakeholders communication

does not come under requirement engineering. These include beliefs,
desires, intentions, and attitudes. To accelerate these, we propose a core
ontology that covers these concerns and is grounded in sound conceptual
foundations resting on a foundational ontology [6]. The new core ontology
for RE paves the way to a new formulation of the requirements problem.
New standards and criteria are inculcated to specify the minimum amount of
ISSN: 1694-2108 | Vol. 2, No. 1. JUNE 2013 4

IJCSBI.ORG
information that has to be represented in RE languages and to determine if
the RE has gone into a favourable completion.
The existing system has goals that are modelled by intrinsic features such as
their type and attributes, and by their links to other goals and two other
elements of a requirements model. The modelling and reasoning about
preferences for stakeholders in the presence of mandatory goals remains
largely unexplored. Goal-oriented techniques in requirements engineering
have a unique way to keep one going between stakeholders wants (goals)
and the means (actions/tasks/plans) to achieve goals [1]. The current goal-
oriented modelling frameworks treat goals as mandatory requirements that
must be fulfilled by any proposed solution. In this respect, such frameworks
cannot accommodate reference (nice-to-have) requirements that might be
posed by stakeholders.
3. METHODOLOGY
3.1 Construction of a Goal Model
Goals are targets that are earned through acute cooperation of actors in the
predestined software. Initial phase of a project is best suited for goal
modeling. The relationships between a system and its environment are
expressed [7]. Goal models are effective in capturing large numbers of
alternative sets of low- level tasks, operations, and configurations that can
fulfil high-level stakeholder requirements.This module contains basic task
and the goals that are represented using GR Tool which is the goal
reasoning tool.
The GR-Tool is graphical tool that is employed to draw goal models and run
the algorithms and tools for forward and backward reasoning [8]. The
tradeoffs between cost, performance, flexibility, security and other goals can
be located and quickly resolved by utilizing goal modelling. It expresses the
relationships between a system and its environment. Goal models have been
found to be effective in concisely capturing large numbers of alternative sets
of low-level tasks, operations, and configurations that can fulfil high-level
stakeholder requirements. The algorithms for the forward reasoning have
been fully developed in java and are embedded in the GR-Tool. Tasks are
represented in hexagonal shapes and goals are in oval shapes. Goals and
tasks are connected with each other using AND- and OR-decompositions. A
goal model based on a book shop detail is shown in Figure 1.
ISSN: 1694-2108 | Vol. 2, No. 1. JUNE 2013 5

IJCSBI.ORG
Figure 1 Goal model
3.2 Identification of Goal Preferences and Priorities

Initial phase of a project is best suited for goal modeling. [5]. A goal model
is effective in concisely capturing large numbers of alternative sets of low-
level tasks, operations, and configurations that can fulfill high-level
stakeholder requirements [9]. Goals and tasks are connected with each other
using AND- and OR-decompositions. AND-decomposition means
decomposing a goal into other goals or tasks, in which all children should
satisfy the fundamental goal or task. Some set of tasks are given below
Table 1. Set of tasks
Tasks
search book
user login
book database
storing book details
stock details
mention books
store details in database
book number
availability
integrity
manage categories
book name
change password
perform privacy control
provide weekly renewal of database
number of books
ISSN: 1694-2108 | Vol. 2, No. 1. JUNE 2013 6

IJCSBI.ORG
If the goal is OR-decomposed into other goals or tasks, then one of these
goals or tasks should satisfy the parent goal. Preference goals are the goals
that are inclined by the stakeholder.The set of goals and its decomposition is
given below.
Table 2. Set of goals
Goals Decomposition
manage internet OR
searching
internet querying OR
catalogue consulting OR
taking detail of AND
transaction
date & time AND
member name AND
select item OR
pick available item OR
pre non ordered item OR
shopping cart AND
add item AND
check out AND
select item AND
get identification AND
details
But the same degree of desirability will not be maintained. Sometimes
preference goals are considered as those goals which help other goals to
benefit from them. Prioritizing requirements is the requirements task of
determining the implementation order of the requirements in an incremental
and iterative development cycle.
3.3 AHP Method

The Analytic Hierarchy Process (AHP) is a theory of measurement through
pair wise comparisons and relies on the judgements of experts to derive
priority scales. It is these scales that measure intangibles in relative terms
[10]. To make decisions, we identify, analyse, and make tradeoffs between
different alternatives to achieve an objective.
The outcome is more likely to be satisfied when the means for analysing and
evaluating the alternatives are more efficient [3]. The Analytic Hierarchy
Process compares alternatives in a stepwise fashion and measures their
contribution systematically to take decisions.
ISSN: 1694-2108 | Vol. 2, No. 1. JUNE 2013 7

IJCSBI.ORG
To what extent an aspect dominates another with respect to a given attribute
is featured by comparing a scale of absolute judgements. The judgements
may be inconsistent. A method to measure inconsistency and improve the
judgements and the possibility of obtaining a better consistency is a concern
of the AHP. In AHP method,
Information is decomposed into a hierarchy of alternatives and
criteria.
Information is then synthesized to determine relative ranking of
alternatives.
Both qualitative and quantitative information can be compared using
informed judgments to derive weights and priorities.
3.4 Determination of Alternative Solution

The AND/OR goal decomposition model represents alternative solutions
for fulfilling the fundamental goal [7]. Each task consists of certain
priorities that may be calculated using AHP. In AHP information is
decomposed into a hierarchy of alternatives and criteria, then synthesized to
determine relative ranking of the alternatives. A set of preference goals is
given below.
Table 3. Set of Preference goals
Preference goals
Whole book details
Not take same book
Arrange book by
volume
Arrange books
Printed robust legal
document
Provides steps to user
Fear internet fraud
Secure form
Standard form
In order to derive weights and priorities both qualitative and quantitative
information are compared using informed judgments. The alternative
solution can be found by taking the goals and task based on AND, OR
decomposition. Each task is prioritized and based on the prioritys task the
goals may be compared .The goal which has the greatest priority may be
considered to be the first and the alternative solution is found out. The
ISSN: 1694-2108 | Vol. 2, No. 1. JUNE 2013 8

IJCSBI.ORG
priorities of goals are calculated by using values in the relative intensity
table.
Algorithm
Calculation of priority among goals
Input: Set of goals based on the preference
Method: Analytical Hierarchy Process
Step 1: Set up n goals in the rows and columns of an n * n matrix.
Step 2: Perform pair wise comparisons of the entire goal according
to the criterion.
Step 3: Use averaging over normalized columns to estimate the
Eigen values of the matrix.
Step 4: Assign to each goal, its relative value based on the estimated
Eigen values.
Output: Plan is generated based on the priorities calculated, and the
tasks connected to each goal are listed. The relative intensity table
is given below on which priorities between requirements is taken
on the basis of the relative intensity
Table 4. Relative intensity table

Relative Definition Explanation
intensity
1 Of Equal Value Two goal are of equal value
3 Slightly more value Experience slightly favours one goal over another
5 Essential or strong value Experience strongly favours one goal over another
7 Very strong value A goal is strongly favoured and its dominance is

demonstrated in practice
9 Extreme value The evidence favouring one over another is of the

highest possible order of affirmation
2, 4, 6, 8 Intermediate values When compromise is needed
between two adjacent
judgments
Reciprocals: If goal i has one of the above numbers assigned to it when compared with goal
j,then j has the reciprocal value when compared with i.
ISSN: 1694-2108 | Vol. 2, No. 1. JUNE 2013 9

IJCSBI.ORG
3.5 Reasoning about preferences
Here, based on the priorities, a model is constructed and preferred plans are
organized. Ordered sequences known as plans are considered as alternative
solutions. The sequence of leaf-level tasks that altogether satisfy the
AND/OR structure of the fundamental goal is the plan for the fundamental
goal. The priorities may be listed as probabilistic values. To render solutions
to goal requirements problems a plan is essential [3]. Solutions in goal
models are specified as configurations of tasks with no dependencies.
Without specifying the system dynamics the stakeholders requirements can
be modelled and solutions to requirements can be determined using
automated constraint solvers. Thus a set of tasks are contained in the plan
that is generated.
4. EXPERIMENT AND RESULTS

A comparison between the performance that is based on the AHP method
and the HTN based on the book shop detail is made [10]. It is evident from
the results that the priority specification is definite in our method. The set of
tasks of the book shop detail and the priority specifications among these
tasks are analysed. The AHP method and Hierarchical Task Plan are utilized
to prioritize the specifications for various tasks. By using AHP method the
percentage of the priority specification is most accurate than the HTN
method. HTN is an approach to automated planning in which the
dependency among actions can be given in the form of networks. The
domains where tasks are naturally organized in hierarchy are an area where
HTN is suitable .The priority specification is more precise while assigning
priorities to those tasks by using the AHP method because of its accuracy.
The task priority graph is shown in Figure 2.
Figure 2. Task priority graph
ISSN: 1694-2108 | Vol. 2, No. 1. JUNE 2013 10

IJCSBI.ORG
5. CONCLUSION
The effort of reasoning, about preferences gives better understanding of the
connection between the stakeholder attitudes and is used to generate a plan
which consists of set of tasks based on preferences. The priorities among
preference goals are defined by the priority specification through the
assignment of numerical weights. A goal model is developed to recognize
alternative solutions of mandatory goals and preferred goals based on
priorities. The priority among the requirements is calculated using
Analytical Hierarchy Process. AHPs pair wise comparison method is used
to assess the relative value of the candidate requirements. The preferences
are analysed based on the prioritization of each task and a definite plan is
generated to view all those tasks according to priority.
6. REFERENCES
[1] Mussbacher G, Amyot D, Araujo Ja, Moreira A, Modelling software product lines
with AoURN, Proc. EA 08, vol 2, pp 12 (8), 2008
[2] Jureta IJ, Borgida A, Ernst NA, Mylopoulos J, Techne, Towards a new generation of
requirements modelling languages with goals, preferences, and inconsistency
handling, Proc of the 18th IEEE RE10, Sydney, Australia, pp 115124, 2010
[3] Giorgini P, Mylopoulos J, Nicchiarelli E, Sebastiani R, Reasoning with goal models,
Proc. ER02, London, UK, pp 167181,2002
[4] Sotirios Liaskos , Sheila A. McIlraith, Shirin Sohrabi, John Mylopoulos,Representing
and Reasoning About Preferences, Springer-Verlag., Vol 16, pp 227- 249, Aug 2011.
[5] Liaskos S, Lapouchnian A, Yu Y, Yu E, Mylopoulos J, On goal-based variability
acquisition and analysis, Proc.RE06, IEEE Computer Society, Minneapolis, pp 79
88, 2006
[6] Azar J, Smith RK, Cords D Value-oriented requirements prioritization in a small
development organization. IEEE Software 24:3237, (2007)
[7] Avesani P, Bazzanella C, Perini A, Susi A, Facing scalability issues in requirements
prioritization with machine learning techniques, Proc (RE05), pp 297305, 2005
[8] Liaskos S, Litoiu M, Jungblut MD, Mylopoulos J Goal Based behavioral
customization of information system, Proc. CAiSE11, London, UK., 2011
[9] Van Lamsweerde A, Requirements Engineering in the year00: a research
perspective, Proc. ICSE00, pp 5-19, 2000
[10] Saaty RW, The analytic hierarchy processwhat it is and how it is used Math
Model 9(3-5): 16117, 1987.
ISSN: 1694-2108 | Vol. 2, No. 1. JUNE 2013 11

IJCSBI.ORG
Unstructured Data Integration

through Automata-Driven
Information Extraction
Maroun Abi Assaf, Kablan Barbar
Dept of applied mathematics
Faculty of Sciences II Lebanese University
Fanar, Lebanon
Youakim Badr
INSA LYON, France
Mahmoud Rammal
Legal Law Center Lebanese University
Sin-Fil, Lebanon
ABSTRACT
Extracting information from plain text and restructuring them into relational databases raise
a challenge as how to locate relevant information and update database records accordingly.
In this paper, we propose a wrapper to efficiently extract information from unstructured
documents, containing plain text expressed with natural-like language. Our extraction
approach is based on the automata formalism to describe the wrapping process running
from text documents to Databases. As usual, relevant information in the text document are
delimited by regular expressions, which define the extracting automaton. Each automaton is
enriched by an output function that automatically generates SQL queries synchronized with
the extracting process in order to insert extracted data into database records. We validate
our extraction approach with automaton-based prototype to extract legal information about
Lebanese official journal decrees and automatically insert them into a relational database.
Keywords
Wrappers, Regular Expressions, Automata, SQL Language, XML.
1. INTRODUCTION
Several approaches to generating wrappers have been introduced in the
literature to extract information from structured, semi-structured and
unstructured documents.
In [2], the authors create wrappers for XML Paragraph-Centric. The
extraction patterns are expressed with regular expressions and translated
into automata. The automata then carry out the information extraction and
produce tuples of values. In [1], the extraction system is based on two
automata. The first automaton decomposes grammatically the given text
into noun groups, verbs groups and others words groups. On the other hand,
ISSN: 1694-2108 | Vol. 2, No. 1. JUNE 2013 1

IJCSBI.ORG
the extraction patterns are encoded into a recognizing patterns automaton
that takes as input all the preceding groups and produces the resulting
entries. In [4], the system runs in two phases, the learning and the testing
phases. The learning phase produces the lexical pattern rules based on a
training collection of reports. The testing phase has three components a
noun phrasing an automaton and a neutral network. The noun phrasing runs
as a syntactical analyzer and produces a list of relevant information. The
automaton makes a matching operation and the neutral network predicts the
matched phrases most possible entity type. In [6] an approach to
automatically locate data-rich regions and to extract relevant attribute-value
pairs of database records from web pages across different sites is presented.
This approach relies on the fact that the attribute-value pairs of the records
usually occurs next to each others in well-designed web pages.
We note that all these approaches do not deal with the context of relevant
information for plain text documents and do not integrate in a unique
formalism both the extracting and the database writing processes.
In contrast to these approaches, we propose an extraction approach, a.k.a.
wrapper, that deal with plain text without a predefined structure (i.e., XML
and HTML) and without a natural language pre-processing (i.e., lexical,
semantic analysis) to identify phase structures and relevant information. Our
extraction approach is based on a formal description of the wrapping
process from text document to databases by means of automata output
functions that generate automatically SQL queries synchronized with the
extracting process. Moreover, on the formal aspect, we express with the
same formalism both the extracting and the database mapping processes.
We particularly rely on the automata and regular expressions formalism to
express the structure of relevant data and locate them. Each relevant
information is described by a regular expression. The sequence separating
two relevant information is also formulated with regular expressions. After
recognizing relevant information, a SQL insert query is generated. For each
relevant information, we specify four elements; its regular expression, the
regular expressions for its preceding and proceeding sequence and the
corresponding SQL query. All, the regular expressions forms the extracting
automaton and the SQL queries are included in the output function of the
automaton. Then, the automata cover both the extracting and the database
writing processes. The research is a first step to build automata generic
system based on attributed grammars like in [3].
The remaining of the paper is organized as follows. Section 2 establishes a
panorama of related works. Section 3 recalls the formalism of automata with
output functions and shows how we translate the relevant information
descriptions into extracting automata. In section 4, we present our
experimental results to extract legal decrees from the text document of the
ISSN: 1694-2108 | Vol. 2, No. 1. JUNE 2013 2

IJCSBI.ORG
Lebanese official journal and insert them into a database. In section 5, we
conclude our work and provide some research trends.
2. RELATED WORKS
Several approaches have been proposed to extract data from text document
using automata. In [1], the authors describe a finite-state processor for
information extraction from real-world text, named FASTUS. FASTUS uses
a pre-defined finite-state machine (Recognizing Phrases Automaton) to
decompose the sentences of the real-world text into noun groups, verb
groups, and several other critical word classes phrases which will form the
candidate list of phrases. Next, the given patterns of interest (extraction
pattern) is encoded into a finite state machine (Recognizing Patterns
Automaton), and the resulted candidate list of phrases is passed as input to
this automaton that identifies phrasal matches with the given patterns of
interest and returns the necessary information. In [2], the authors deal with
the problem of creating a wrapper for XML paragraph-centric documents
named Xtractor. They propose a specification language to write the
extraction patterns, based on Regular Expressions, but is more simple and
easier to read than Regular Expressions themselves. This specification
language includes a set of meta-words that are referenced in a provided
dictionary for the domain of interest, and associates lexemes with these
meta-words. The extraction pattern, which is a regular language, is
translated into an equivalent finite-state automaton in the first phase. Then
in the second phase, the finite-state automaton will carry out the information
extraction, and the result is returned as data tuples of attributes.
In a similar approach, in [4] authors deals with the problem of extracting
meaningful entities, such as person names, addresses, narcotic drugs, or
vehicle names, from free texts of police narrative reports through a neural
network-based entity extractor. The system has three major components;
Noun phrasing, Finite state machine and lexical lookup, and Neural
network. The Noun phrasing component is a form of a syntactical analyzer
that extracts the noun phrases from the reports and will form the candidate
list of relevant information values. The Finite state machine and lexical
lookup component is a finite state machine that finds matches between the
words of the extracted noun phrases list and the items of a provided lexicon
that contains lists of possible values of the entities of interest. The Neural
network component, a feedforward/backpropagation neural network,
predicts the matched phrases most possible entity type. The system also has
two states; a learning state where it identifies the lexical pattern rules
(extraction pattern rules) based on a training collection of reports, where the
entities are manually identified by a human, and saves these rules in the
neutral network as synaptic weight, and a testing state where the system
ISSN: 1694-2108 | Vol. 2, No. 1. JUNE 2013 3

IJCSBI.ORG
extracts the phrases and find item matches from the reports and predicts
their entity type based on the obtained extraction pattern rules from the
learning state.
In [6], authors describe a web information extraction system that extracts
attribute-value pairs, like Product code, Manufacturer, Price, and
Description of products from web pages that present product descriptions.
The system introduces a notion called Structural-Semantic Entropy to
identify if a web page contains a data-rich region or not. The data-rich
region of a web page is the region that contains the actual relevant
information. Structural-Semantic Entropy means the measure of
randomness of the leaf nodes that contain a semantic role of a given non-
leaf node. A leaf node contains a semantic role if it contains a keyword label
associated with the attribute-value pairs. This keyword label list is known
beforehand and given as input to the wrapper. The more this randomness is
higher the more the non-leaf node that is inspected is likely to be a data-rich
node. For every non-leaf node in the DOM tree of the web page its
Structural-Semantic Entropy is calculated, and every node having a
Structural-Semantic Entropy higher than a given threshold, which is
learned by experimentation, is considered to be a data-rich node that
contains a record. Finally, for every discovered data-rich node, the contents
of the next text-nodes of its leaf nodes that are annotated with a semantic
role are usually extracted as the value of the semantic roles (or attributes).
We conclude that most of these representative approaches do not provide
formal description of the whole extracting and database writing processes
and they do not run on non-structured text documents.
3. METHODOLOGY
Automata are used to support syntactical analyzer in the compiling process
of programming languages. They act like a virtual machine to recognize
words belonging to rational languages by passing from one state to another
according to a symbol of the initial alphabet.
Formally, an automaton is a 5-tuple A=< , S, s0, T, > where:
is the alphabet
S is the set of states
s0 is the initial state and an element of S
T is the set of terminal states and a subset of S
is the transition function defined by:
: S x S
The transition function is represented by a graph and leads the recognizing
process. As in [5], we can define a semantic for automata transition. A
semantic has the same domain as the transition function but it associates
ISSN: 1694-2108 | Vol. 2, No. 1. JUNE 2013 4

IJCSBI.ORG
semantic values with each transition. This technique allows the enrichment
of the automaton and the possibility of executing of complementary
semantic operations in parallel with the automaton transition function. Then,
we formalize an extended automaton by a 7-tuple A=< , S, s0, T, , D,
>where , S, s0, T, remains the same as preceding and where D is a
semantic domain and is the semantic function defined by:
: S x D
As application of the data integration process from text document to
database, we consider the data integration process applied to the Lebanese
Official Journal (LOJ) in the Center of Documentation and Research of the
Lebanese University. The main unit in LOJ is the decree that contains
general information like decree number, date,, many 'based on' clauses,
many articles, a 'decrees' clause, a place, a date and many signatures (see
Figure 1). Each week, the center receives an electronic copy of LOJ as a text
document. The relevant information in the decrees are selected and
transferred manually to database by a copy-paste operation from text
document to database. This operation takes too much time and produces
errors. In the center, they want to develop a program that both selects
automatically the relevant information from the decrees of LOJ and insert
them in the database. That is why we are going now to use automata to
formalize the data integration process for LOJ decrees.
The relational schema of the database associated with the LOJ is described
as follows:
decree(decree_id_pk, type, number, title, location, date)
basedon(basedon _id_pk, decree_id_fk, value)
andafter(andafter _id_pk, decree_id_fk, value)
article(article_id_pk, decree_id_fk, value)
signature(signature_id_pk, decree_id_fk, value)
The list of relevant information with their values from the example decree of
Figure 1, and their corresponding regular expressions and database
operations are given in Table 1.
In addition, we give finite state automata where the transition function is
represented by a graph and the output function is formally identified. Each
state in the automata is implemented as a string to which we add the symbol
recognized at each transition. The domain D contains four operations:
1. insert: to insert a table record
2. update: to update a field of a table record
3. concatenate: to concatenate two strings
4. clear: to clear the string state.
ISSN: 1694-2108 | Vol. 2, No. 1. JUNE 2013 5

IJCSBI.ORG
Decree No. 6873
Giving the Ministry of Finance treasury advance to pay the last installment of a
series of differences of ranks and salaries of civilian and military
The President of the Republic,
Based on the Constitution,
Based on the law of public accounting, as amended, particularly Article 203 and
subsequent of it,
Based on the suggestion of the Minister of Finance,
And after the approval of the Council of Ministers on 15/11/2011,
Decrees the following:
Article 1: To give the Ministry of Finance treasury advance worth / 500.000.000.000 /

LP (only five hundred billion pounds).
Article 2: It is not allowed for the party taking the advance, to use it for any other end
other than the one that is given for it.
Article 3: The Director of Treasury, director of the disbursements, director of budget
and expenditure control in the Ministry of Finance, and the director of Administrative
Affairs at the Directorate General of the Ministry of Finance, within their competences,
should pursue the repayment of the advance.
Article 4: This Decree shall be published and reported where needed.
Baabda, the 18th of November 2011

Signature: Michel Sleiman
Issued by the President of the Republic
President of the Ministers Council
Signature: Muhammad NajibMikati
Minister of Finance
Signature: Muhammad Al Safady
Figure 1. A decree of the Lebanese Official Journal
Table 1. Relevant Information
Preceding Wanted Entity

Relevant Value in Following Regular
Regular Regular Database Operation
Information Example Decree Expression
Expression Expression
Insert a new record in

the table decree and
Type Decree [a-zA-Z ]+ No.
update the type field
in this record
Update the number
field of the current
Number 6873 No. [0-9]+ [a-zA-Z]
record in the decree
table
Giving the
Ministry of
Finance treasury Update the title field
The President of the
Title advance to pay [0-9]* [a-zA-Z ]+ of the current record
Republic,
the last in the decree table
installment of a
series of
ISSN: 1694-2108 | Vol. 2, No. 1. JUNE 2013 6

IJCSBI.ORG
Preceding Wanted Entity

Relevant Value in Following Regular
Regular Regular Database Operation
Information Example Decree Expression
Expression Expression
differences of
ranks and salaries
of civilian and
military
the
Insert a new record in
Constitution
the basedOn table
BasedOn the suggestion Based on [a-zA-Z,.0-9 ]+ Based on | And after
and update the value
of the Minister of
field in this record
Finance
the approval of Insert a new record in
the Council of And after | Decrees the andAfter table
AndAfter And after [a-zA-Z,.0-9 ]+
Ministers on the following: and update the value
15/11/2011 field in this record
To give the
Ministry of
Finance treasury
Article [0-9]*: | [a-zA- Insert a new record in
advance worth /
Z]*, the [0-9]{2}th of the article table and
Article 500.000.000.000 Article [0-9]*: [a-zA-Z,.0-9 ]+
(January||December update the value field
/ LP (only five
) [0-9]{4} in this record
hundred billion
pounds).

Article [0-9]*:
This Decree
, the [0-9]{2}th of Update the location
shall be
Location Baabda [a-zA-Z]+ (January||December field in current record
published and
) [0-9]{4} of the decree table
reported where
needed.
the [0-9]{2}th of Update the date field
the 18th of
Date [a-zA-Z]+, (January||Dece Signature: in current record of
November 2011
mber) [0-9]{4} the decree table
Michel
Sleiman, Insert a new record in
Muhammad the signature table
Signature Signature: [a-zA-Z ]+ Issued by | Minister
NajibMikati, and update the value
Muhammad Al field in this record
Safady
ISSN: 1694-2108 | Vol. 2, No. 1. JUNE 2013 7

IJCSBI.ORG
The automata associated with the decree relevant information are:
FSA( Type ):
a-z|A-Z| |
q0 a-z|A-Z| | q1 No. q2
Figure 2. Type Finite State Automaton

D = {Insert(qi, 'decree', 'type'), Concatenate(qi, c) }
:QxD
(q0, c) = Concatenate(q0, c) where c a-z|A-Z| |
(q1, No.) = Insert(q0+q1, 'decree', 'type')
This automaton will extract all the alphabetical characters including a space
character between the start of the decree and a No. keyword and insert
them as the value of the type attribute of a new record in the decree table.
FSA( Number ):
0-9
q0 No. q1 0-9 a-z|A-Z

q2 q3
Figure 3. Number Finite State Automaton
D= { Update(qi, 'decree', 'number'), Concatenate(qi, c) }

:QxD
(q0, No.) =
(q1, c) = Concatenate(q1, c) where c 0-9
(q2, a-z|A-Z) = Update(q1+q2, 'decree', 'number')
This automaton will extract all the numerical characters between the No.
keyword and any alphabetical character and update the number attribute of
the last inserted record of the decree table with this value.
ISSN: 1694-2108 | Vol. 2, No. 1. JUNE 2013 8

IJCSBI.ORG
FSA( Title ):
0-9 a-z|A-Z| |
The President
of the
0-9 a-z|A-Z| |
q0 q1 q2 Republic, q
3
Figure 4. Title Finite State Automaton
D = { Update(qi, 'decree', 'title'), Concatenate(qi, c) }

:QxD
(q0, 0-9) =
(q1, 0-9) =
(q2, c ) = Concatenate(q2, c) where c a-z|A-Z| |
(q2, The President of the Republic,) = Update(q1+q2, 'decree', 'title')
character between the delimiter formed by any numerical characters and the
keywords The President of the Republic,, and updatethe title attribute of
the last inserted record of the decree table with this value.
FSA( BasedOn ):
a-z|A-Z| |0-9
Based on And after

q0 q1 a-z|A-Z| |0-9 q2 q3
Basedon
Figure 5. BasedOn Finite State Automaton
D = {Insert(qi, 'basedon', 'value'), Concatenate(qi, c),Clear(qi)}

:QxD
(q0, Based on) =
(q1, c) = Concatenate(q1, c) where c a-z|A-Z| |0-9
(q2, c) = Concatenate(q2, c) where c a-z|A-Z| |0-9
(q2, Based on) = Insert (q1+q2, 'basedon', 'value'); Clear(q1); Clear(q2)
(q2, And after) = Insert(q1+q2, 'basedon', 'value')
ISSN: 1694-2108 | Vol. 2, No. 1. JUNE 2013 9

IJCSBI.ORG
This automaton will extract all the alphanumeric characters including a
space character between every Based on keyword and another Based on
or And after keywords, and insert them as the value of the value
attribute of a new record in the basedon table.
FSA( AndAfter ):
a-z|A-Z| |0-9
Decrees the
following:
q0 And after q1 a-z|A-Z| |0-9 q2 q3
And after
Figure 6. AndAfter Finite State Automaton
D = {Insert(qi, 'andafter', 'value'),Concatenate(qi, c), Clear(qi)}

:QxD
(q0, And after) =
(q2, And after) = Insert (q1+q2, 'andafter', 'value'); Clear(q1); Clear(q2)
(q2, Decrees the following:) = Insert(q1+q2,'andafter', 'value'); Clear(q1);
Clear(q2)
space character between every And after keyword and another And
after or Decrees the following: keywords, and insert them as the value of
the value attribute of a new record in the andafter table.
FSA( Article ):
ISSN: 1694-2108 | Vol. 2, No. 1. JUNE 2013 10

IJCSBI.ORG
D = {Insert(qi, 'article', 'value'), Concatenate(qi, c), Clear(qi)}
:QxD
(q0, Article) =
(q1, 0-9) =
(q1, : ) =
(q2, c) = Concatenate(q2, c) where c a-z|A-Z| |,|.
(q3, c) = Concatenate(q3, c) where c a-z|A-Z| |,|.
(q3, Article) =
(q4, 0-9) =
(q4, :) = Insert (q2+q3, 'article', 'value'); Clear(q2); Clear(q3)
(q3, a-z|A-Z) =
(q5, a-z|A-Z) =
(q5, , the) =
(q6, 0-9) =
(q6, th of) = Insert(q2+q3, 'article', 'value');
space, comma, and dot characters between every Article n: keyword
(where n is a number) and another Article n: keyword or any word
followed by a date delimiter, and insert them as the value of the value
attribute of a new record in the article table.
FSA( Location ):
a-z|A-Z
This Decree shall
be published and
reported where
needed. a-z|A-Z q2 , the
q0 q1 q3
Figure 8. Location Finite State Automaton
D = { Update(qi, 'decree', 'location'), Concatenate(qi, c) }

:QxD
(q0, This Decree shall be published and reported where needed.) =
(q1, c) = Concatenate(q1, c) where c a-z|A-Z
(q2, c) = Concatenate(q2, c) where c a-z|A-Z
(q2, , the) = Update(q1+q2, 'decree', 'location')
This automaton will extract the word between the two sequences This
Decree shall be published and reported where needed. and , the
keywords and update the location attribute of the last inserted record of the
decree table with this value.
ISSN: 1694-2108 | Vol. 2, No. 1. JUNE 2013 11

IJCSBI.ORG
FSA( Date ):
a-z|A-Z 0-9 0-9
January||
, the December Signature:
a-z|A-Z
q0 q1 q2 q3 th of q4 q5 q6
Figure 9. Date Finite State Automaton
D = { Update(qi, 'decree', 'date'), Concatenate(qi, c) }

:QxD
(q0, a-z|A-Z) =
(q1, a-z|A-Z) =
(q1, , ) =
(q2, the) = Concatenate(q2, the)
(q3, th of) = Concatenate(q3, th of)
(q4, w) = Concatenate(q4, w) where w January|| December
(q5, Signature:) = Update(q2+q3+q4+q5, 'decree', 'date')
This automaton will extract the date according to the given format and that
is located between any word and a Signature: keyword, and update the
date attribute of the last inserted record of the decree table with this
value.
FSA( Signature ):
a-z|A-Z| |
Issued
q0 Signature: q1 a-z|A-Z| | q2 by|Minister q3
Figure 10. Signature Finite State Automaton
D = { Insert(qi, 'signature', 'value'), Concatenate(qi, c) }

:QxD
(q0, Signature:) =
(q2, Issued by|Minister ) = Insert(q1+q2, 'signature', 'value')
character between every Signature: and an Issued by or Minister
ISSN: 1694-2108 | Vol. 2, No. 1. JUNE 2013 12

IJCSBI.ORG
keywords, and insert them as the value of the value attribute of a new
record in the signature table.
4. RESULTS
In order to validate our extraction approach, we have developed a .NET-
based prototype, including two modules, the extracting module and the
database writing module. The extracting module is the implementation of
the regular expressions of all relevant information with their preceding and
following sequences and also includes the output function for the SQL
operations. It extracts relevant information values from decrees of the text
document of the Lebanese official journal, and inserts these values in the
SQL queries that are saved in an XML document. The database writing
module takes as input the XML document for the SQL operations, selects a
SQL query, makes a connection to the database and finally executes this
query.
We have tested our prototype application on a sum of 100 decrees from
several LOJs. Then we manually checked the database entries with the
initial 100 decrees texts for consistency and errors. We found that out of the
100 decrees, 98 decrees where discovered and extracted, and out of the 98
extracted decrees, 5 errors were found in some of the attributes. To measure
the correctness of our prototype, we calculate two statistical classifications
used in pattern recognition and information retrieval, the Precision (also
called positive predictive value) which is the fraction of retrieved instances
that are relevant, and the Recall (also known as sensitivity) which is the
fraction of relevant instances that are retrieved. Based on the previously
found numbers, the number of relevant instances is equal to the 98 extracted
decrees minus the number of extracted decrees with errors which is 5,
giving us a 93 for the number of relevant instances. This will result in a
Precision that is equal to 93/98 which is approximately 95%, and a Recall
that is equal to 93/100 which is 93%.
5. CONCLUSIONS
In order to deal with information extraction from plain text without a
predefined structure (i.e., XML and HTML) and without a natural language
pre-processing, we have presented an automata-based model to describe a
wrapper that extracts meta-data from a text document containing the decrees
of the Lebanese Official Journals, generates the corresponding SQL writing
queries, and executes them in order to insert these extracted data into a
relational database. In this model, we specify with the automata formalism
both the extracting and the database writing processes. At present time, we
have developed a prototype extractor by directly coding the Finite State
ISSN: 1694-2108 | Vol. 2, No. 1. JUNE 2013 13

IJCSBI.ORG
Automaton with the Output Function. The prototype extractor showed
prominent results in extracting correct meta-data about the articles and
successfully inserting them in their correct place in the relational database.
In our future works, we would like to target the extracting process by
considering conflicts when relevant information occurs in a wrong place.
Moreover, we are investigating a new specification language for automata
with output function in order to build a generic system that produces
automatically the automata program or the extraction and writing programs.
6. REFERENCES
[1] D. E. Appelt, J. R. Hobbs, J. Bear, D Israel and M. Tyson, FASTUS: A Finite-state
Processor for Information Extraction from Real-world Text. IJCAI 1993: Chambry,
1993.
[2] Y. BADR, Xtractor: A Light Wrapper for XML Paragraph-Centric Documents.
Proceedings of the 1st International Conference on Signal-Image Technology and
Internet-Based Systems, SITIS 2005, 2005.
[3] K. Barbar, Automatic generation of XML-based editors for autonomic systems,
International Journal of Autonomic Computing, v.1 n.3, p.246-262, May 2010
[4] M. Chau, J. J. Xu, H. Chen, Extracting Meaningful Entities from Police Narrative
Reports. National Conference on Digital Government Research 2002, 2002.
[5] Knuth, D., "Semantics of context-free languages", Math. Systems Theory 5, 127-145,
1968.
[6] X. Zheng, Y. Gu, Y. Li, Data Extraction from Web Pages Based on Structural-
Semantic Entropy. Proceedings of the 21st international conference companion on
World Wide Web. Pages 93-102. ACM New York, NY, USA, 2012
ISSN: 1694-2108 | Vol. 2, No. 1. JUNE 2013 14

IJCSBI.ORG
Wireless Solution for Water Saving In

Agriculture Using Embedded System
Venkata Narayana Eluri, K. Madhusudhana Rao, A. Srinag
Electronics and Communication Engineering,
KKR & KSR Institute of Technology and Sciences,
Guntur, Andhra Pradesh, India.
ABSTRACT
Agriculture is a source of livelihood of majority Indians and has great impact on the
economy of the country. In a country like India, where climatic conditions vary
substantially and irrigation facilities are poor. Agriculture is timely and sufficient supply of
water. Water pumps are crucial in agriculture where electricity is indeed. The frequent,
intermittent, low voltage supply of power to the agriculture sector has caused problems to
the farmers who are spending their time monitoring the supply of power without which no
progress of their work. The power supply with frequent power cuts have not only lowered
the efficiency of farmers but also have led to the frustration of the farmer to give up
agriculture and move to urban areas for better prospects in the globalized world. In this
paper we presented a system which shows, how mobile communication can benefit
millions of farmers in rural India by providing a solution for the irrigation problems caused
by intermittent electrical power supply. Information is exchanged in form of
messages/miscalls between the system and the user cell phones. The system is based on
ATMEGA32 micro-controller and includes protection against fluctuations in power supply.
For measuring time and temperature RTC DS1307 and DS18S20 are used. By using this
system the hardships of farmers relating water distribution can be relieved.
Keywords
AT command supporting GSM mobile phone, ATMEGA32 Micro controller, Relays,
Sensors, LCD for monitoring the current reading of all the parameters.
1. INTRODUCTION
Even in the modern era of industrialization, agriculture plays a very
significant role in the overall socio-economic development of India. India
has an agriculture based economy. 43% of Indias territory comes under
agricultural lands. Agriculture along with other related fields like forestry
and logging provides employment to 52% of Indias population. Agriculture
also accounts for 8.56% of the countrys total exports. According to a
survey made in 2007, agriculture accounts for 16.6% of Indias Gross
Domestic Product. Hence, it is agriculture that is the most influential field
as compared to others in India. This importance on agriculture leads to an
emphasis on better agricultural practices. The underground water level is
slowly falling down and forests are being cut which reduces the rainfall as
well. With increasing area available for cultivation and the need for
increasing the productivity from the farm land, there is a growing need for
ISSN: 1694-2108 | Vol. 2, No. 1. JUNE 2013 1

IJCSBI.ORG
electrical energy for irrigation. The generation of electricity is not growing
proportionately to the demand.
The supply to agriculture is limited to few fixed hours throughout the day.
In a sample survey of village Komarolu, located at distance of 17kms from
Giddalur (Taluka in Prakasam District, Coastal Region, Andhra Pradesh
State, India), it was observed that 3-phase supply is normally available for 7
hours a day from 2.00 am to 9.00 am for a week and from 11.00 am to 6.00
pm for next week and changes alternately. Due to increase in demand of
electricity the schedule given was not followed and sometimes power given
at night times and fluctuations in voltage level leads to motor damage[1]. So
farmers, after carrying strenuous physical activities in farm during day, have
to return back to their farms in night to carry water distribution using 3-
phase motor pumps. Agriculture receives power mostly during mid night
(off-peak) as this reduces the cost of electricity supply for the transmission
and Distribution Company. The farmers have to be on their guard all the
time due to the unpredictable nature of supply of electrical energy. And the
farmers have to switch on their motor after electricity supply resumes. The
reduced amount of yielding, man power wastage, and idle state of
equipment can observed in its results.
Due to the fast development in tele-communication technologies, it is
believed that wireless solution for irrigation in agriculture. This system has
fully utilized wireless sensor network, Global System for Mobile
Communication (GSM) and short message service (SMS)/missed call to
carry out data from the sensors to computers or directly alert the farmers
through their mobile phone and to control the remote watering process [2-
4], also through the mobile phone. This practice eliminates the use of wired
technology, improves old method of collecting data in farming areas and
allows farmers to control their sprinklers remotely. Also, It has been
observed that source of electrical energy generation is slowly depleting
using solar energy. We are going to discuss an example of how the mobile
technology can benefit millions of farmers by providing a solution for the
irrigation problems. Simple cell phones having just voice call and
messaging facility or cell phones with non-working display can be easily
adapted for remote control applications.
2. SYSTEM DESCRIPTION
A system is developed for optimum water distribution in the fields through
motor pump. The crop yield is maximized to a great extent by providing
proper amount of water at suitable time intervals based on climatic
conditions. The system offers attractive features like automatic control
based on parameters specified through keyboard /SMS/ number of miscalls;
provides protection against single phasing, over-current, over-voltage, dry
ISSN: 1694-2108 | Vol. 2, No. 1. JUNE 2013 2

IJCSBI.ORG
running and probable bearing faults; alerts users in case of abnormal
conditions like power failure, dry-running, etc. and provides audible
indication through buzzer/ miscall on completion of task.
The figure 1 shows the Block diagram of the scheme. Using keyboard the
parameters of the system can be set or received in form of SMS/number of
miscalls in specified duration from user mobile through serial cable
connected to control system cell phone (Model Nokia 6610). Based on
commands received microcontroller system sends signals to switch on / off
motor through Starter using relays controlled by its ports. Three phase 5
hp,7 hp,10.5 hp induction motor working on Direct-on-line Starter are
chosen for agriculture purpose.
3 PHASESUPPLY
R Y B
CELLULAR
USER LINK ATMEGA32 STARTER
SYSTEM
MOBILE MICROCONTROLLER (MOTOR
MOBILE
BASED SYSTEM CONTRACTER)
LCD RTCSENS KEY

MOTORPUMP
DISPLAY ORS BOARD
Figure 1. System Block Diagram

2.1 Cell phone Interface:
Cell phone 6610 is connected to AVR Microcontroller board through
RS232C serial interface. AT commands are sent by sending text strings A,
T, along with specified command strings through serial port to cell phone
and are executed on receipt of carriage return [5-6]. The result codes are
sent by cell phone to system (TE) to indicate the status after execution of
command.
1) SMS Approach: SMS is store and forward way of transmitting messages
to and from cell phones. The major advantage of using SMS is provision of
intimation to the sender when SMS is delivered at the destination and ability
of SMSC to continue efforts for delivery of message for the specified
validity period if network is presently busy or called user is outside the
coverage area. Using CMGS command the text message is sent to cell
phone. CNMI command is used to indicate to TE about the receipt of
incoming SMS message from the network. On receipt of the SMS message,
unsolicited result code +CMT is obtained from which text message is
ISSN: 1694-2108 | Vol. 2, No. 1. JUNE 2013 3

IJCSBI.ORG
extracted and checked with predetermined format, which includes
password, desired time or ON/OFF commands or status query.
Microcontroller carries out for the valid control message. In this application,
any incoming SMS message is directly routed to micro-controller (TE) and
any outgoing text message is directly sent by micro-controller to designated
cell phone number without being stored in control system cell phone
memory. As a result, phone memory is not inundated with messages in spite
of many messages being transferred.
2) Miscall Approach: The operational cost of communication between user
and control system cell phones is further reduced by using concept of
miscall where in no charges are incurred by using only ring signal for
information transfer. Miscalls are treated in two situations one is calling
party disconnects after receiving ring tones and second one when called
party does not respond to call within mentioned 5 minutes. The system cell
phone is designed to send specified number of miscall(s) within five
minutes duration to user cell phone to report various conditions as shown in
Table 1. Similarly, user cell phone sends commands to system cell phone by
making specified number of miscalls as shown in Table 2.
Table 1. Messages based on missed calls from system cellphone
No. of missed calls in 5 min. Message indication
01 No Power
02 Power failure
03 Resumption after normal conditions
04 Task completion
05 Probable motor faults/dry running
Table 2. Command based on missed calls from user cell phone

No. of missed calls in 5min. Message indication
01 Switch on
02 Switch off
03 Increase on time by 0ne hour
04 Decrease on time by one hour
05 Present status of the motor
This novel concept of miscalls results in substantial savings without

comprising the utility of system. Another advantage of miscall over SMS is
that during night time, ringing tone can easily wake-up farmer to carry out
necessary arrangement like shifting pipes to new locations, etc..CLIP
ISSN: 1694-2108 | Vol. 2, No. 1. JUNE 2013 4

IJCSBI.ORG
command is used in recognizing particular saved recipient. Micro-controller
checks incoming number with user cell phone when reception of unsolicited
code RING along with CLIP occurs. Micro-controller waits for five minutes
duration to check total number of calls and carries out specified task if
match founds. Similarly, microcontroller carries out number of voice calls
from control system cell phone to user cell phone within five minutes time
duration using ATD command. And SMS mode activated when the BUSY
or ERROR occurred while making voice call.
3. MICROCONTROLLER SYSTEM
AT Mega32 microcontroller has RISC architecture with 32 KB of in-system
programmable Flash, 1k E2PROM, 2k SRAM, 32-bit General purpose I/O,
8channel 10-bit ADC, TWI, USART, SPI, JTAG interface support etc [7-8].
Ponyprog software was used for flash programming [9]. The software was
developed in C language using GCC compiler.
3.1 Interfacing
Interfacing diagram of micro-controller system is shown in Fig. 2. 8-bits of
Port A are configured as analog inputs ports. PA0-PA2 bits monitor the
present values of supply phase voltages. Water level of well indicated by
port A 3 and 4 pins while PA5 to PA7 are used to sense whether water has
reached the desired final destination of the regions. PB0-PB5 pins are
connected to 2 16 characters LCD display in 4-bit data length mode. Two
relays are controlled by upper 2 bits of Port B. Start (green) pushbutton of
starter for automated starting of the pump from micro-controller board is
connected to one relay while other relay is connected in series with stop
(red) pushbutton for stopping the motor pump from micro-controller. Upper
4 bits of Port C and upper 4 bits of Port D are used to interface 4x4
keyboard matrixes. DS1307 (RTC) is chosen for implementation of timing
applications. It is connected through TWI interface (I2C) i.e. PC0 (SCL)
and PC1 (SDA) pins. Two temperature sensors (DS18S20) are used having
single wire interface for connectivity. PC2 bit is used for single-wire
interface. Internally RxD (PD0) and TxD (PD1) are connected to 9-pin
RS232 female connector through MAX 232 IC for TTL-
RS232Csignaltranslation.
3.2 Real Time Clock
The DS1307 serial real-time clock (RTC) is a low power, full binary-coded
decimal (BCD) clock/calendar [10]. Address and data are transferred
serially through an I2C, bi-directional bus (TWI). The clock/calendar
provides all timing information from seconds to years. With the help of
keyboard and LCD display, present date and time are written into
corresponding internal memory locations of this IC using I2C protocol.
ISSN: 1694-2108 | Vol. 2, No. 1. JUNE 2013 5

IJCSBI.ORG
Whenever on-time duration for pump is specified, the two registers of AT
Mega32 are used to store time duration in minutes and hours.
R PD7
VOLTAGE 4X4
Y ADC0 (PA0) PD4
SENSING
B BLOCK ADC2 (PA2) PC4 KEY BOARD
PC7
WATER ADC3 (PA3) TXD (PD1)

LL TXD(P MAX232
LEVEL
HL ADC4 (PA4) RXDD1)
(PD0)
SENSOR
RXD
+5V
(PD0)
WATER
R1
REACH ADC (PA5)
R2 R
SENSOR ADC (PA7) C
R3
3
ATMEGA32
DS RESET
PC2
18S20 RES
ET
PB7 REL1 MOTOR
DS 1307 SCL (PC0) PB6 REL2 STARTER
SDA (PC1)
XTAL1
XTA PB5 D4
C1
L1 PB4 D5
PB3 D6 LCD
PB2 D7 DISPLAY
C2 PB1 ENIN
XTAL2 PB0 RS
GND
XTA
L2
Figure 2. Microcontroller system interfacing
GND
The pump is switched on through relays using ports of micro-controller.
After passage of every minute (known through RTC) relevant registers of
microcontroller are updated if normal conditions exists and when their
values reach null, by using PB7 bit the pump is switched off. The
occurrence of abnormal conditions like unbalanced phase voltages, dry-
running, etc. causes pump to be switched OFF and the counting is
temporary stopped and is resumed on restoration of normal conditions. Thus
it is ensured that proper quantity of water is distributed by keeping pump
ON for specified time duration under normal conditions and user is
informed about restoration of normal conditions through miscalls.
3.3 Phase Voltage Measurement
In order to measure phase voltages, three transformers of equal ratings (6-0-
6) are used to step down voltage. These voltages are converted into
appropriate dc levels at analog inputs of micro-controller. Phase voltages
ISSN: 1694-2108 | Vol. 2, No. 1. JUNE 2013 6

IJCSBI.ORG
are read at regular intervals by micro-controller. The values of phase
voltages are compared with one another. If the result of any comparison
exceeds specified value, signal is sent to switch off the pump along with
error message indicating unbalanced phase voltage condition and conveying
this information to user cell phone through miscalls. Internal 10-bit ADC of
micro-controller is configured to work in left justified format and only most
significant 8-bit values are used for comparison.
3.4 Temperature measurement
There has been tremendous research in fault diagnosis and protection
schemes for induction motor [11]. The basic over-current protection for
motor is provided by bimetallic strip of starter. Maximum reliability ensured
by mounting temperature sensor on body of motor and another temperature
sensor is mounted at a suitable location to measure ambient temperature.
Whenever temperature difference between the two sensors exceeds
specified safety limit (250C), signal is sent to switch off pump along with
error indication to LCD display and conveying miscalls to user cell phone to
indicate probable fault occurrence. This arrangement ensures that
catastrophic event like burning of motor due to any fault like over-current,
bearing blockage and insulation failure are avoided. And preventive
maintenance is carried out at substantially lower cost. The chosen
temperature sensor is DS 18S20 manufactured by Dallas Semiconductor
(Maxim) [12-13].It has operating temperature range of 55oC to +125oC. A
major advantage associated with this sensor is the availability of output
directly in digital form obviating the need of analog to digital conversion.
Moreover, this sensor provides inherent error-detection capability through
CRC technique.
3.5 Water Level Sensor
In order to prevent dry running of motor and allow automatic restart of
motor when sufficient water level is regained, three wires are inserted into
the well. One wire (GND) is inserted at bottom of well while second wire is
adjusted just above foot-valve of the suction pipe of motor pump (LL) and
third one at suitable level above second wire (HL). Microcontroller switches
OFF the motor and sends miscalls indicating empty well (dry running)
conditions whenever water falls below LL level. Whenever water rises
beyond HL level and sends miscalls indicating resumption of task, the
micro-controller switches ON the motor again. The flexibility of system can
be increased by using water level sensor and prevents the damage of motor.
ISSN: 1694-2108 | Vol. 2, No. 1. JUNE 2013 7

IJCSBI.ORG
Figure 3. Implementation
3.6 Water Reach Sensor
There are two approaches for water distribution in this system. One method
is to specify the ON time duration of pump and second approach is to
specify the area of water distribution before stopping motor. First approach
is more suitable for sprinkle-based irrigation system while second approach
is chosen for ground-level water irrigation. For ground level water
irrigation, two wires are extended to extreme end of region where water is
to be reached. In this scheme, three such regions are presently supported
and one end of wire of these regions (R1, R2, R3) are connected to PA5,
PA6 and PA7 port bits of microcontroller using pull-up resistors of 100k
while other ends are connected to GND.
4. CONCLUSION AND RECOMMENDATIONS

By using this system optimum water distribution can be achieved. All the
electric defects are vanished and the maximum utilization of power supply
achievement is possible with the proposed system and during normal
conditions the motor will be restarted automatically. The system result in
uniform distribution of water at given time periods, minimizing man power
effort, water wastage reduction, minimization of occurrences of motor faults
and intimation to user about the completion of task through miscalls/SMS
system proves to be great boon to farmers whose pump sets are located far
away from their homes and intimation about any abnormal conditions.
With introduction of MMS message support, it is possible to capture images
from field using higher end cell phone and disease-pest control management
can be carried out by analysis of these images by agriculturists. Various
parameters such are warmness, moisture in the weather, etc., can be noted at
regular intervals on daily basis and time duration of motor, amount and type
of fertilizers and pesticides can be decided based on analysis of acquired
data. The technological assistance to farmers can tremendously boost the
ISSN: 1694-2108 | Vol. 2, No. 1. JUNE 2013 8

IJCSBI.ORG
productivity of food grains and bring prosperity to this hardworking
population ensured with the usage of proposed system.
5. REFERENCES
[1] Ilonen J. Kamarainen J. K, Lindh T, Ahola J, Kalviainen H andPartanen,
Diagnosis tool for motor condition monitoring. IEEE Trans. Ind. Application,
Vol 41, No. 4, 2005 pp. 963-971.
[2] KuniakiUmino, Yasuhiro Ohyama, Jin-Hua She and Hiroyuki Kobayashi,
Remote Controlled Embedded System, 4th China-Japan International
Workshop on Internet Technology and Control Applications, Hunan, China, 21-
26 October, 2005.
[3] A Weaver J Luo, and S Zhang, Monitoring and Control using Internet and
Java, Proc. 25th Annual Conference on IEEE Industrial Electronics IECON99,
Vol. 3 , pp 1152-1158, 1999.
[4] Trohandl C, Proske M &Elmeureich W, Remote Target Monitoring in
Embedded System Lab courses using a Sensor Network, Proc. 32nd Annual
Conference on IEEE Industrial Electronics IECON 2006, 6-10 Nov 2006, pp
5433-5438.
[5] AT Commands Set for Nokia GSM and WCDMA products,Version 1.2, July
2005, available online: http://forum.nokia.com.
[6] ETSI TS 127 007 V5.3.0 (2003-03) Digital cellular telecommunications system
(Phase2+); Universal Mobile Telecommunications System (UMTS); AT
command set for 3G User Equipment (UE) (3GPP TS 27.007 version 5.3.0
Release 5, available online: http://www.etsi.org.
[7] ATMELAVRATMega32 Datasheet, available online:
http://www.atmel.com/dyn/resources/prod_documents/2503S.pdf.
[8] Richard Barnett, Larry OCull and Sarah Cox, Embedded C Programming and
Atmel AVR, Delmar-Cengage Learning India Pvt. Ltd., 2007 Edition (India).
[9] Ponyprog software, available online: http://www.lancos.com/
[10] DS1307 64 x 8, Serial, I2C Real-Time Clock Datasheet, availableonline:
http://pdfserv.maxim-ic.com/en/ds/DS1307.
[11] Benbouzid, M.E.H, A review of induction motors signature analysis as a
medium for faults detection, IEEE Trans. Ind.Electronics, Vol 47, No.5, 2000
pp. 984-993.
[12] DS18S20 High-Precision 1-Wire Digital Thermometer, available online:
http://pdfserv.maxim-ic.com/en/ds/DS18S20.
[13] Overview of 1-Wire Technology and Its Use, available
online:http://www.maxim-ic.com/an1796.
ISSN: 1694-2108 | Vol. 2, No. 1. JUNE 2013 9

IJCSBI.ORG
Upcoming Trends of Virtual

Experiments for Laboratories
Bhaskar Y. Kathane
PCD ICSR, VMV College Campus, Wardhaman Nagar, Nagpur (MS), India
Pradeep B. Dahikar
Kamla Nehru Mahavidyalaya Sakkardara, Nagpur (MS), India
Satish J. Sharma
Department of Electronics and Computer Science, R.T.M. Nagpur University, Nagpur
ABSTRACT
The scope of this paper includes development and implementation of virtual laboratories.
Virtual Experiments can be performed using the concept of virtual Intelligent SoftLab
(VIS). The virtual experiments described here will help students to perform it at anytime
and anywhere. The model accepts inputs using virtual instruments and observed virtual
results on the screen. There is a facility for constructing an experimental circuit, change the
input values and observed outputs respectively. The graphical and simulated effects of
virtual experiments are visible on the screen. SoftLab will help Electronics, Computer
Science and Engineering students perform and practice experiments to improve their
understanding of the subject.
Keywords
SoftLab, Virtual, VIS Model, Virtualization, Virtual Lab.
1. INTRODUCTION
For the last two decades, Electronics and Computer Science technology
have been among the fastest growing areas in the research field. Due to this,
there is a traditional laboratory in keeping up with some of the modern
industry's requirements. The concept of traditional real laboratories in
educational institutions has many limitations. The availability of resources
may be too limited to provide basic as well as advanced laboratory
equipments, and there may be a paucity of qualified faculty members. Even
if the facilities are there, in many cases a student may end up as a spectator
in the conduct of experiments. Students are not free to do experiments
according to their own schedules, as the time slot for an experiment may be
limited to the usual working hours. Numbers of subjects like Physics and
Electronics students have to perform a large number of experiments in an
academic year. Many times students do not get time to repeat experiments
which they have performed during the session. Also many of the
laboratories lack of resources to perform experiments in which sophisticated
ISSN: 1694-2108 | Vol. 2, No. 1. JUNE 2013 1

IJCSBI.ORG
instruments are required. The new era of education demands a revolution
and modernization in techniques for different subjects. One gap in subjects
having component of practice is the lack of a complete set of e-experiments
that is a full scale virtual laboratory, instead of little virtual experiments.
Science subjects always have a component of practical. In subjects like
Physics and Electronics students have to perform a large number of
experiments in an academic year. Many times students do not get time to
repeat experiments which they have performed during the session. Also
many of the laboratories lack of resources to perform experiments in which
sophisticated instruments are required. Therefore to provide access to
laboratory experiments, anytime anywhere, concept of virtual laboratory is
developed. This virtual laboratory can cater to students at undergraduate
(UG) and post graduate (PG) levels. Some softwares like Mat Lab and Lab
VIEW are available for simulation of experiments and for other purposes.
However, these softwares are generally available only in big
Institutes/Laboratories and the student can use them only during college
hours. It is therefore decided to develop software for performing individual
experiments virtually on a computer screen. An attempt is made to develop
software for electronics experiments from basic to advance level. The
facilities of the laboratories in educational institutions are generally
insufficient when the number of students is considered. Implementation of a
laboratory to meet the requirements has a very high price.
The Virtualization of experiments is one of the most efficient ways for the
modernization of traditional laboratories. Virtualization is basically the
conversion of real experiments into virtual experiments with the help of
information and communication technologies (ICT), which provide a real
laboratory environment and 'feel' to perform the experiment. The basic goal
for Virtualization is to provide a facility to perform the experiment by using
either the Internet or s specific computer program. This can provide a highly
interactive and powerful learning environment for the engineering and
science disciplines and enables a learner to select and control all the related
parameters of the experiments. A complete graphical interface with the
adequate learning components and a scientific approach can provide the
'feel' of performing experiments efficiently. The Virtualization of an
experiment and its application by Internet based remote techniques can
provide a relevant and meaningful practical learning experience. This
relatively new concept of Virtualization is cost effective and there is no time
bound for the users as experiments could be accessed "24X7" anywhere any
time. Virtual experiments can be used extensively for teaching, e-learning,
and other computer-based education [1].
ISSN: 1694-2108 | Vol. 2, No. 1. JUNE 2013 2

IJCSBI.ORG
2. LITERATURE REVIEW
In 1990, Mercer, Lynn from Canada University of Regina proposed the
concept of the Virtual Lab and built a simple one. During the next 20 years,
a lot of researchers further built and enriched the concept of Virtual Lab.
Experimental simulations have appeared with discipline specific texts, such
as Jones and Childers (2001), and Christian and Belloni (2001), or on the
World Wide Web at sites such as the Physical Sciences Resource Center
[2]. Many of these simulations are implemented using Java applets, and are
well designed and executed. Other online experimentation includes the
Stevens Institute of Technology in their Remote Dynamical Systems
Laboratory [3], which incorporated several online experiments but is
subject to limited use and limited data. Mercer University has implemented
the Online Interactive Chaotic Pendulum [4]. The site has an exemplary
interface and data presentation, but doesn't provide easy access to data.
Several mechanical engineering courses have online experimentation,
including Curtin University [5]. These sites tend to be highly technical and
their online availability appears to be limited. Other organizations that have
executed online experimentation include the Advanced Liquid Crystalline
Optical Materials (ALCOM) Science and Technology Center [6] of the Kent
State University. Tan, Tang and Paterson at Queensland University of
Technology had a project on developing a web based remote controlled
mechanical vibrations laboratory via the Internet.
The system allowed flexibility for students to access a range of laboratory

experiments at any time and any place where there is an Internet connection.
Coito et al. (2007), Bauer and Fedk (2007) and Bachnak et al. (2003)
discussed the ability to control and monitor processes from remote locations
through a PC-based data acquisition for online and off line analyses.
Advances in networking technologies and development of measurement
hardware and software have turned PCs into platforms capable of
continuous remote monitoring using the Internet (Fountain & Wright, 2000).
Software programs that facilitate the developers of such applications are
available. A user can publish data on the web with little or no programming
experience. The software creates user interfaces on a Tiwari and Singh 673
web page to give the user access to the system (Travis, 2000). Antunes,
Paulino & Piteau (1998) developed predictive methods to remotely analyze
heat exchanger tube responses and wear for realistic multi-supported tubes
and flow configurations. Buckman (2000) presented an introduction to the
virtual instrumentation to increase the ease of use for students, and the
capability to add new measurements that were otherwise unaffordable.
Overstreet and Tzes (1998) describes the design and development of generic
virtual instrument used for real time experimentation in the control
engineering laboratory of their institute in a remote-access environment.
ISSN: 1694-2108 | Vol. 2, No. 1. JUNE 2013 3

IJCSBI.ORG
Salzmann (1999) proposed the steps of remote, real time control over the
Internet, and demonstrated the feasibility of using a distributed online
laboratory to complement and enhance the traditional laboratory. Ertugrul
(2000) reported such software tool applications, and his paper aimed to
provide some background knowledge about the tools and to address
common problems encountered by users. Resendez and Bachnak (2003) and
Bachnak et al. (2003) showed that such software tools could be used to
perform data acquisition and remotely control hardware devices through the
Internet. Trevelyan (2002) aimed to provide a cost effective online
laboratory to staff and students. Their first project was a tele-robot, which
was written in special purpose software written in C, C++ and Java.
Trevelyan (2004) reviewed the entire principal lessons their group learned
since 1994 and briefly described Telelabs, a cost-effective framework to
provide an extendable series of online labs that could be sustained from
normal operating budgets. Hofmann and Bubb (2003) explained about the
virtual environment for the typical industrial application. Almgren and
Cahow (2005) explained the way to make engineering education more
innovative by using the Internet.
Pheatt and Ballester (2003) discussed the design and implementation of web
courses and remote experiments, and the incorporation of these techniques
into the curriculum without acquiring equipment, setting up equipment or
creating a laboratory environment. Che (2005) presented the development of
biological engineering education along with a discussion on the
development of e-learning with a time sharing mainframe model and
providing a centralized, remote controllable biology lab. Feisel and Rosa
(2005) explored the major factors influencing conventional laboratories.
They described the various limitations which affect the effectiveness of
laboratory work and also its importance in science and engineering
education. Ozeki et al. (2006) produced remote experiments with a digital
certificate and encryption of communication data to protect a supervisory
control system against illegal access. They did the development of the
remote experiment system with testing and demonstration. Jeschke et al.
(2008) discussed the integration of experimental setups into a virtual
cooperative knowledge space, so that availability and accessibility can be
enhanced for a wide range of users. They worked out the architecture and
implementation of the remote experiment. Grber, Vetter, Eckert & Jodl
(2007) described a remote laboratory and its helpfulness to provide a tool to
sustain this shift towards a student-centered engineering teaching approach.
In papers by Bauer and Fedk (2007) and Bauer et al. (2008) a distance
measurement application for educational purposes was described, and the
monitoring of industrial applications was studied through web based
applications. Macas and Mndez (2008) describes the automation of their
ISSN: 1694-2108 | Vol. 2, No. 1. JUNE 2013 4

IJCSBI.ORG
laboratory and rapid integration of automation systems into most of the
engineering processes. Cui, Wang, Zhang and Akujuobi, 2011, 27 (4)
(2008) focused on the reasons for traditional laboratories lagging behind and
emphasize the basic demands behind the modern trend to increased
virtualisation. Kolias et al. (2007) provided a categorization of remote
experimentation in education according to the platform used for conducting
the experiments and the scientific field represented. Gadzhanov and
Nafalski (2010) reviewed the pedagogical effectiveness of distance
education, with a special focus on remote laboratories for measurement and
control. Machotka et al. (2010) developed the remote laboratory NetLab at
the University of South Australia. NetLab was developed from the
beginning as a collaborative learning environment that enables students to
cooperate while conducting remote experiments via the Internet on both the
domestic and international levels. Herrera and Fuller (2011) proposed a
model for the implementation of remote experimentation laboratories in a
distributed collaborative scenario, focusing on two crucial key elements,
namely the shared knowledge and the interaction for the collaboration. They
contributed towards the implementation of remote experimentation using
collaborative scenarios. R Tiwari and K Singh summarized their experiences
during the design, evaluation and implementation of Virtualization for a
number of engineering experiments, enabling these to be accessed through
the Internet as virtual experiments. They also covered the features required
to give the 'feel' of performing experiments, inclusion of learning
components, incorporation of easy navigation, testing of the effectiveness of
learning, and development of useful feedback mechanisms.
From the geographical point of view, the educational environment can be

divided into two categories called classroom environment and lab
environment. The most common modern classroom environment is the
multimedia classroom which includes: projectors and large screens, object
display platform, DVD player and computers. Among them, the computer
is increasingly becoming the center of multimedia classrooms. The virtual
classroom is also called live classroom, which can provide real-time
teaching function through web technology. Teachers can use text, images,
handwritings, voices, videos and other forms of media to teach students with
theoretical knowledge. Students can use text, voices and other forms of
media to ask questions at any time. The learning process is similar to the
classroom environment.
3. VIRTUALIZATION OF LABORATORIES
Virtualization is the creation of a virtual rather than actual version of
something, such as a hardware platform, an operating system, a storage
ISSN: 1694-2108 | Vol. 2, No. 1. JUNE 2013 5

IJCSBI.ORG
device or a network resource. Virtualization is a technique that allows you
to run more than one server (or other infrastructure component) on the same
hardware [7]. The virtual machine concept was in existence since 1960s
when it was first developed by IBM to provide concurrent, interactive
access to a mainframe computer. Each virtual machine (VM) used to be an
instance of the physical machine that gave users an illusion of accessing the
physical machine directly. It was an elegant and transparent way to enable
time-sharing and resource-sharing on the highly expensive hardware. Each
VM was a fully protected and isolated copy of the underlying system. Users
could execute, develop, and test applications without ever having to fear
causing a crash to systems used by other users on the same computer.
Virtualization was thus used to reduce the hardware acquisition cost and
improving the productivity by letting number of users work on it
simultaneously. As hardware got cheaper and multiprocessing operating
systems emerged, VMs were almost extinct in the 1970s and 1980s. With
the emergence of wide varieties of PC based hardware and operating
systems in 1990s, the Virtualization ideas were in demand again [8]. There
can be innumerous reasons how Virtualization can be useful in practical
scenarios, a few of which are the following:
Operating System (OS) Level Virtualization: An OS-level VM is a

virtual execution environment that can be forked instantly from the base
operating environment. OS-level Virtualization has been widely used to
improve security, manageability and availability of todays complex
software environment, with small runtime and resource overhead, and
with minimal changes to the existing computing infrastructure.
Hardware Level Virtualization: Hardware-level Virtualization
technologies simulate the complete or a subset of hardware to run
unmodied guest OS, or modify the guest OS to utilize the underlying
physical hardware. Examples of the hardware-level virtualization
technologies are Bochs, VMware, Virtual PC, Xen.
Application Virtualization: Application Virtualization creates isolated
virtual environments on a host OS to sandbox the deployment and
execution of selected application programs. Such virtual environments
are mainly used for four purposes: resolving application conflicts,
migrating running processes, isolating untrusted programs, and
supporting portable application environments.
Cloud Computing: Cloud Computing is the delivery of computing as a
service rather than a product, whereby shared resources, software, and
information are provided with computers and other devices as a metered
service over a network (typically the Internet). Cloud Computing may
look like Virtualization because it appears that your application is
running on a virtual server detached from any reliance or connection to a
ISSN: 1694-2108 | Vol. 2, No. 1. JUNE 2013 6

IJCSBI.ORG
single physical host. However, Cloud Computing can be better described
as a service where Virtualization is part of a physical infrastructure.
Cloud Computing builds on top of a virtualized infrastructure (compute,
storage, network) by using standardization and automated delivery to
provide service management. This makes monitoring the virtualized
resources and the responsible deployment of these resources possible.
3.1 Need of Virtual Lab
Physical distances and the lack of resources make us unable to perform
experiments, especially when they involve sophisticated instruments. Also,
good teachers are always a scarce resource. Web-based and video-based
courses address the issue of teaching to some extent. Conducting joint
experiments by two participating institutions and also sharing costly
resources has always been a challenge. With the present day internet and
computer technologies the above limitations can no more hamper students
and researchers in enhancing their skills and knowledge. In India costly
instruments and equipment need to be shared with fellow researchers to the
extent possible. Web enabled experiments can be designed for remote
operation and viewing so as to enthuse the curiosity and innovation into
students. This would help in learning basic and advanced concepts through
remote experimentation. Today most equipment has a computer interface
for control and data storage. It is possible to design good experiments
around some of this equipment which would enhance the learning of a
student. Internet-based experimentation further permits use of resources
knowledge, software, and data available on the web, apart from encouraging
skillful experiments being simultaneously performed at points separated in
space.
3.2 Objectives of the Virtual Labs
To provide remote-access to Labs in various disciplines of Science and
Engineering. These Virtual Labs would cater to students at the
undergraduate level, postgraduate level as well as to research scholars.
To enthuse students to conduct experiments by arousing their curiosity.
This would help them in learning basic and advanced concepts through
remote experimentation.
To provide a complete Learning Management System around the Virtual
Labs, where the students can avail the various tools for learning,
including additional web-resources, video-lectures, animated
demonstrations and self evaluation.
To share costly equipment and resources, which are otherwise available
to a limited number of users due to constraints on time and geographical
distances?
ISSN: 1694-2108 | Vol. 2, No. 1. JUNE 2013 7

IJCSBI.ORG
3.3 Salient Features of the Virtual Lab
Virtual Labs will provide to the students the result of an experiment by
one of the following methods (or possibly a combination)
Modeling the physical phenomenon by a set of equations and carrying
out simulations to yield the result of the particular experiment. This can,
at-the-best, provide an approximate version of the real-world
experiment.
Providing measured data for virtual lab experiments corresponding to
the data previously obtained by measurements on an actual system.
Remotely triggering an experiment in an actual lab and providing the
student the result of the experiment through the computer interface. This
would entail carrying out the actual lab experiment remotely.
Virtual Labs will be made more effective and realistic by providing
additional inputs to the students like to accompany audio and video
streaming of an actual lab experiment and equipment.
For the touch and feel part, the students can possibly visit an actual
laboratory for a short duration.
Laboratories play an important role in engineering education. The laboratory

work provides an opportunity to witness classroom-learned theoretical
concepts at work; operate instruments used in the experimental set-up;
measure different quantities and thereafter analyze the experimental data;
and work collaboratively. However, physical distances and the lack of
resources make us unable to perform experiments, especially when they
involve sophisticated instruments. Also, trained teachers are always a scarce
resource. Rapid development and ever increasing use of Information and
Communication Technology (ICT) has revolutionized the education system
by eliminating limitations in terms of physical distance, time and access, by
providing equal opportunity to everyone, irrespective of place and time.
Virtual laboratories have been developed in different areas, to reproduce

experiments that were made in physical laboratories. Virtual labs are useful
for pre-practice and post-analysis of experiments developed in physics labs,
and in some cases they can replace th
e physical lab itself. Although virtual labs may have many limitations, they
have many advantages over physical labs. For example, some physical labs
have a scarcity of resources (in equipment and staff), limiting the
researcher's performance. Virtual labs have relatively low costs,
experiments can easily be repeated, and there are no inconveniences in
failing experiments, because the virtual environment is controlled, and there
are no risks for natural systems. It is desirable that virtual labs exploit the
advantages of virtual reality, multimedia, and the Internet. Virtual labs have
ISSN: 1694-2108 | Vol. 2, No. 1. JUNE 2013 8

IJCSBI.ORG
been developed for different areas, such as physics, electronics, robotics,
physiology, chemistry, engineering, economics, and ecology[9].
Develop the Soft Lab for the following purposes.
Lower Your Measurement Costs with Virtual Instrumentation the costs
of a measurement application.
Reducing system specification time cost.
Lowering hardware and software prices.
Minimizing set-up and configuration time costs.
Decreasing application software development time costs.
Lowering system validation & hardware calibration time costs.
4. EXISTING VIRTUAL LABS

An electronics student always looking for computer simulation softwares to
do their experiments and hobby circuits with PC thereby reducing the cost
for researches for projects and laboratory works. The electronic simulation
software provides a platform for design, wiring the circuits virtually and we
can connect it to a number of equipments which is programmed by the
software vendor with special parameters for each. So here we provide a top
list of essential software that becomes handy during the course time and so
on. Some of them have download links provided.
4.1 IIT Virtual Labs

An Initiative of Ministry of Human Resource Development (MHRD) Under
the National Mission on Education through ICT design in IIT Delhi,
Bombay, Kanpur, Kharagpur, Madras, Guwahati and all its participating
Institutes. IIT design the virtual lab for Electronics and communication,
Computer Science and Engineering, Electronic Engineering, Mechanical
Engineering. This system is more advanced for the Engineering faculty [10].
4.2 LabVIEW
LabVIEW (short for Laboratory Virtual Instrumentation Engineering
Workbench) is a system design platform and development environment for
a visual programming language from National Instruments. LabVIEW is a
software development environment originally designed for computer-
controlled data acquisition. It has since grown to be a general purpose
programming environment with a wide range of applications. The basic
object in LabVIEW is a Virtual Instrument (VI). It has two parts:
The Front Panel or GUI.
The Block Diagram or programming guts.
The VI Front Panel was originally designed to mimic the front panel of a
laboratory instrument, but it can be set up to resemble pretty much any sort
ISSN: 1694-2108 | Vol. 2, No. 1. JUNE 2013 9

IJCSBI.ORG
of user interface. National Instruments has a website entitled LabVIEW 101
that contains a series of instructional videos and PDFs of the instructions.
LabVIEW is commonly used for data acquisition, instrument control, and
industrial automation on a variety of platforms including Microsoft
Windows, various versions of UNIX, Linux. The latest version of LabVIEW
is version LabVIEW 2011, released in August 2011. In LabVIEW 8.20,
you have the freedom to choose your preferred syntax for technical
computing. Whether developing algorithms, analyzing results, processing
signals, or exploring DSP concepts, you can combine the intuitive
LabVIEW graphical dataflow programming with LabVIEW MathScript, a
math-oriented text-based programming language that includes more than
600 commonly used functions for math, signal processing and analysis [11].
4.3 Circuit Maker

Circuit maker is a simple, powerful schematic capture and downloading
tools. It is most useful for a normal user to design and simulate his various
circuits using circuit maker [12].
4.4 Digital Works

Digital works is a simple graphical design tool to construct various digital
electronic circuits and virtually simulates to analyze their behavior through
real time simulation [13].
4.5 MATLAB
MATLAB (matrix laboratory) is a numerical computing environment and
fourth generation programming language. Developed by MathWorks,
MATLAB allows matrix manipulations, plotting of function and data,
implementation of algorithms creation of user interfaces, and interfacing
with programs written in other languages, including C, C++, Java, and
FORTRAN. Although MATLAB is intended primarily for numerical
computing, an optional toolbox uses the MuPAD symbolic engine, allowing
access to symbolic computing capabilities. An additional package,
Simulink, adds graphical multi-domain simulation and Model-Based Design
for dynamic and embedded systems [14].
4.6 The Laboratory Web Site
A lab website is developed using Google Sites for local as well as remote
students. General lab safety and radiation safety information are made
available. Instructions to use Skype and the web cameras are also on the lab
website. In addition, the website includes separate links for all of the
experiments in NPRE 451 and provides references and links of relevant
information about each experiment. The course outline and the handout for
each experiment are also available. Video recording of the lecture for each
ISSN: 1694-2108 | Vol. 2, No. 1. JUNE 2013 10

IJCSBI.ORG
experiment can also be made available. A frequently asked question (FAQs)
section is continuously being developed for each experiment.
4.7 Virtual Labs

The laboratory consists of eight experiments ranging from introducing the
student how to use the equipment to build a resonant circuit and active
filters. The computer-based laboratory was created to duplicate an actual
electronics laboratory that has been given for many years at Vanderbilt
University as an initial practical experience with electrical circuits. The
computer environment created consists of a high fidelity simulation of the
physical environment embedded in a computer-structured copy of the
original, physical laboratory curriculum. After an introductory tour of the
laboratory procedure, the student can elect to work the first laboratory or to
go through a step-by-step tutorial. Tutorials are implemented for the
components (i.e., resistor, capacitor, and inductor) and the instruments
(breadboard, function generator, and oscilloscope). The component tutorials
provide very basic information on how to select component values. In the
case of resistors, after completing the tutorial, the student is asked to
identify the specific value of a resistor from its color code. For the function
generator and oscilloscope, the tutorial provides the name, location, and
function of each control and indicator for both instruments [15].
5. VIRTUAL INTELLIGENT SOFTLAB

Electronic devices and the lack of resources make us unable to perform
experiments, especially when they involve sophisticated instruments. Web-
based and video-based courses address the issue of teaching to some extent.
Conducting joint experiments by two participating institutions and also
sharing costly resources has always been a challenge. With the present day
internet and computer technologies the above limitations can no more
hamper students and researchers in enhancing their skills and knowledge.
The SoftLab philosophy challenges us to link the physical laboratory

experiment with its theoretical simulation model within a unified and
interactive environment. The goal for each instance of a SoftLab laboratory
is to create a software environment where experimental research, simulation
and education coexist and interact with each other. As part of the SoftLab
project, we are investigating the issues involved in the design and
development of the SoftLab Model for Electronics, Computer science and
engineering. The VIS describes how the SoftLab philosophy has used to
design and implements.
The Virtual SoftLab forces us to address the challenge of solving
experiments. Such systems require a wide range of expertise plus a flexible
ISSN: 1694-2108 | Vol. 2, No. 1. JUNE 2013 11

IJCSBI.ORG
and diverse array of equipment. The SoftLab framework should provide the
infrastructure and facilities that serve the needs of basic research.
The basic facilities include with VIS:
High-performance graphics processors to support scientific
visualization, and multimedia facilities.
Software that allows application researchers to make full use of these
facilities to evaluate their experiments without any risk.
Virtual laboratories that provide scientists with a full range of
functionality, both for experimental activities and for the mathematical
modeling and simulation of the experimental processes.
The laboratory presents an ideal environment in which to accept and meet

the challenges of computational science and engineering. Issues that must be
addressed include mathematical software, electronic prototyping, geometric
modeling, parallel algorithms, databases, software engineering, and
computer systems. This facility is the use of research in electronics,
computer science and engineering Department.
In order to support the various activities that take place in a physical
laboratory, scientists must be able to use the virtual laboratory -
To perform physical experiments by interacting with and controlling
remote instruments.
To simulate physical experiments through computer modeling of the
experimental process.
To compare physical and simulated data by playing back recorded
experiments.
To train scientists to understand the physical experiment, the
computational model of the experiment, and the relationships between
them.
Virtual SoftLab can benefit thousands of students and professors to learn
difficult experiments. The Virtual laboratory will bridge physical distance
and availability of resources. Today it is possible to design good
experiments among students for better learning. Virtual Labs would helpful
to students at the graduate level, postgraduate level as well as to research
scholars. This would help them in learning basic and advanced concepts
through Virtual Intelligent SoftLab (VIS) experimentation. This system
provides a complete Learning Management system around the virtual Labs
where the students can learn, observe and research the various tools. Video-
lectures, animated demonstrations and self evaluation program available in
future. This system also shares costly equipment and resources with visual
effects. This relatively new concept of Virtualization is cost effective and
there is no time bound for the users as experiments could be accessed "24 x
7". Virtual experiments can be used extensively for teaching, clearing, and
ISSN: 1694-2108 | Vol. 2, No. 1. JUNE 2013 12

IJCSBI.ORG
other computer-based education.The functionality provided by this problem
solving environment will therefore be:
Collect the experimental data in real time.
Provide a visualization of the experiments.
Control the experiment.
Input the experimental data to the computational model.
Control the computational model and visualize the numerical simulation
solution.
Professionals confirmed certain characteristics of the virtual lab.
Creating new intellectual models in education better than the real, and
more beautiful than the imagination.
Knowledge-building and inculcate information.
Encouraging and guiding students.
Registering students' information and evaluating them automatically.
Performing experiments, which are difficult to be performed in the
traditional lab due to its danger and high cost.
Reducing the learning time spent in the traditional lab.
Develop an exploration based on scientific assumptions and processes.
Permanently updated.
6. CONCLUSIONS
SoftLab will help Electronics; Computer Science and Engineering students
perform and practice experiments to improve their understanding of the
subject. The design of the VIS model is more effective and realistic as
necessary variable inputs and outputs are visible on the monitor screen.
This model created for the client based system, can be converted into a
client - server based application system. Through this model, the examiner
can conduct the examination, in which the student can perform their
experiments in a given time period, by applying timer. Some modifications
may be done, by applying a video camera. Through this video camera, the
lab instructor or examiner can watch the video clip of the experiment
performed by the student in stepwise so that the instructor/examiner can find
the errors if any. This virtual experiment provides practice for students for
the touch & feel part they have already performed in the laboratory.
7. ACKNOWLEDGEMENTS
We are very much thankful to Dr. D. A. Deshpande, Dr. P. K. Butey and Dr.
U. A. Lanjewar for their valuable inputs, constant guidance and their
extensive support and encouragement in this work.
ISSN: 1694-2108 | Vol. 2, No. 1. JUNE 2013 13

IJCSBI.ORG
8. REFERENCES
[1] Tiwari, R. & Singh, K (2011). Virtualization of engineering discipline experiments

for an Internet-based remote laboratory. Australasian Journal of Educational
Technology, 27(4), 671-692.
[2] The Physical Science Resource Center(PSRC) http://www.psrc-online.org/, Dec 2012.
[3] Remoter Dynamical System Laboratory, STEVENS, Institute of Technology
http://www.stevens.edu/remotelabs/, Dec 2012.
[4] Mercer University Online Interactive Chaotic Pendulum,
http://physics.mercer.edu/pendulum, Retrieved on Dec 2012.
[5] http://www.cage.curtin.edu.au/mechanical/info/vibrations, Retrieved on Dec 2012.
[6] http://www.lci.kent.edu/ALCOM/alcom.html, Retrieved on Dec 2012.
[7] Digit Fast Track to Virtualization, Volume 07, issue 04, April 2012.
[8] Susanta Nanda and Tzi-cker Chiueh, A Survey on Virtualization Technologies
Department of Computer Science SUNY, Stony Brook Stony Brook, NY 11794-4400.
[9] A Virtual Laboratory, March 2011; http://www.virtlab.com/main.aspx.
[10] Virtual Labs: Dec 2012, http://www.vlab.co.in/.
[11] Peter A. Blume, The LabVIEW Style Book, February 27, 2007, Prentice Hall. Part
of the National Instruments Virtual Instrumentation Series series. ISBN 0-13-145835-
3
[12] Circuit maker, Feb 2013, https://www.circuitlab.com/
[13] Digital Works, Feb 2013,
http://www.shsu.edu/~csc_tjm/fall2003/cs272/dworks_howto.html
[14] MATLAB Central File Exchange Library of over 12,000 MATLAB files and
toolboxes, mostly distributed under the BSD License.
[15] The Virtual Laboratories, Feb 2013 www.virtlab.com/
[16] Agelidis, V. G. (2005). A laboratory-supported power electronics and related
technologies undergraduate curriculum for aerospace engineering students.
International Journal of Engineering Education, 21(6), 1177-1188.
[17] ALCOM Education Project (2002). Online experiments.
http://olbers.kent.edu/alcomed.
[18] Almgren, R. C. & Cahow, J. A. (2005). Evolving technologies and trends for
innovative online delivery of the engineering curriculum. International Journal of
Online Engineering, 1(1).
[19] De Araujo, M., Antunes, J. & Piteau, P. (1998). Remote identification of impact
forces on loosely supported tubes: Part 1 - basic theory and experiments. Journal of
Sound and Vibration, 215(5), 1015-1041.
ISSN: 1694-2108 | Vol. 2, No. 1. JUNE 2013 14

IJCSBI.ORG
Negotiation Based Resource

Allocation for Distributed
Environments
Krishnamoorthy M, Senthil Murugan B
School of Information Technology and Engineering, VIT University,
Vellore, Tamil nadu, India
ABSTRACT
Resource allocation is an important task in distributed systems. A dynamic allocation is
most effective task in decreasing execution time and improving overall system
performance. The request for the resources might be generated at any time in environment
and it has to be allocated without any delays and conflicts. The workload distributed among
the system raises the major concern in the context of optimal allocation of resources. The
primary objective of this system is to achieve the high performance and availability.
Thereby reducing the computation and communication delays happens in dynamic
environment. It generates good results in synchronization between the systems.
Keywords
Resource Allocation, Negotiation, APR protocol, Load balancing, Airplane rerouting.
1. INTRODUCTION
Nowadays we are using computing platform for various purposes. The
purpose and usage may differ from one business to other business. Based on
their usage level, the appropriate computing hardware is chosen and the
users try to optimize its usage, thereby obtaining maximum profit from the
minimal hardware. Distributed system provides the environment to process
large amount of data with a short span of time. The total workload is
distributed among the systems, then it is processed within small amount of
time and its results could be aggregated. The job allocation can take place
based upon the resource availability. But if the system gets overloaded with
more number of jobs then it will create unnecessary delays and traffic,
hence there is also a delay in executing jobs submitted before. This results
in load imbalance problem and conflicts. So the load can be properly
distributed, and the jobs must be completed without any resource conflicts
and delays.
There exists a situation in which the resource required by a job, but it is
hold by some other process. But there is no usage of holding the resource.
So it must be reallocated as soon as when the process finishes its job. This
necessitates incorporates the purpose of bargaining in which the resources
must be reallocated when finishes its job, by that both process gets the
mutual benefit without any incurring any loss.
ISSN: 1694-2108 | Vol. 2, No. 1. JUNE 2013 1

IJCSBI.ORG
2. RELATED WORK
Load balancing has been previously studied mainly for the purpose of
balancing workload in centralized environment. So we remember some
balancing concept for distributed environment which is presented in some
papers. In contextual based resource allocation they viewed the system in
terms of physical and social contexts. Resource allocations is done based on
it maximal usage. In this system, if the task enters into the system, it
receives and puts into agent waiting queue. If the waiting queue reaches
maximum limit, it stops further allocations. The agents can execute the jobs
with required resources. If not, agents try to negotiate with other agents to
accomplish their desired tasks. The aim of this approach is to reduce
average queues size for the agents and maximizes the resource usage [1].
In Agent based simulations they have to proper monitoring and migrating
mechanism to correct load problems. Initially it detects heterogeneity of
system connected in environment and its loads. Balancing operations is
performed through various phases. Namely, Monitoring phase, filtering and
selecting operation phase, redistribution phase and migration phase. Better
approach to handle the dynamic changes, happen due to unanticipated
resources enters into the system. But we have more computation and
communication load to handle and need complex structure to monitoring
resources [2].
We have adaptive algorithm to effectively schedule and manage the shared
resources. And the failures can be handled very easily. If the main container
fails, it replicated main containers to become active. The problem of
centralized system fails or overload at centralized system then it totally
becomes imbalanced and recovery takes more time. This problem is solved
using distributed system through approximate optimized scheduling
algorithm with partial information. It is trying to reduce the delays
happening between the communication agents. This model can increase
distributed simulation performance by minimizing is communication loads
[3].
Task allocation is done in variety of ways. The multiagent system is
organized in network structure manner that can interact with its neighbour
nodes to increase computation speed and to decrease the communication
delays between agents. Through, Semi-Supervised approach with load
balancing is providing the good results compared to other methods. In full
supervised allocation method can reduce the task allocation based on the
probability of task allocation in future. The comparative analysis of both
locality and centrality can give alternative benefits with communication
costs and load performance [4].
ISSN: 1694-2108 | Vol. 2, No. 1. JUNE 2013 2

IJCSBI.ORG
Pipelined computation helps in giving the necessary information and it has
been passed from one system to other systems. The pipelined computation
is implemented in both centralized and distributed environment. The
distributed adaptive scheduling algorithm provides the better load balance
compared to centralized systems. The distributed, heterogeneous and shared
nature of large system gives more results to complete the work within
stipulated amount of time [5].
Load balancing is well handled in agent based distributed system. Here they
are tried to reduce the load from heavyweight to light weighted loads. Credit
based system is used to handle overload that may happens in the
environment. Through regression analysis method we can find out the
system information can find out parameters Location credit (LC) and Load
balancing coordinator (LBC), that decide when to migrate the load between
hosts The overall complexity of this approach is O (n2) [6].
In another approach they are presented the load balancing in game theory
point of view. Workload is distributed using round robin algorithm (it
allocates equal time slice to execute each task). The local interaction with
neighbours is the way to reduce the communication delays. The model
which followed here is to assigns the job each node with equal amount of
capacities to run. If the node finds overload it tries move the work to any
other node with load is very less by giving some payments. Dynamic load
balancing is achieved through local interactions itself [7].
3. PROPOSED WORK
3.1 Resource Allocation for Distributed Environment
Before getting detail into our proposed work, let see some of the problems
which might occurs in airspace management. In airspace management, we
have many numbers of airplanes who takes the journey in particular region.
But here, the unmanned little air planes need a protocol to negotiate the
route or in software applications where the multiagent system must suggest
the user possible alternative reorganizations of the route of many planes.
Regarding the airspace management, an accurate research has been made by
the Agent technology centre that developed AGENTFLY, a multi-agent
system enabling large-scale simulation of civilian and unmanned air traffic.
This system integrates advanced flight path planning, decentralized collision
avoidance with highly detailed models of the airplanes and the
environment. The AGENTFLY project is still maintained and its authors
are continuously improving it in many ways.
And also let us the problem of load balancing in distributed systems. The
distributed systems processes large amount of data by decomposing its work
into smaller segments and it gets completed. At each system we have certain
limit for the processing the number of tasks. So we need to distribute the
ISSN: 1694-2108 | Vol. 2, No. 1. JUNE 2013 3

IJCSBI.ORG
workload among the systems properly and maintain the system from load
imbalance state. Each system must be operated with equal amount of
workload. The allocation of jobs is based upon the resource requirement
level. The allocation must to provide more benefit when we do the
allocation process as well as the reallocation process. It helps in faster
execution of jobs submitted. And the resource allocation in distributed
environment approach gives more benefits in getting maximum resource
usage level.
So the implementation of airspace management is possible to do with kind
of distributed environment. The multiagent concept is an easiest approach to
make the negotiation between the airplane (Agents) systems. If we want to
say accurately, it provides more advantage over the communication and for
further operations.
3.2 NegotiationBasedResourceAllocationfor DistributedEnvironment

Negotiation is a concept in distributed multiagent system. The main purpose
of negotiation is to reach an agreement, and in particular, agreement in the
presence of conflicting goals and preferences. Negotiation usually proceeds
in a series of rounds, with some proposal made at every round. Proposal is
an agreement defined by the agent to avoid the conflicts. With the help of
reallocating the resources among one another, agents gets the mutually
benefit. Consider the negotiation with the context of resource allocation in
distributed environment. Allocation has to reduce the conflicts between the
systems and increase the throughput of the overall system. The resource
allocation is defined as tuple of:
<A, Z, v1...vn>
Where:
A is a set of agents
Z is a set of resources
vi: 2Z Ris a valuation function for each agent i belong to
A.
The allocation of Z to A is a partition Z1, Z2 Zn of the resources Z over
the agents A. Now, starting from some initial allocation P0= Z1, Z2 Zn
agents can bargain with each other in an attempt to improve the value of
their holding. If I have a resource that I do not value highly but you do, and
you have resource that you do not value highly but I do, then exchanging
resources, we can mutually benefit. Formally we considered for each agent i
as making a payment pi. In order to do the new allocation, i can be given
some payment, sufficient to compensate for the resulting loss in utility.
Consider the system is implemented with multiagent system. The algorithm
which is used for negotiation is as follows.
ISSN: 1694-2108 | Vol. 2, No. 1. JUNE 2013 4

IJCSBI.ORG
1. We start off with the initial allocation Z0; this is the initial allocation
of resources to agents.
2. We define the current allocation to be Z0 with 0 side payments.
3. Any agents is permitted to put forward a deal (Z, Z, p) as a proposal
where Z is the current allocation. If all agents to that deal (i.e. are
satisfied with the allocation Z and payments p) and the relevant
termination condition has been satisfied, then negotiation terminates
and the deal Z is implemented with payments p.
4. If every agent to the deal but the termination condition is not reached,
then the current allocation is set to Z with payments p, and go to
step3.
5. If some agent is unsatisfied with deal, then the current allocation
remains is unchanged and go to step3.
3.3 Comparsion Mechanism for Finding theCollision State
The system which is prevents from collision state is started. All the airplane
systems which is moving state, starts signalling to the base systems. The
base system is having some algorithm for finding the collision state. The
algorithm for finding the collision state is described below with pseudo
code. The comparator interface is always keep checks the condition
whenever the airplane is in signalling state (Airplane system is in moving
state).
Procedure Comparator_Interface (A: list of Agents)
Repeat
Flag = false
For it.next () to length (A) inclusive do:
For it.next () to length (A-1) inclusive do:
Count++
If A [i-1].location ==A [j].location then
If A [i-1].height==A [j].height then
If A [i-1].speed==A[j].speed then
Add (A [i-1] to List);
End if
ISSN: 1694-2108 | Vol. 2, No. 1. JUNE 2013 5

IJCSBI.ORG
End if
End if
End for
List ();
For k=0 to length (count)
It.previous
End for
Until End of list
End procedure
At this juncture, the loop is executed infinitely till the touches the condition,
there is no airplane is in Service Agents region. It predominately repeats this
process to avoid the collisions. The list A is continually getting updated and
traversed with each and every agent, to detect the airplane colliding state. If
the loop is traversed once, it activates the handler and manger to continue
further process.
4. SYSTEM ARCHITECTURE
In this section, we discuss the architecture of negotiation based resource
allocation system. The main reason behind this architecture is to reduce the
conflict that arises between the airplane systems.
At the initial stage, service agent (which is present in base systems) is
started with all the components which are required to manage all the
operations. It is considered as the basement system, which keep tracks the
airplane system. The base system is consists of three components namely,
Process Request, Service Handler, Manager. Each component works
together to handle the data which is generated from the airplane systems. It
decides whether the airplane what has to do when the uncertain event occurs
at air plane systems. And it is capable of handling any number of flight
systems which travels in its zone. So whenever the airplane starts it shares
the GPS information with the basement system. Service Agent stores the
airplane system location information in its database. At each and every
second, the airplane GPS information is updated in service agent database.
Every time the data about each flight is got synchronized to take more
optimized decisions in avoiding the collisions. The updating information in
database is happened through the process request component only.
ISSN: 1694-2108 | Vol. 2, No. 1. JUNE 2013 6

IJCSBI.ORG
Figure 1. Data flowing airplane system

The vital role of service agent is to check the conditions for the clashes and
to inform corresponding plane agents. This operation is performed by the
service handler component. And it continually checks the information
provided the process request component. If it finds the collision then it adds
to collision list, and it triggers the manager component starts its further
operations. The basement makes the negotiation process with the conflicting
airplane systems. The negotiation can take place with the context of
resource (i.e. route) and payments. The negotiation process starts in the
form of communication. Here the four types of messages considered such as
INFORM, REQUEST, and PROPOSE, ACCEPT_PROPOSAL and
REJECT_PROPOSAL. After finishing the negotiating process the airplane
get clear route that never conflicts with any other airplane systems.
5. EXPERIMENTS
In this section, we demonstrated this approach with real time application of
aircraft traffic control system. Here we have considered the airplane
rerouting progression. In aircraft traffic control system, each aeroplane is
considered as autonomous system (agents), which is travel in airspace with
generating signals. And it continually makes interaction with basement
ISSN: 1694-2108 | Vol. 2, No. 1. JUNE 2013 7

IJCSBI.ORG
system to avoid collisions. This system is implemented with distributed
multiagent system JADE (Java Agent Development Framework). JADE
provides the distributed agent platform. The agent platform can be split
among several hosts. Only one Java application, and therefore only one
JVM, is executed on each host.Agents are implemented as Java threads and
live within Agent Containersthat provide the runtime support to the agent
execution. The computer is present at each airplane system helps to load and
run the agents. Agents extract the system information (i.e. airplane
information) and then it properly sends its signal information to basement
systems.
The system which is present at middle named as Service Agent and also
considered as basement system, and its keep on receiving the signal from
airplane systems. It not only manages all airplane system information and
also it does the route allocating process. It keep tracks each airplane to
avoiding clashes, and to maintain respecting constraints on the path they can
use, time to make a decision, distance they need to maintain between them
and so on.
Here we considered airspace as resource which is to be used by the planes
to make the journey. All the airplane system (agents) information is got
synchronized with service agent in the form of messages. The reason behind
synchronizing is to detect the location of planes. At every time it updates
the information about airplane system in the servicing agent system.
It always checks the algorithmic conditions for collision occurrence. The
condition for finding collision state is comparing the distance and height
between the planes including with its speed. If it satisfies the collision state
condition, servicing agent informs to corresponding agents. For instance,
the distance difference between the airplane systems is less than 25miles is
considered for the collision and informed. Negotiations are fully handled by
Manager Component. Then corresponding agents receives the INFORM
message and immediately it requests the servicing agent for route clearance.
Here the purpose of negotiation comes between the agents. The servicing
agent responds with REQUEST message contains the proposals. The
proposal includes system (route) information to modify and side payment
information. The PROPOSE message may contains either
ACCEPTED/REJECTED conditions.
If the agent accepts the proposals then it has to make the payment and
process the reroute information for the clearance. After modifying it system
information it continues with its journey. The route information including
height and speed information it has to maintain till gets away from
conflicting airplanes. If the agent rejects the proposals, then servicing agent
responds with another PROPOSE message. It is the final proposal message
ISSN: 1694-2108 | Vol. 2, No. 1. JUNE 2013 8

IJCSBI.ORG
and it must be accepted by the airplane agent and to make the payments.
Because reject again and again it never the benefit of this maximal usage.
The benefit behind the both proposals is somewhat advantages to either
service agent or airplane system. But if agent accepts the proposal at first
gets more advantage and the agent rejects the proposal gets less benefit
only. Consider we have many number of airplanes may enters into the
condition of collision. It has been well handled by the process request and
service handler components. At this point the servicing agent is standing as
more reliable and available one to manage any number of airplane systems
data.
6. RESULTS
The system implementation is made with interconnection of aircraft systems
(i.e. Remote system) with service agent. At each system the airplane system
information is get loaded. After the start up, it sends current aeroplane
system location information to service agent and the corresponding plane
information are updated in service agent database. Basically the service
agent keeps tracks the aeroplanes for collision, so it must to have updated
airplane location information. Since the airplane is moving object, so we
cant predict the data exactly at all the intervals.
Figure 2. Load performance

At initially, the air systems are connected with various configurations. But
the underlying software system (Service Agent) is common for all the
systems. At figure 2, load performance is initially checked with
interconnection of 100 airplane systems. It generates good results without
any delays. At another interval, 200 systems are connected to find out the
ISSN: 1694-2108 | Vol. 2, No. 1. JUNE 2013 9

IJCSBI.ORG
variation. It gives what we already defined in case of 100 systems. The
results is slightly may vary due to load performance characteristics. Using
this approach get more benefits in low memory consumption, because if
airplane reaches its destination, related about the airplane system state is get
removed the service agent. So there is no pointless to maintain information
about the airplanes, after the end of the journey.
This experiment provides good results with certain number of systems. We
connect as many number of service agent which we need when service
reaches maximum capability. So here, we have considered one service agent
for each sector or region. So we concluded the airplane systems may
contact its nearest service agent to get the service. With this formation of
several service agents we dont get any unreachable conditions or not able
to connect situation. Gradually, we can get better performance compare to
single service agent.
For instance, at the figure 2, we put limits the number of airplane systems
for each service agent to connect. At first instant, we are connected 50 to
150 at one service agent. And if requests go beyond 200, it activates another
service agent for further requests processing of airplane system. Based on
the requirements, we can increase the service agents to get better
performance. But we dont have that much problems, because it maintains
the airplane systems information without redundancy.
7. CONCLUSION
The simulation experiment demonstrates the effectiveness of this approach.
Our on-going work includes distributed quality optimization for effectively
handling the conflicts and error handling mechanisms. As well as
information exchange mechanisms among the agents are handled well. We
can use this approach for load balancing between the servers and resource
allocation. So we have the possibility of sharing computer resources in
networking system. It helps to negotiate for obtaining the CPU, RAM,
PRINTER, and other computing resources. The main advantage behind this
distributed multiagent system approach is the better way to navigate the
workload if it reaches maximum limit. And it is easier to maintain the
location information in different system (i.e. in its corresponding region of
airplane systems), with replication of servicing agent in each sector or
region of traffic control system. And the distributed multiagent system
provides environment to handle any kind of failure that happens between
the systems.
It is more applicable in load balancing between the servers. Most probably it
increases the resource usage level at computing environments. The most
important thing with this application is airplane systems and basement
ISSN: 1694-2108 | Vol. 2, No. 1. JUNE 2013 10

IJCSBI.ORG
systems must have to proper devices to make communications. At any point
if we miss or lose signal information, we got severe problem.
8. REFERENCES
[1] Yichuan Jiang, Member, IEEE, and Jiuchuan Jiang, Contextual Resource
Negotiation-Based Task Allocation and Load Balancing in Complex Software
Systems, IEEE TRANSACTIONS ON PARALLEL AND DISTRIBUTED
SYSTEMS, VOL. 20, NO.5, MAY 2009.
[2] Robson E. De Grande, Azzedine Boukerche, Dynamic balancing of
communication and computation load for HLA-based simulations on large-scale
distributed systems, Journal of Parallel Distributed computing 71 (2011) 40-52.
[3] Qingqi Long, Jie Lin, Zhixun Sun, Agent scheduling model for adaptive dynamic
load balancing in agent-based distributed simulations, Simulation Modelling
Practice and Theory 19 (2011) 1021-1034.
[4] Yichuan Jiang AB, Zhaofeng Li C, Locality-sensitive task allocation and load
balancing in networked multiagent systems: Talent versus centrality, Journalof
Parallel Distributed computing 71 (2011) 822-836.
[5] IoannisRiakiotakis A, Florina M. Ciorba B, Theodore Andronikos C, George
Papakonstantinou A, Distributed dynamic load balancing for pipelined
computations on heterogeneous systems, Parallel Computing 37 (2011) 713-729.
[6] Maha A. Metaweia B, Salma A. Ghoneim A, Sahar M.Haggaga, Salwa M. Nassar
B, Load balancing in distributed multiagent computing system, Ain Shams
Engineering journal (2012) 3,237-249.
[7] Sameena Shah A, Ravi Kothari B, Convergence of the Dynamic Load balancing
problem to Nash equilibrium using distributed local Interactions, Information
Sciences 221 (2013) 297-305.
ISSN: 1694-2108 | Vol. 2, No. 1. JUNE 2013 11

IJCSBI.ORG
Wireless Solution for Water Saving In

Agriculture Using Embedded System
Venkata Narayana Eluri, K. Madhusudhana Rao, A. Srinag
Electronics and Communication Engineering,
KKR & KSR Institute of Technology and Sciences,
Guntur, Andhra Pradesh, India.
ABSTRACT
Agriculture is a source of livelihood of majority Indians and has great impact on the
economy of the country. In a country like India, where climatic conditions vary
substantially and irrigation facilities are poor. Agriculture is timely and sufficient supply of
water. Water pumps are crucial in agriculture where electricity is indeed. The frequent,
intermittent, low voltage supply of power to the agriculture sector has caused problems to
the farmers who are spending their time monitoring the supply of power without which no
progress of their work. The power supply with frequent power cuts have not only lowered
the efficiency of farmers but also have led to the frustration of the farmer to give up
agriculture and move to urban areas for better prospects in the globalized world. In this
paper we presented a system which shows, how mobile communication can benefit
millions of farmers in rural India by providing a solution for the irrigation problems caused
by intermittent electrical power supply. Information is exchanged in form of
messages/miscalls between the system and the user cell phones. The system is based on
ATMEGA32 micro-controller and includes protection against fluctuations in power supply.
For measuring time and temperature RTC DS1307 and DS18S20 are used. By using this
system the hardships of farmers relating water distribution can be relieved.
Keywords
AT command supporting GSM mobile phone, ATMEGA32 Micro controller, Relays,
Sensors, LCD for monitoring the current reading of all the parameters.
1. INTRODUCTION
Even in the modern era of industrialization, agriculture plays a very
significant role in the overall socio-economic development of India. India
has an agriculture based economy. 43% of Indias territory comes under
agricultural lands. Agriculture along with other related fields like forestry
and logging provides employment to 52% of Indias population. Agriculture
also accounts for 8.56% of the countrys total exports. According to a
survey made in 2007, agriculture accounts for 16.6% of Indias Gross
Domestic Product. Hence, it is agriculture that is the most influential field
as compared to others in India. This importance on agriculture leads to an
emphasis on better agricultural practices. The underground water level is
slowly falling down and forests are being cut which reduces the rainfall as
well. With increasing area available for cultivation and the need for
increasing the productivity from the farm land, there is a growing need for
ISSN: 1694-2108 | Vol. 2, No. 1. JUNE 2013 1

IJCSBI.ORG
electrical energy for irrigation. The generation of electricity is not growing
proportionately to the demand.
The supply to agriculture is limited to few fixed hours throughout the day.
In a sample survey of village Komarolu, located at distance of 17kms from
Giddalur (Taluka in Prakasam District, Coastal Region, Andhra Pradesh
State, India), it was observed that 3-phase supply is normally available for 7
hours a day from 2.00 am to 9.00 am for a week and from 11.00 am to 6.00
pm for next week and changes alternately. Due to increase in demand of
electricity the schedule given was not followed and sometimes power given
at night times and fluctuations in voltage level leads to motor damage[1]. So
farmers, after carrying strenuous physical activities in farm during day, have
to return back to their farms in night to carry water distribution using 3-
phase motor pumps. Agriculture receives power mostly during mid night
(off-peak) as this reduces the cost of electricity supply for the transmission
and Distribution Company. The farmers have to be on their guard all the
time due to the unpredictable nature of supply of electrical energy. And the
farmers have to switch on their motor after electricity supply resumes. The
reduced amount of yielding, man power wastage, and idle state of
equipment can observed in its results.
Due to the fast development in tele-communication technologies, it is
believed that wireless solution for irrigation in agriculture. This system has
fully utilized wireless sensor network, Global System for Mobile
Communication (GSM) and short message service (SMS)/missed call to
carry out data from the sensors to computers or directly alert the farmers
through their mobile phone and to control the remote watering process [2-
4], also through the mobile phone. This practice eliminates the use of wired
technology, improves old method of collecting data in farming areas and
allows farmers to control their sprinklers remotely. Also, It has been
observed that source of electrical energy generation is slowly depleting
using solar energy. We are going to discuss an example of how the mobile
technology can benefit millions of farmers by providing a solution for the
irrigation problems. Simple cell phones having just voice call and
messaging facility or cell phones with non-working display can be easily
adapted for remote control applications.
2. SYSTEM DESCRIPTION
A system is developed for optimum water distribution in the fields through
motor pump. The crop yield is maximized to a great extent by providing
proper amount of water at suitable time intervals based on climatic
conditions. The system offers attractive features like automatic control
based on parameters specified through keyboard /SMS/ number of miscalls;
provides protection against single phasing, over-current, over-voltage, dry
ISSN: 1694-2108 | Vol. 2, No. 1. JUNE 2013 2

IJCSBI.ORG
running and probable bearing faults; alerts users in case of abnormal
conditions like power failure, dry-running, etc. and provides audible
indication through buzzer/ miscall on completion of task.
The figure 1 shows the Block diagram of the scheme. Using keyboard the
parameters of the system can be set or received in form of SMS/number of
miscalls in specified duration from user mobile through serial cable
connected to control system cell phone (Model Nokia 6610). Based on
commands received microcontroller system sends signals to switch on / off
motor through Starter using relays controlled by its ports. Three phase 5
hp,7 hp,10.5 hp induction motor working on Direct-on-line Starter are
chosen for agriculture purpose.
3 PHASESUPPLY
R Y B
CELLULAR
USER LINK ATMEGA32 STARTER
SYSTEM
MOBILE MICROCONTROLLER (MOTOR
MOBILE
BASED SYSTEM CONTRACTER)
LCD RTCSENS KEY

MOTORPUMP
DISPLAY ORS BOARD
Figure 1. System Block Diagram

2.1 Cell phone Interface:
Cell phone 6610 is connected to AVR Microcontroller board through
RS232C serial interface. AT commands are sent by sending text strings A,
T, along with specified command strings through serial port to cell phone
and are executed on receipt of carriage return [5-6]. The result codes are
sent by cell phone to system (TE) to indicate the status after execution of
command.
1) SMS Approach: SMS is store and forward way of transmitting messages
to and from cell phones. The major advantage of using SMS is provision of
intimation to the sender when SMS is delivered at the destination and ability
of SMSC to continue efforts for delivery of message for the specified
validity period if network is presently busy or called user is outside the
coverage area. Using CMGS command the text message is sent to cell
phone. CNMI command is used to indicate to TE about the receipt of
incoming SMS message from the network. On receipt of the SMS message,
unsolicited result code +CMT is obtained from which text message is
ISSN: 1694-2108 | Vol. 2, No. 1. JUNE 2013 3

IJCSBI.ORG
extracted and checked with predetermined format, which includes
password, desired time or ON/OFF commands or status query.
Microcontroller carries out for the valid control message. In this application,
any incoming SMS message is directly routed to micro-controller (TE) and
any outgoing text message is directly sent by micro-controller to designated
cell phone number without being stored in control system cell phone
memory. As a result, phone memory is not inundated with messages in spite
of many messages being transferred.
2) Miscall Approach: The operational cost of communication between user
and control system cell phones is further reduced by using concept of
miscall where in no charges are incurred by using only ring signal for
information transfer. Miscalls are treated in two situations one is calling
party disconnects after receiving ring tones and second one when called
party does not respond to call within mentioned 5 minutes. The system cell
phone is designed to send specified number of miscall(s) within five
minutes duration to user cell phone to report various conditions as shown in
Table 1. Similarly, user cell phone sends commands to system cell phone by
making specified number of miscalls as shown in Table 2.
Table 1. Messages based on missed calls from system cellphone
No. of missed calls in 5 min. Message indication
01 No Power
02 Power failure
03 Resumption after normal conditions
04 Task completion
05 Probable motor faults/dry running
Table 2. Command based on missed calls from user cell phone

No. of missed calls in 5min. Message indication
01 Switch on
02 Switch off
03 Increase on time by 0ne hour
04 Decrease on time by one hour
05 Present status of the motor
This novel concept of miscalls results in substantial savings without

comprising the utility of system. Another advantage of miscall over SMS is
that during night time, ringing tone can easily wake-up farmer to carry out
necessary arrangement like shifting pipes to new locations, etc..CLIP
ISSN: 1694-2108 | Vol. 2, No. 1. JUNE 2013 4

IJCSBI.ORG
command is used in recognizing particular saved recipient. Micro-controller
checks incoming number with user cell phone when reception of unsolicited
code RING along with CLIP occurs. Micro-controller waits for five minutes
duration to check total number of calls and carries out specified task if
match founds. Similarly, microcontroller carries out number of voice calls
from control system cell phone to user cell phone within five minutes time
duration using ATD command. And SMS mode activated when the BUSY
or ERROR occurred while making voice call.
3. MICROCONTROLLER SYSTEM
AT Mega32 microcontroller has RISC architecture with 32 KB of in-system
programmable Flash, 1k E2PROM, 2k SRAM, 32-bit General purpose I/O,
8channel 10-bit ADC, TWI, USART, SPI, JTAG interface support etc [7-8].
Ponyprog software was used for flash programming [9]. The software was
developed in C language using GCC compiler.
3.1 Interfacing
Interfacing diagram of micro-controller system is shown in Fig. 2. 8-bits of
Port A are configured as analog inputs ports. PA0-PA2 bits monitor the
present values of supply phase voltages. Water level of well indicated by
port A 3 and 4 pins while PA5 to PA7 are used to sense whether water has
reached the desired final destination of the regions. PB0-PB5 pins are
connected to 2 16 characters LCD display in 4-bit data length mode. Two
relays are controlled by upper 2 bits of Port B. Start (green) pushbutton of
starter for automated starting of the pump from micro-controller board is
connected to one relay while other relay is connected in series with stop
(red) pushbutton for stopping the motor pump from micro-controller. Upper
4 bits of Port C and upper 4 bits of Port D are used to interface 4x4
keyboard matrixes. DS1307 (RTC) is chosen for implementation of timing
applications. It is connected through TWI interface (I2C) i.e. PC0 (SCL)
and PC1 (SDA) pins. Two temperature sensors (DS18S20) are used having
single wire interface for connectivity. PC2 bit is used for single-wire
interface. Internally RxD (PD0) and TxD (PD1) are connected to 9-pin
RS232 female connector through MAX 232 IC for TTL-
RS232Csignaltranslation.
3.2 Real Time Clock
The DS1307 serial real-time clock (RTC) is a low power, full binary-coded
decimal (BCD) clock/calendar [10]. Address and data are transferred
serially through an I2C, bi-directional bus (TWI). The clock/calendar
provides all timing information from seconds to years. With the help of
keyboard and LCD display, present date and time are written into
corresponding internal memory locations of this IC using I2C protocol.
ISSN: 1694-2108 | Vol. 2, No. 1. JUNE 2013 5

IJCSBI.ORG
Whenever on-time duration for pump is specified, the two registers of AT
Mega32 are used to store time duration in minutes and hours.
R PD7
VOLTAGE 4X4
Y ADC0 (PA0) PD4
SENSING
B BLOCK ADC2 (PA2) PC4 KEY BOARD
PC7
WATER ADC3 (PA3) TXD (PD1)

LL TXD(P MAX232
LEVEL
HL ADC4 (PA4) RXDD1)
(PD0)
SENSOR
RXD
+5V
(PD0)
WATER
R1
REACH ADC (PA5)
R2 R
SENSOR ADC (PA7) C
R3
3
ATMEGA32
DS RESET
PC2
18S20 RES
ET
PB7 REL1 MOTOR
DS 1307 SCL (PC0) PB6 REL2 STARTER
SDA (PC1)
XTAL1
XTA PB5 D4
C1
L1 PB4 D5
PB3 D6 LCD
PB2 D7 DISPLAY
C2 PB1 ENIN
XTAL2 PB0 RS
GND
XTA
L2
Figure 2. Microcontroller system interfacing
GND
The pump is switched on through relays using ports of micro-controller.
After passage of every minute (known through RTC) relevant registers of
microcontroller are updated if normal conditions exists and when their
values reach null, by using PB7 bit the pump is switched off. The
occurrence of abnormal conditions like unbalanced phase voltages, dry-
running, etc. causes pump to be switched OFF and the counting is
temporary stopped and is resumed on restoration of normal conditions. Thus
it is ensured that proper quantity of water is distributed by keeping pump
ON for specified time duration under normal conditions and user is
informed about restoration of normal conditions through miscalls.
3.3 Phase Voltage Measurement
In order to measure phase voltages, three transformers of equal ratings (6-0-
6) are used to step down voltage. These voltages are converted into
appropriate dc levels at analog inputs of micro-controller. Phase voltages
ISSN: 1694-2108 | Vol. 2, No. 1. JUNE 2013 6

IJCSBI.ORG
are read at regular intervals by micro-controller. The values of phase
voltages are compared with one another. If the result of any comparison
exceeds specified value, signal is sent to switch off the pump along with
error message indicating unbalanced phase voltage condition and conveying
this information to user cell phone through miscalls. Internal 10-bit ADC of
micro-controller is configured to work in left justified format and only most
significant 8-bit values are used for comparison.
3.4 Temperature measurement
There has been tremendous research in fault diagnosis and protection
schemes for induction motor [11]. The basic over-current protection for
motor is provided by bimetallic strip of starter. Maximum reliability ensured
by mounting temperature sensor on body of motor and another temperature
sensor is mounted at a suitable location to measure ambient temperature.
Whenever temperature difference between the two sensors exceeds
specified safety limit (250C), signal is sent to switch off pump along with
error indication to LCD display and conveying miscalls to user cell phone to
indicate probable fault occurrence. This arrangement ensures that
catastrophic event like burning of motor due to any fault like over-current,
bearing blockage and insulation failure are avoided. And preventive
maintenance is carried out at substantially lower cost. The chosen
temperature sensor is DS 18S20 manufactured by Dallas Semiconductor
(Maxim) [12-13].It has operating temperature range of 55oC to +125oC. A
major advantage associated with this sensor is the availability of output
directly in digital form obviating the need of analog to digital conversion.
Moreover, this sensor provides inherent error-detection capability through
CRC technique.
3.5 Water Level Sensor
In order to prevent dry running of motor and allow automatic restart of
motor when sufficient water level is regained, three wires are inserted into
the well. One wire (GND) is inserted at bottom of well while second wire is
adjusted just above foot-valve of the suction pipe of motor pump (LL) and
third one at suitable level above second wire (HL). Microcontroller switches
OFF the motor and sends miscalls indicating empty well (dry running)
conditions whenever water falls below LL level. Whenever water rises
beyond HL level and sends miscalls indicating resumption of task, the
micro-controller switches ON the motor again. The flexibility of system can
be increased by using water level sensor and prevents the damage of motor.
ISSN: 1694-2108 | Vol. 2, No. 1. JUNE 2013 7

IJCSBI.ORG
Figure 3. Implementation
3.6 Water Reach Sensor
There are two approaches for water distribution in this system. One method
is to specify the ON time duration of pump and second approach is to
specify the area of water distribution before stopping motor. First approach
is more suitable for sprinkle-based irrigation system while second approach
is chosen for ground-level water irrigation. For ground level water
irrigation, two wires are extended to extreme end of region where water is
to be reached. In this scheme, three such regions are presently supported
and one end of wire of these regions (R1, R2, R3) are connected to PA5,
PA6 and PA7 port bits of microcontroller using pull-up resistors of 100k
while other ends are connected to GND.
4. CONCLUSION AND RECOMMENDATIONS

By using this system optimum water distribution can be achieved. All the
electric defects are vanished and the maximum utilization of power supply
achievement is possible with the proposed system and during normal
conditions the motor will be restarted automatically. The system result in
uniform distribution of water at given time periods, minimizing man power
effort, water wastage reduction, minimization of occurrences of motor faults
and intimation to user about the completion of task through miscalls/SMS
system proves to be great boon to farmers whose pump sets are located far
away from their homes and intimation about any abnormal conditions.
With introduction of MMS message support, it is possible to capture images
from field using higher end cell phone and disease-pest control management
can be carried out by analysis of these images by agriculturists. Various
parameters such are warmness, moisture in the weather, etc., can be noted at
regular intervals on daily basis and time duration of motor, amount and type
of fertilizers and pesticides can be decided based on analysis of acquired
data. The technological assistance to farmers can tremendously boost the
ISSN: 1694-2108 | Vol. 2, No. 1. JUNE 2013 8

IJCSBI.ORG
productivity of food grains and bring prosperity to this hardworking
population ensured with the usage of proposed system.
5. REFERENCES
[1] Ilonen J. Kamarainen J. K, Lindh T, Ahola J, Kalviainen H andPartanen,
Diagnosis tool for motor condition monitoring. IEEE Trans. Ind. Application,
Vol 41, No. 4, 2005 pp. 963-971.
[2] KuniakiUmino, Yasuhiro Ohyama, Jin-Hua She and Hiroyuki Kobayashi,
Remote Controlled Embedded System, 4th China-Japan International
Workshop on Internet Technology and Control Applications, Hunan, China, 21-
26 October, 2005.
[3] A Weaver J Luo, and S Zhang, Monitoring and Control using Internet and
Java, Proc. 25th Annual Conference on IEEE Industrial Electronics IECON99,
Vol. 3 , pp 1152-1158, 1999.
[4] Trohandl C, Proske M &Elmeureich W, Remote Target Monitoring in
Embedded System Lab courses using a Sensor Network, Proc. 32nd Annual
Conference on IEEE Industrial Electronics IECON 2006, 6-10 Nov 2006, pp
5433-5438.
[5] AT Commands Set for Nokia GSM and WCDMA products,Version 1.2, July
2005, available online: http://forum.nokia.com.
[6] ETSI TS 127 007 V5.3.0 (2003-03) Digital cellular telecommunications system
(Phase2+); Universal Mobile Telecommunications System (UMTS); AT
command set for 3G User Equipment (UE) (3GPP TS 27.007 version 5.3.0
Release 5, available online: http://www.etsi.org.
[7] ATMELAVRATMega32 Datasheet, available online:
http://www.atmel.com/dyn/resources/prod_documents/2503S.pdf.
[8] Richard Barnett, Larry OCull and Sarah Cox, Embedded C Programming and
Atmel AVR, Delmar-Cengage Learning India Pvt. Ltd., 2007 Edition (India).
[9] Ponyprog software, available online: http://www.lancos.com/
[10] DS1307 64 x 8, Serial, I2C Real-Time Clock Datasheet, availableonline:
http://pdfserv.maxim-ic.com/en/ds/DS1307.
[11] Benbouzid, M.E.H, A review of induction motors signature analysis as a
medium for faults detection, IEEE Trans. Ind.Electronics, Vol 47, No.5, 2000
pp. 984-993.
[12] DS18S20 High-Precision 1-Wire Digital Thermometer, available online:
http://pdfserv.maxim-ic.com/en/ds/DS18S20.
[13] Overview of 1-Wire Technology and Its Use, available
online:http://www.maxim-ic.com/an1796.
ISSN: 1694-2108 | Vol. 2, No. 1. JUNE 2013 9

IJCSBI.ORG
Digital Image Tamper Detection

Techniques - A Comprehensive Study
Minati Mishra
Department of Information and Communication Technology
Fakir Mohan University, Balasore, Odisha, India
Flt. Lt. Dr. M. C. Adhikary

Department of Applied Physics and Ballistics
Fakir Mohan University, Balasore, Odisha, India
ABSTRACT
Photographs are considered to be the most powerful and trustworthy media of expression.
For a long time, those were accepted as proves of evidences in varied fields such as
journalism, forensic investigations, military intelligence, scientific research and
publications, crime detection and legal proceedings, investigation of insurance claims,
medical imaging etc. Today, digital images have completely replaced the conventional
photographs from every sphere of life but unfortunately, they seldom enjoy the credibility
of their conventional counterparts, thanks to the rapid advancements in the field of digital
image processing. The increasing availability of low cost and sometimes free of cost image
editing software such as Photoshop, Corel Paint Shop, Photoscape, PhotoPlus, GIMP and
Pixelmator have made the tampering of digital images even more easier and a common
practice. Now it has become quite impossible to say whether a photograph is a genuine
camera output or a manipulated version of it just by looking at it. As a result, photographs
have almost lost their reliability and place as proves of evidences in all fields. This is why
digital image tamper detection has emerged as an important research area to establish the
authenticity of digital photographs by separating the tampered lots from the original ones.
This paper gives a brief history of image tampering and a state-of-the-art review of the
tamper detection techniques.
Keywords
Digital Image, Tampering, Splicing, Cloning, Steganography, Watermarking .
1. INTRODUCTION
Today, digital images not only provide forged information but also work as
agents of secret communication. Users and editing specialists manipulate
digital images with varied goals. Scientists and researchers manipulate
images for their work to get published; medical images are tampered to
misrepresent the patients diagnostics, photo and yellow journalists use the
trick for creating and giving dramatic effect to their stories, politicians,
lawyers, forensic investigators use tampered images to direct the opinion of
people, court or law to their favour and so on. Hence, distinguishing the
original images from faked lots and establishing the authenticity of digital
ISSN: 1694-2108 | Vol. 2, No. 1. JUNE 2013 1

IJCSBI.ORG
photographs have become some of the greatest challenges of the present
time. Retouching, splicing, copy-pasting, cropping, cloning etc are some of
the popular techniques used for image manipulations. In additions to these
techniques there also exists a wide range of Steganographic methods those
use this popular digital media for secret data transmission.
This paper is organized as follows. A brief history of Image tampering is

given in the next section. Classification of tampering techniques and
distinction between image tampering and Steganography is made in Section
3. A state-of-the-art review of the existing tamper detection techniques is
given in section 4 which is followed by the summary and conclusion at the
end.
2. HISTORY OF IMAGE TAMPERING
According to Oxford dictionary, the literary meaning of tampering is
interfering with something so as to make unauthorized alterations or
damages to it [1]. Though it emerged as a critical problem with the
availability of digital cameras but Photo fakery is not an issue of 21st
century. Its history can be dated back to as early as in 1840s. Hippolyte
Bayard, the first person to create a fake image as recorded by history, is
famous for a picture of him committing suicide (Figure. 1). It was later
found that the photograph was forged out of frustration because he had lost
the chance of becoming 'the inventor of photography' to Louis Daguerre.
Daguerre patented a photography process earlier than him and owned all the
glory [2].
Figure 1. Hippolyte Bayards suicidal photograph
Before computers, photo manipulations were performed by retouching with

ink, paint, by double-exposing, airbrushing piecing photos or negatives
together in the darkroom, or scratching Polaroid. In the early days of
photography, the use of technology was not as advanced and efficient as it is
now. The outputs of such manipulations were very much similar to digital
manipulations but they were more difficult to create [3].
ISSN: 1694-2108 | Vol. 2, No. 1. JUNE 2013 2

IJCSBI.ORG
2.1 Conventional Vs Digital Image Tampering
Digital Image tampering is very much similar in nature to that of the
conventional photo fakeries where the conventional photographs are
replaced by their digital counterparts. One of the key characteristics of a
digital image is; it is easier to modify or manipulate a digital image in
comparison to its conventional counterpart. The process of digital image
tampering have been made even more simple and easier, thanks to the
availability of low cost (or sometimes free of cost) and powerful image
editing software nowadays.
History of digital image tampering can be dated back to the late twentieth
century to support political propagandas [3]. According to the Wall Street
Journal, even in 1989, out of all the colour photographs published in United
States, 10% were digitally altered or retouched! [4] An article in the journal
Nature reports on the impact of digital photography and image-manipulation
software in science. Mike Rossner, editor of the Journal of Cell Biology,
estimates that roughly 20% of accepted manuscripts to his journal contain at
least one figure that has to be remade because of inappropriate image
manipulation. And, in 1990, 2.5% of allegations examined by the U.S.
Office of Research Integrity, which monitors scientific misconduct,
involved contested scientific images. By 2001, this figure was nearly 26%
[5]. We suspect, as on date, more than 90% of commercially captured and
published digital photographs, if not tampered with some ulterior motive,
are at least retouched to improve the look, colour, contrast or background.
Few of the well-known historical tampered images include: The iconic

portraits of U.S President Abraham Lincoln (Figure. 2c) and Benito
Mussolini (Figure. 2d), the Iraq soldier picture of the Los Angeles Times
(March 1, 2004), the Internet image of a tourist just before the 911 WTC
attack (Sept 2001) (Figure. 2b) and that showing Jane Fonda and John Kerry
sharing the same speaker platform (Feb 2004) (Figure. 2a). Coming to
Indian context, the sensational photograph that sent the Congress MLA,
Kalpana Parulekar into police custody on 3rd February, 2012 was a tampered
photograph of State Lokayukta P P Naolekar. She has used that photo to
show that Naolekar was a former RSS worker (Figure. 2e). The former two
photographs given here represent the conventional photo fakeries whereas
the later are the results of sophisticated digital image tampering.
ISSN: 1694-2108 | Vol. 2, No. 1. JUNE 2013 3

IJCSBI.ORG
Figure 2b. This famous altered image

Figure 2a. This image was released smacked emails and went viral after the
during the 2004 presidential election 911 attacks in New York. The photo was
campaign which shows John Kerry and forwarded with the title the tourist guy,
Jane Fonda speaking together at an the accidental tourist, Waldo or the
anti-Vietnam war protest which later WTC Guy and displayed a person
turned out to be a politically motivated standing on the Observation deck of one
forged composite of two different images of the World Trade Centre towers
of Kerry taken on June 13, 1971 at an seconds before the plane hit the tower.
anti-war rally in Mineola, N.Y. and Later it was found to be an edited image
another of Jane Fonda taken in August, of Pter Guzli, a 25 year old Hungarian
1972 speaking at a Political rally at man taken during his visit to relatives in
Miami [6] [7]. N. Y. on 28/11/1997 [8].
Figure 2c. This iconic portrait of U.S Figure 2d. The horse handler had been
President Lincoln is a composite of airbrushed from the original
Lincolns head and the Southern photograph of Benito Mussolini to make
politician John Calhouns body [9]. him look more epic and heroic [9].
Figure 2e. Congress MLA Kalpana Parulekar showing a doctored photograph of

State Lokayukta P P Naolekar to show that he was a former RSS worker.[10]
3. YPES OF DIGITAL IMAGE TAMPERING

There are many more such cases of digital image tampering available and
the list is increasing every second with addition of newer cases. Based on
whether the manipulation is performed to the visible surface of the image or
ISSN: 1694-2108 | Vol. 2, No. 1. JUNE 2013 4

IJCSBI.ORG
to invisible planes, the manipulation techniques can be classified as image
tampering or image Steganography. Image tampering again can be
performed either by making changes to the context of the scene elements or
without the change of the context. In the second case, the recipient is duped
to believe that the objects in an image are something else from what they
really are but the image itself is not altered [11]. Figure.3 shows such an
image published in November 1997 after 58 tourists were killed in a
terrorist attack at the temple of Hatshepsut in Luxor Egypt in which a
puddle of water were digitally altered to appear as blood flowing from the
temple [12].
Figure 3. A digitally altered photograph (left) of temple of Hatshepsut (right)

after a terrorist attack.
The context based image tampering methods, according to the manipulation
process used, are further divided into three major classes such as:
Retouching, Splicing, and Copy-Move or Cloning. These manipulations
generally involve deletion or addition of scene elements of an image or
combination of scenes from multiple images [13]. Besides these above
mentioned techniques there is another class of photo fakery known as
Computer Graphics created photography. Though it requires a little more
effort still by the use sophisticated computer graphics rendering software
highly photorealistic images are created which are impossible to be
differentiated from the photographic images [14] [15] [16]. Figure.4 shows
two such photorealistic images.
Figure 4. Computer Graphics Created photorealistic image (left)

source: [12] and a photorealistic painting (right) source: [17]
ISSN: 1694-2108 | Vol. 2, No. 1. JUNE 2013 5

IJCSBI.ORG
3.1 Digital Image Tampering Vs Digital Image Steganography
Tampering and Steganography, though both the techniques manipulate a
digital image from its original capture but, they differ from each other at
their vary purposes. One manipulates an image for the purpose of hidden
communication whereas the other manipulates it to fake a fact and mislead
the viewer to misbelieve the truth behind a scene [18][19].
Image Steganography is the process of secret communication where a piece

of information (a secret message or an image, preferably encrypted) is
encoded into the bits of an innocent looking cover image, in such a manner
that the very existence of the secret information remains concealed without
raising any suspicion in the minds of the viewers [20] whereas, image
tampering generally replaces one or more parts of a host image with those
from the same or from some different images [21].
Because of its inherent purpose of data hiding, Steganography requires the

original and the Stego image to look alike but tampering on the other hand
aims at creating an image that appears to be an original camera output by
copy pasting parts from one or more images and preferably applying
geometric transformations such as scaling, rotation, cropping as well as
smoothing operations such as edge blurring, blending etc. to the copied parts
to ensure the tampered image to look as natural as possible and the
tampering undetectable to human vision.
Steganography is more global in nature and offer vary little or no change to

the image content in comparison to tampering which makes dramatic
changes to the image content those are more local in nature[21].
Figures.2c and 2d show two different tampered images and their original
counterparts and figure.5, given below, shows a Stego Lena image that has
been created by embedding a piece of hidden information into the original
Lena image. The differences between the pairs of tampered images and their
original ones are clearly visible where as there exists no visible difference
between both the Lena images.
Figure 5. Original Lena image, Stego Lena and the Retrieved hidden Information
ISSN: 1694-2108 | Vol. 2, No. 1. JUNE 2013 6

IJCSBI.ORG
4. TAMPER DETECTION TECHNIQUES
Digital image tamper detection techniques can be broadly classified into two
groups such as active detection techniques and passive (blind) techniques.
The active techniques require a pre-processing step and suggest embedding
of watermarks or digital signatures to images so as to authenticate them. The
major difficulty with this method is that it requires the watermark to be
embedded at the time of image capturing and for this; all digital cameras
should have a standard inbuilt watermark. Few questions need to be
answered in this regard are: whether all the camera manufacturing
companies will agree to manufacture cameras with some standard
watermarks signals inbuilt into them? Whether the costumers will be ready
to accept the probable degradation in the image quality due to the embedded
watermark? What about the processing time and complexity that involves
the embedding and retrieval of the watermark? Most importantly how to
deal with all those millions of pre manufactured digital cameras already
available in market as well as with users and can false watermarking be
completely ruled out? All these questions make the image authentication
and active tamper detection technique a remote possibility in practice. On
the other hand, the passive detection techniques do not require pre
embedding of any watermark or digital signatures to the images and hence
are commonly used for the purpose of tamper detection in digital images.
4.1 Active Methods of Tamper Detection

Active taper detection techniques due to their inherent limitation, though,
are not as common as those of the passive techniques still these are
considered to be most efficient image authentication methods and a lot of
research has been done in this field. These active image authentication
techniques are commonly classified into two categories: the first method
uses a fragile watermark, which localizes and detects the modifications to
the contents. While the rate of tamper detection is very high for these
methods they cannot distinguish between the simple brightness, contrast
adjustments and replacement or addition of scene elements. Increasing the
gray scales of all pixels by one would indicate a large extent of tampering
by this method, even though the image content remains unchanged for all
practical purposes [23]. The second method uses a semi-fragile
watermarking, that only detects the significant changes in the image while
permitting content-preserving processing.
The fragile watermark though has good localization and security properties
but cannot differentiate forgeries such as addition or removal of parts of
image, from the innocent image processing operations such as brightness or
contrast adjustments. J. Fridrich [23] solves this problem through his new
hybrid image authentication watermarking scheme that combines both the
ISSN: 1694-2108 | Vol. 2, No. 1. JUNE 2013 7

IJCSBI.ORG
fragile and a robust watermark. The hybrid watermark can be used to
accurately pinpoint changes as well as distinguish forgeries from other
innocent operations. This work is further improved and a secured hybrid
method [24] is presented in by Deguillaume and Voloshynovskiy. Several
researchers worked in these active tamper detection and authentication
schemes and developed a number of fragile, semi-fragile, robust, public as
well as private key based watermarks for copyright protection,
authentication and tamper detection [25-29] out of which, some either failed
to effectively address the problems or sacrifice tamper localization accuracy
of the original methods while few of them were proved to be highly efficient
and effective. However, the hierarchical digital watermarking method
proposed by Phen et.al is a simple but efficient method that not only
localizes and detects tampering but also is capable of tamper recovery with a
little degradation to the image quality. The precision of tamper detection
and localization of this method is 99.6% and 100% after level-2 and level-3
inspection, respectively. The tamper recovery rate is better than 93% for a
less than half tampered image [22].
4.2 Passive Detection Techniques

The passive methods are regarded as evolutionary developments in the area
of tamper detection. In contrast to the active authentication techniques these
methods neither require any prior information about the image nor
necessitate the pre embedding of any watermark or digital signature into the
image. The underlying assumption that is the basis of these schemes is,
though the carefully performed digital forgeries do not leave any visual clue
of alteration, they are bound to alter the statistical properties of the image.
The passive techniques try to detect digital tampering in the absence the
original photograph as well as without any pre inserted watermark just by
studying the statistical variations of the images [30]. Researchers of passive
detection techniques generally focus on two types of passive methods, the
copy-move forgery detection or cloning and splicing.
4.2.1 Cloning Detection

To clone or copy and paste a part of the image to conceal an object or
person is one of the most commonly used image manipulation techniques.
When it is done with care, it becomes almost impossible to detect the clone
visually and since the cloned region can be of any shape and size and can be
located anywhere in the image, it is not computationally possible to make an
exhaustive search of all sizes to all possible image locations.
According to [31], any Copy-Move forgery introduces a correlation between

the original image segment and the pasted one which can be used as a basis
for successful detection of this type of forgeries. Because the tampered
ISSN: 1694-2108 | Vol. 2, No. 1. JUNE 2013 8

IJCSBI.ORG
image will likely be compressed and because of a probable use of the
smoothing or other post processing operation, the segments may only
match approximately not exactly. The authors in this paper give two
different detection schemes: exact and robust matching those successfully
detects duplicate regions in an image even when the images are post
processed following a copy-paste. Methods based on blur movement
invariants and DWT, SVD, PCA based sorted neighbourhood approaches
are suggested in [32][33][34] for robust detection of cloned regions in an
image.
4.2.2 Splicing Detection Techniques

Digital splicing of two or more images into a single image is another
commonly used image manipulation technique. When performed carefully,
the borders between the spliced regions can be visually imperceptible. It is a
popular way to distort the semantic content of an image so as to fool the
viewer to misbelieve the truth behind a scene. Image splicing is a
fundamental operation in image forgery and is characterized by simple cut-
and-paste operation that takes a part of an image and puts it onto either the
same or another image without performing any post-processing smoothing
operation such as edge blurring, blending to it. By Image tampering, it
generally means splicing followed by the post-processing operations so as to
make the manipulation imperceptible to human vision [18].
Splicing detection is more challenging in comparison to cloning detection as

unlike cloned images spliced images do not have any duplicate regions and
unavailability of the source images offer no clue about the forgery. In [31],
however, the authors have shown that splicing disrupts higher-order Fourier
statistics, which can subsequently be used to detect splicing. Tian-Tsong Ng
and Shih-Fu Chang in [35] suggest a bio- coherence feature based splicing
model. Yun Q. Shi, Chunhua Chen, Wen Chen in [36] proposed an effective
splicing detection approach based on a natural image model that consists of
statistical features extracted from the given test image as well as 2-D arrays
generated by applying multi-size block DCT transform to the test images.
With the assumption that fusion of multiple statistical features can improve
the performance of splicing detection, Jing Zhang, Yun Zhao, Yuting Su in
their paper proposed a new splicing detection approach based on the
features utilized for steganalysis. They merge Markov process based
features and DCT features for splicing detection. The proposed approach
achieved up to 91.5% accuracy with a 109-dimensional feature vector [37].
In [38] the authors proposed an automatic detection framework to identify a
spliced image based on a human visual system (HVS) model in which visual
saliency and fixation are used to guide the feature extraction mechanism.
Zimba and Xingming in their paper propose a novel method for detecting
ISSN: 1694-2108 | Vol. 2, No. 1. JUNE 2013 9

IJCSBI.ORG
image splicing by thresholding transition region measures of DWT
coefficients of a suspicious image in chroma spaces. Only the low frequency
sub-band of the DWT of the suspected image is extracted to reduce the size
of the image and improve the performance [39]. Because splicing combines
image parts from multiple images so, careful study of the lighting conditions
can provide a better clue on detection of these types of manipulations.
5. SUMMARY AND CONCLUSIONS

A lot of research has been done on active as well as passive tamper
detection techniques and still a lot of work is going on worldwide to
successfully detect tampering in digital images. In this paper we have
reviewed the two popularly used passive detection techniques, splicing and
cloning. There exits many other techniques such as detection based on
examining the lighting environment, camera feature based detections,
studying the statistical and geometric properties. But interestingly the
tamper detection techniques and the tampering techniques are evolving
together. With the development of a new image authentication and tamper
detection technique, newer variants of detection resistant tampering methods
are evolving making the issue more challenging.
6. REFERENCES
[1] The Oxford dictionary online. http://oxforddictionaries.com
[2] P Lester. Photojournalism: An Ethical Approach. Lawrence Erlbaum Associates.
91-92; 1991.
[3] H. Farid, A Survey of Image Forgery Detection, IEEE Signal Processing
Magazine, 26(2):16-25, 2009
[4] C. Amsberry, Alterations of photos raise host of legal, ethical issues," TheWall
Street Journal, Jan 1989.
[5] http://www.fourandsix.com/photo-tampering-history
[6] http://en.wikipedia.org/wiki/Kerry_Fonda_2004_election_photo_controversy
[7] http://www.snopes.com/photos/politics/kerry2.asp
[8] http://www.famouspictures.org/mag/index.php?title=Altered_Images#Tourist_of_
Death
[9] http:// izismile.com/ 2012/02/08/ historic photographs_ which are known _to _be
_altered 13_pics.html
[10] http://samvada.org/2012/news-digest/congress-mla-arrested-for-tampering-rss-
chief-mohan-bhagwat-photo
[11] Kusam, Pawanesh Abrol and Devanand, Digital Tampering Detection
Techniques: A Review, BVICAMs International Journal of Information
Technology, Vol.1, No.2,
[12] Giuseppe Messina, "DCT Based Digital Forgery Identification", Presented to
Interpol Crime Against Children Group, Lyon, 25 March 2009.
[13] Sangwon Lee, David A. Shamma, Bruce Gooch, Detecting false captioning using
common-sense reasoning, Digital Investigation, Elsevier, 2006:S65-S70.
[14] F. Cutzu, R. Hammoud, and A. Leykin, Estimating the degree of photorealism of
images: Distinguishing paintings from photographs, in Proc. IEEE Conf.
Computer Vision Pattern Recognition, Madison, WI, 2003.
ISSN: 1694-2108 | Vol. 2, No. 1. JUNE 2013 10

IJCSBI.ORG
[15] Siwei Lyu and Hany Farid, "How Realistic is Photorealistic?, IEEE
TRANSACTIONS ON SIGNAL PROCESSING, VOL. 53, NO. 2, FEBRUARY
2005 845-850.
[16] V. Athitsos, M. J. Swain, and C. Frankel, Distinguishing photographs and
graphics on the world wide web, in Proc.Workshop Content-Based Access Image
Video Libraries, San Juan, PR, 1997.
[17] http://www.lostateminor.com/2013/04/23/nathan-walshs-photorealistic-paintings-
of-america/
[18] http://www.autodesk.com/eng/etc/fakeorfoto/about.html
[19] Patchara Sutthiwan, Yun-Qing Shi, Jing Dong, Tian-Tsong Ng, New
Developments in Color Image Tampering Detection, 978-1-4244-5309-2/10,
IEEE, Pp3064-3067, 2010.
[20] Minati Mishra, A.R. Routray, Sunit Kumar: High Security Image Steganography
with modified Arnolds cat map, IJCA, Vol.37, No.9:16-20, January 2012.
[21] Yun Q. Shi, Chunhua Chen, Guorong Xuan, and Wei Su, Steganalysis versus
Splicing Detection, International Workshop on Digital Watermarking, December
2007.
[22] Phen Lan Lina, Chung-Kai Hsiehb, Po-Whei Huangb, A hierarchical digital
watermarking method for image tampers detection and recovery, Pattern
Recognition, Elsevier, Pp. 2519-2528, 2005.
[23] Jiri Fridrich, A HYBRID WATERMARK FOR TAMPER DETECTION IN
DIGITAL IMAGES, Fifth International Symposium on Signal Processing and its
Applications (ISSPA) 1999.
[24] F. Deguillaume, S. Voloshynovskiy, T. Pun, Secure hybrid robust watermarking
resistant against tampering and copy attack, Signal Processing, Elsevier, Vol.
83,2003, pp. 21332170.
[25] C.-S. Woo, J. Du, B. Pham, "Multiple watermark method for privacy control and
tamper detection in medical images," in Proc. APRS Workshop on Digital Image
Computing, Australia, 2005, pp.59-64.
[26] Byun, S. C., et al. "A public-key based watermarking for color image
authentication." Multimedia and Expo, 2002. ICME'02. Proceedings. 2002 IEEE
International Conference on. Vol. 1. IEEE, 2002.
[27] P.W. Wong, A public key watermark for image verification and authentication,
in: Proceedings of the IEEE International Conference on Image Processing,
Chicago, IL, October 1998, pp. 455459.
[28] D. Kundur, D. Hatzinakos, "Toward a telltale watermarking technique for tamper-
proofing", in: Proceedings of the IEEE International Conference on Image
Processing, vol. 2, 1998, pp. 409413.
[29] J. Fridrich, Image watermarking for tamper detection, in: Proceedings of the IEEE
International Conference on Image Processing, vol. II, Chicago, IL, USA, October
1998, pp.404408.
[30] Hany Farid, Image Forgery Detection: A survey,IEEE SIGNAL PROCESSING
MAGAZINE , MARCH 2009, Pp: 16-25.
[31] Jessica Fridrich, David Soukal, and Jan Luk, Detection of Copy-Move Forgery
in Digital Images, Proc. Digital Forensic Research Workshop, IEEE Computer
Society, Cleveland, OH, August 2003
[32] G. Li, Q. Wu, D. Tu, and S. Sun, A sorted neighborhood approach for detecting
duplicated regions in image forgeries based on DWT and SVD, in IEEE Int.
Conf. Multimedia and Expo, Beijing, China, 2007, pp. 17501753.
[33] W. Luo, J. Huang, and G. Qiu, Robust detection of region-duplication forgery in
digital images, in Proc. Int. Conf. on Pattern Recognition, Washington,D.C.,
2006, pp. 746749.
ISSN: 1694-2108 | Vol. 2, No. 1. JUNE 2013 11

IJCSBI.ORG
[34] B. Mahdian and S. Saic, Detection of copy move forgery using a method based
on blur movement invariants, Forensic Sci. Int., vol. 171, pp. 180189, 2007.
[35] Tian-Tsong Ng, Shih-Fu Chang, A MODEL FOR IMAGE SPLICING,
International conference on Image Processing (ICIP), IEEE, Pp. 1169-1172, 2004.
[36] Yun Q. Shi, Chunhua Chen, Wen Chen, A Natural Image Model Approach to
Splicing Detection, ACM Conference07, Dallas, Texas, USA., September 20
21, 2007.
[37] Jing Zhang, Yun Zhao, Yuting Su, A New Approach Merging Markov and DCT
Features for Image Splicing Detection, IEEE, 390-394, 2009.
[38] Detect Digital Image Splicing with Visual Cues, Zhenhua Qu, Guoping Qiu, and
Jiwu Huang, S. Katzenbeisser and A.-R. Sadeghi (Eds.): IH 2009, LNCS 5806, pp.
247261, 2009.Springer-Verlag Berlin Heidelberg 2009.
[39] Michael Zimba, Sun Xingming, Digital Image Splicing Detection Based on Local
Complexity of Local Maximum Partial Gradient of DWT Coefficients,
International Journal of Digital Content Technology and its Applications (JDCTA)
Volume6, Number5, Pp. 1:9, March 2012.
ISSN: 1694-2108 | Vol. 2, No. 1. JUNE 2013 12

IJCSBI.ORG
A Multi-parametric based W-PAC

Mechanism in Ad Hoc Network
using IPv6 and IPv4 Address
S. Thirumurugan
Department of Computer Applications
J. J. College of Engineering and Technology
Trichy,Tamilnadu, India.
Dr. E. George Dharma Prakash Raj

Department of Computer Science and Engineering
Bharathidasan University
Trichy,Tamilnadu, India.
ABSTRACT
The wireless network plays an indispensable role in the present communication scenario.
This network functionality has been improved with the advent of clustering mechanism.
The type of nodes which are used in forming these clusters will decide the nature of the
applications the network could be utilized. These nodes can be either IPv4 or IPv6 nodes.
There are various queries need to be resolved while IPv4 or IPv6 nodes have been
considered to devise the clusters. This paper work brings out the discussion on cluster
formation using IPv4 nodes and IPv6 nodes. This study proposes IPv4 cluster creation,
IPv6 cluster creation and also a novel duplicate address detection strategy for stateless
address autoconfiguration. The cluster formation work has been simulated with the help of
OMNET++.
Keywords
W-PAC, IPV4, IPV6.
1. INTRODUCTION
The ad hoc wireless network formation has been based on IPv4 or IPv6
nodes. The IPv4 nodes are configured with the 32bit IP address and IPv6
nodes are configured with 128bit address. The configuration of node with
IPv4 address has almost got saturated in this present scenario. Due to this
non availability of unique IPv4 address the situation drives us to move
towards IPv6 address. The configuration of IPv4 node could be done
manually or dynamically using DHCP server. The IPv6 node can be
configured dynamically using DHCPv6 or stateless auto configuration
without DHCPv6. Generally the IPv6 node can self configure automatically.
The DHCPv6 has been similar to DHCPv4 used in IPv4 nodes
configuration. This study considers the clustering mechanism on ad hoc
scenario network to support scalability and improves the performance of
ISSN: 1694-2108 | Vol. 2, No. 1. JUNE 2013 1

IJCSBI.ORG
network in terms of bandwidth utilization. The IPv4 or IPv6 nodes are
considered to form the clusters. This work discusses the cluster formation
and issues need to be faced when IPv4 nodes or IPv6 nodes are used to
make clusters.
This paper has been organized as follows. Section 1 deals with introduction.
Section 2 says about the related works. Section 3 speaks about the existing
W-PAC algorithm to form clusters. Section 4 tells about the cluster creation
with the help IPv4 nodes. Section 5 gives out the cluster formation using
IPv6 nodes. Section 6 puts down the experimental results and analysis.
Section.7 signifies the future works on this study. Section.8 ends up with the
conclusion.
2. RELATED WORKS
The existing single parameter cluster formation algorithm k-means has
faced problem in obtaining an efficient cluster in terms of time. The PAC
[1] algorithm has solved that problem. But this algorithm forms clusters at
the initial level and leaves many nodes as non clustered node. It also doesnt
able to form maximum clusters that is possible for the given sample set of
nodes.
The Ex-PAC [2] algorithm, comes under single parametric algorithm, has
been an extension of the PAC mainly to overcome the drawbacks faced by
the PAC. In this algorithm the sample size also high for experimental
results. This also could be able to form the maximum number of clusters
possible for the given sample set of nodes.
The multi-parametric cluster formation algorithm W-PAC [3] forms the
cluster based on weight computation involves multiple parameters. This
method also suggests cluster maintenance strategy which deals with the load
distribution and mobility.
The cluster formation includes essential parameters on creating clusters
dynamically. The PSO-PAC [4] applies the swarm intelligence to device
clusters by considering the essential parameters. This ultimately reduces the
computation time towards optimizing the cluster creation process.
The cluster formation approach could be integrated with the existing well
performing protocols [5] has been realized. The applications [6] show
significance performance hike while the clustering has been incorporated
with the existing protocols at the network layer.
The IPv4 address no longer will serve the world has been realized. The IPv6
address has got the focus in the present scenario. The configuration of
address can be done in two ways. The configuration uses DHCPv6 server
and stateless autoconfiguration [7]. This address can be local link address or
global address. This will be decided based on the application requirement.
But there must be awareness on the restrictions of using the address range
has been essential.
ISSN: 1694-2108 | Vol. 2, No. 1. JUNE 2013 2

IJCSBI.ORG
The stateless autoconfiguration of IPv6 nodes should have duplicate address
detection mechanism. Since when the nodes move across the clusters there
is a possibility of address duplication. To eliminate this passive duplicate
address detection mechanism [8] has been introduced. This method shows
improved results than passive autoconfiguration for mobile ad hoc
networks.
The purpose of autoconfigured address has been realized while the ad hoc
network has come to existence. The limitations of IPv4 and the need of IPv6
[9][10] have been understood clearly. Having understood the difference
between IPv4 and IPv6 the scenarios will demand the specific way of
addressing the nodes. These works are dealing the autoconfiguration or
manual configuration of IPv4 and IPv6 nodes. The clustering as a
mechanism comes for this IPv4 or IPv6 autoconfigured node to make the
routing simple and also confirms the efficient utilization of bandwidth and
other resources.
3. W-PAC
The clustering mechanism W-PAC (weighted Partitioning Around Cluster
head) forms the cluster based on the weight value of the node. This has three
phases in clustering process. They are cluster creation, cluster head election
and maintenance phase. The cluster creation phase devises the clusters and
cluster head election phase selects the cluster head.
Cluster Creation
(1) Initialize set of nodes as M.
(2) Compute the degree of node Ni.
(3) Deg (Ni) = 0.
(5) If ( i not equal to j)
(4) j = 1.
begin
MD (Ni, Nj) = MOD X2 X1 + Y2 Y1
If ( Manhattan Dist( Ni , Nj) < Radius )

begin
Add ( Ni, Cm) // add to cluster
Deg(Ni) = Deg(Ni) + 1
j=j+1
end
else
Add ( Ni , NCn) // add to Non cluster
ISSN: 1694-2108 | Vol. 2, No. 1. JUNE 2013 3

IJCSBI.ORG
end
(6) Repeat the step 5 until j = M.
Cluster Head Election
(1) Create Clusters using W-PAC cluster creation.
(2) Cluster = Ci, P = Number of nodes in Ci.
(3) j = 1; Ni = (Ut , Vt ); Nj = (Ut1 , Vt1 );
(4) If ( i not equal to j)
begin
If ( Manhattan Dist(Ni,Nj) < Radious )
begin
Compute the Mobility speed of Node Ni belongs to Ci.
T
1
M(Ni) = MOD Ut Ut1 + Vt Vt1
T
t=1
Compute the Distance between Ni and Nj.
D(Ni) = MOD Ut Ut1 + Vt Vt1

t=1
j=j+1
end
end
(5) Repeat the step 4 until j = P.
(6) Assume the energy consumption of nodes E(Ni).
(7) The weight of node Ni computed as follows,
W(Ni) = q1*Deg(Ni) + q2*M(Ni) + q3*D(Ni) + q4*E(Ni)

(8) Repeat the step 7 for all nodes belong to Ci.
(9) CHk = Min { W(N1),W(N2),W(N3)W(NM) }.
(10) Repeat the step 2 through 9 for i = 1..no of clusters.
ISSN: 1694-2108 | Vol. 2, No. 1. JUNE 2013 4

IJCSBI.ORG
4. IPV4
The clusters with cluster head created using W-PAC procedure have been
considered for implementing the clusters using IPv4 nodes. By default all
the nodes are configured with IPv4 address.
192.168.25.2
N2
192.168.25.1 192.168.25.0
192.168.25.4
N4
N1 CH
192.168.25.3
N3
Figure 1. IPv4 Network Model

Algorithm: IPv4 nodes Configuration.
1) Input the Clusters Created using W-PAC procedure.
2) Assume CH(Cluster head) node as DHCPv4 server.
3) for each cluster Ci
Begin
Consider each node within Ci as DHCPv4 Client.
The Client sends Request message to DHCPv4 server.
The DHCP Server Configures the DHCPv4 Client
Node.
End
4) Repeat the step 3 for i = 1..... Number of clusters.
The Fig.1 shows the network model comprised of IPv4 nodes. The DHCPv4
server runs at Cluster head node. This will have the IPv4 address to be
assigned dynamically upon the clients request. This server periodically
sends the probing message to all the member nodes to know the status of
them. This clustering technique reduces the burden of handling the number
of clients by the cluster head and also eases the monitoring process of the
cluster head. One of the major benefit of the clustering mechanism would be
the frequencies which are used by the cluster can be reused by some other
cluster. The network model shows that the nodes are configured with IPv4
address by DHCPv4 server dynamically.
Figure 2 shows the communication happens between client and server as a
process of member node affiliation with the cluster head. The W-PAC
ISSN: 1694-2108 | Vol. 2, No. 1. JUNE 2013 5

IJCSBI.ORG
algorithm forms the clusters with cluster head. The cluster head node which
runs server process broadcasts the probe message to all the member nodes
belong to the same cluster. The member nodes put the request to cluster
head to configure themselves if the nodes have received the message first
time. The server periodically probes the client node to have the member
nodes under its affiliation.
Client DHCPv4
Request
Server
Reply (Configuration)
DHCPv4
Client Probe
Server
ACK / No ACK
Figure 2. Communication between Client and Server
The member node which runs the client process sends the acknowledgement
to the server process runs under cluster head node while the message
received count is more than once. This will update the members list table in
cluster head. The absence of the acknowledgement indicates the movement
of node beyond the cluster boundary. In this way the cluster head will be
aware of the member nodes under its control.
192.168.25.2 192.168.25.0
N2
192.168.25.1 192.168.25.0
192.168.25.4
N4
N1 CH
192.168.25.3
N3
Figure 3. Mobility of IPv4 Cluster

Head
Figure 3 shows the mobility scenario of cluster head. In this cluster head
mobility scenario the re-election of cluster head through broadcasting the
request message would be an overhead in the process of ad hoc network
maintenance. Thus, the result of W-PAC has already listed out the order of
the cluster heads. If the node at the second top position has adequate energy
ISSN: 1694-2108 | Vol. 2, No. 1. JUNE 2013 6

IJCSBI.ORG
to hold the position of cluster head, then this would be considered for the
cluster head position otherwise this will be continued till the node with the
sufficient energy level has been identified within the cluster.
192.168.25.2 192.168.25.2
N2
N2
192.168.25.1 192.168.25.0
192.168.25.4
N4
N1 CH
192.168.25.3
N3
Figure 4. Mobility of IPv4

The Figure 4 shows the mobility node
scenario of the member node. The node
movement across the cluster boundary will be known by the DHCPv4
server. Since the server monitors the nodes status by periodically sending
the probing packet. The server will reuse the IPv4 address of the node which
has moved out of the cluster. If the node enters to new cluster the probe
message will reach the member node from the cluster head of that cluster.
The node may send a request to the cluster head while the probe message
reaches first time to the member node.
5. IPV6
The clusters created using W-PAC procedure has been considered as input
for implementing the clusters using IPv6 nodes. The Figure 5 shows the
network model comprised of IPv6 nodes. These IPv6 nodes may be
configured in two ways. They can be either assigned with IPv6 address
using DHCPv6 server in a stateful or IPv6 stateless address
autoconfiguration way. These IPv6 address starts with fe80 for the local
network address. The DHCPv6 server will have the addresses in the
database which can be assigned for every client on their entry or presence to
the cluster. In the case of stateless method every IPv6 node need to identify
the unique address in the network where the node has been associated.
The functionality of this network has been similar to IPv4 network. The
cluster head runs DHCPv6 server. This server configures the member node
which runs client process. The clients are assigned with IPv6 address as part
of configuration.
ISSN: 1694-2108 | Vol. 2, No. 1. JUNE 2013 7

IJCSBI.ORG
fe80:302:11:1:
1220:0:0:2
N2
fe80:302:11:1: fe80:302:11:1 fe80:302:11:1:

1220:0:0:1 :1220:0:0:0 1220:0:0:4
N4
N1 CH
fe80:302:11:1:
1220:0:0:3
N3
Figure 5. IPv6 Network Model with Stateful address configuration
1) IPv6 nodes stateful address Configuration

Algorithm:
1) Input the Clusters Created using W-PAC procedure.
2) Assume CH(Cluster head) node as DHCPv6 server.
3) for each cluster Ci
Begin
Consider each node within Ci as DHCPv6 Client.
The Client sends hello message to DHCPv6 server.
The DHCP Server Configures the DHCPv6 Client
Node.
End
4) Repeat the step 3 for i = 1..... Number of clusters.
The DHCPv6 server assigns address dynamically to the IPv6 nodes. These
addresses can be reused while the node moves out of the cluster.
fe80:302:11:1
:1220:0:0:0
fe80:302:11:1:
1220:0:0:2
N2 CH
fe80:302:11:1: fe80:302:11:1 fe80:302:11:1:

1220:0:0:1 :1220:0:0:0 1220:0:0:4
N4
N1 CH
fe80:302:11:1:
1220:0:0:3
N3
Figure 6. Mobility of IPv6 Cluster

head
ISSN: 1694-2108 | Vol. 2, No. 1. JUNE 2013 8

IJCSBI.ORG
The Figure 6 shows the mobility of the IPv6 node acts as cluster head. The
movement of cluster head will elect the member node as next cluster head.
This will be decided based on the weight values computed for all nodes. The
next node which has least weight will be elected as cluster head. This
process will be limited by the clusters boundary.
fe80:302:11:1:
fe80:302:11:1: 1220:0:0:2
1220:0:0:2 N2
N2
fe80:302:11:1: fe80:302:11:1 fe80:302:11:1:

1220:0:0:1 :1220:0:0:0 1220:0:0:4
N4
N1 CH
fe80:302:11:1:
1220:0:0:3
N3
Figure 7. Mobility of IPv6 node

The Figure 7 shows the mobility scenario of the member node. The node
movement out of the cluster boundary will be known by the DHCPv6
server. Since the server monitors the nodes status by periodically sending
the probing packet. The server will reuse the IPv6 address of the node which
has moved out of the cluster. If the node enters to new cluster the probe
message will reach the member node from the cluster head of that cluster.
The node may send a request to the cluster head while the probe message
reaches first time to the member node.
2) IPv6 nodes stateless address Autoconfiguration
Broadcasting
N1
IPv6address
Cluster
head
N2
N3
Figure 8. Cluster head broadcasting IPv6

address
The Figure 8 shows the cluster head broadcasting of IPv6 address to all the
member nodes of the same cluster. This cluster head is elected with the help
of W-PAC algorithm.
ISSN: 1694-2108 | Vol. 2, No. 1. JUNE 2013 9

IJCSBI.ORG
Algorithm: Autoconfiguration of IPv6 nodes

(1) Input the clusters formed through W-PAC.
Clusters : C1, C2......Cn
(2) Cluster head is elected based on weight.
(3) i = 0.
// Cluster head broadcasts IPv6 address
(4) For each cluster Ci
Begin
Repeat
Begin
j=j+1
CH broadcast IPv6 address to Nj
End
Until ( j < number of Nodes under Ci )
End
(5) Repeat the step 4 for i = 1........number of clusters.
// Node Affiliation
(6) For each node Nj of Ci
Begin
Nj sends IPv6 address to cluster head
Cluster head adds to member list
Cluster head verifies the IPv6 address of Node Nj
Cluster head replies Unique/Duplicate to member node.
End
(7) Repeat the step 6 for j = 1....number of nodes of Ci
(8) Repeat the step 7 for i = 1........ number of clusters.
Node Cluster
IPv6 address
head
Unique/ Duplicate
Figure 9. Node confirms unique IPv6 address with the cluster head
Figure 9 shows the node communication with the cluster head to detect the
duplication in the selected address under stateless autoconfiguration method.
The cluster head reply conveys the uniqueness of the address which is
obtained by the member node.
Algorithm: CHDAD
(1) ADD_LIST ={ Empty }
(2) Duplicate = false.
(3) Member node N sends IPv6 address to CHi of the cluster Ci.
(4) For each cluster Ci
ISSN: 1694-2108 | Vol. 2, No. 1. JUNE 2013 10

IJCSBI.ORG
Begin
The Cluster head CHi validates the address
Begin
If ( IPv6 of N ADD_LIST ) then
Reply to member node N about the duplication of
IPv6 address.
Duplicate = true
End
i=i+1
End
(5) If (Duplicate = false) then
Add the IPv6 address of node N to ADD_LIST.
(6) Repeat the steps 2 through 5 for all the node of the cluster Ci.
(7) Repeat the steps 1 through 6 for all the clusters of the network.
The CHDAD (Cluster heads duplicate address detection) mechanism takes
the IPv6 address decided by the member nodes . This will be probed against
the duplication by the respective CH head by checking the existence of the
address in their list ADD_LIST or neighbors list. If the address is not found
in this list then the address gets added to the list of the respective cluster
head. The cluster head node should have the information stored in the form
of table as shown in Table.1. This will be exchanged among the cluster
heads on demand.
Table 1. Data structure of CHs Info Packet Fields
List of Affiliated List of Neighbor

Nodes IPv6 address CHs IPv6 address
( ADD_LIST )
The Figure 10 shows the IPv6 address autoconfiguration in the network

model.
In this model the cluster head is elected based on the weight value as similar
to IPv4 network model. The cluster head role has been changed in this
network. The nodes need to configure themselves since there is no DHCP
server as for as this network model has been concerned. The cluster head is
useful on detecting duplicate address. Initially the cluster head will
configure itself automatically then it will broadcast the address to all the
nodes belong to the cluster. After that, each node chooses the IPv6 address
ISSN: 1694-2108 | Vol. 2, No. 1. JUNE 2013 11

IJCSBI.ORG
and immediately informs the cluster head node. The cluster head maintains a
list of IPv6 addresses which are allotted.
fe80:302:11:1:
1220:0:0:2
N2
fe80:302:11:1: fe80:302:11:1: fe80:302:11:1:

1220:0:0:1 1220:0:0:0 1220:0:0:4
N4
N1 CH
fe80:302:11:1:
1220:0:0:3
N3
Figure 10. Stateless address autoconfiguration of IPv6

nodes
Thus, the cluster head will be aware of the addresses already in use. The
member node communicates with cluster head to detect the duplicate
address. In this way the bandwidth consumption has been reduced
significantly. Since the broadcasting of the node address to detect the
duplicate address has been eliminated. As a result this the performance
using the clustering mechanism has shown an improvement in terms of
efficiently utilizing the service parameters.
6. EXPERIMENTAL RESULTS
The simulation has been done with the help of OMNET++. The following
Table.2 shows the simulation parameters that are considered for the
simulation.
Table 2. Simulation Parameters
Parameter Values
N (Number of IPv6
10
Nodes)
Space (area) 30 30
Tr (Transmission
6m
range)
Ideal nodes percentage 90%
Simulation Time 5 sec

Weighing Factors 0.7,0.2,
(w1,w2,w3,w4) 0.05,0.05
Threshold(Energy) in
500
Units
ISSN: 1694-2108 | Vol. 2, No. 1. JUNE 2013 12

IJCSBI.ORG
Table 3. Nodes Weight
Cluster Node Weight

(WNi)
- N0 73.45
C2 N1 107.90
C1 N2 158.70
- N3 208.40
C1 N4 258.70
C1 N5 308.75
C1 N6 58.70
- N7 133.40
C1 N8 352.95
C2 N9 127.90
The Table.3 shows the computed weight value of the nodes. The results
show that the cluster1 and cluster2 will have cluster heads as node6 and
node1 respectively. These are decided based on the weight value of the
nodes belong to each cluster. The Figure 11 shows the clusters formed using
set of nodes. The W-PAC cluster formation algorithm forms two clusters for
the given sample set. These nodes later on configured with IPv6 address
either by stateless or stateful way.
30
25 20, 24
Node Position(m)
20 26, 20
C1
15 20, 12 24, 14 30, 14
27, 12 32, 12
10 28, 10
10, 10 21, 8
5
C2
0
0 10 20 30 40
VII. FUTURE DIRECTION
Node position(m)
Figure 11. Result of W-PAC
The Figure 12 shows the cluster C1 comprised of member nodes and cluster
head as node6. The CH6 node runs DHCPv6 to configure the nodes under
ISSN: 1694-2108 | Vol. 2, No. 1. JUNE 2013 13

IJCSBI.ORG
stateful method using IPv6 address. The node6 will broadcast the IPv6
address to the member nodes in order to communicate with the server node.
The member node sends the request and gets the IPv6 address from the
server as their unique address.
Broadcasting
N2 N5
CH6 Request
Reply
N4
N8
Figure 12. Configuration of Cluster1 nodes
400
350 352.95
300 308.75
Weight of CH
250 258.7
200 208.4
150 158.7
133.4127.9
100 107.9
73.45 58.7
50
0
0 1 2 3 4 5 6 7 8 9
CH Nodes
Figure 13. CH Nodes based on Weight
The Figure 13 shows the cluster heads order based on weight to change the
cluster heads when the existing cluster head moved away.
7. CONCLUSIONS
This study describes the cluster formation of IPv4 and IPv6 nodes. It has
pointed out the methods of forming clusters while using IPv6 nodes. The
cluster maintenance strategy after the cluster formation using W-PAC has
also been discussed. The simulation with the help of OMNET++ shows the
study to be applied in a realistic manner. In future this work has to be
extended for huge sample set and also consider forming the clusters
dynamically.
ISSN: 1694-2108 | Vol. 2, No. 1. JUNE 2013 14

IJCSBI.ORG
REFERENCES
[1] S. Thirumurugan and E. George Dharma Prakash Raj, PAC - A Novel approach For
Clustering Mechanism in Adhoc Network, ICSCCN 2011, pp.593-598.
[2] S. Thirumurugan and E. George Dharma Prakash Raj, Ex-PAC: An Improved
Clustering Technique in Ad Hoc network, RACSS12, 2012, pp.195-199.
[3] S. Thirumurugan and E. George Dharma Prakash Raj, W-PAC: An Efficient Weighted
Partinioning Around Cluster head Mechanism for Adhoc Network, CCSEIT12, pp.182-
188.
[4] S. Thirumurugan and E. George Dharma Prakash Raj, PSO-PAC: An Intelligent
Clustering Mechanism in Ad hoc Network, NETCOM 12.
[5] S. Thirumurugan, Direct sequenced C-IAODV Routing Protocol, International
Journal of Computer Science and Technology, Vol.1, Issue.2, Dec 2010, pp. 108-113.
[6] S. Thirumurugan, C-AODV: Routing Protocol for Tunnels Network, International
Journal of computer science and technology, Vol.2, Issue.1, Mar2011, pp.113-116.
[7] Li Qing, An IP address auto-configuration scheme for MANET with global
connectivity, ICIME10, 2010, pp.244-247.
[8] M. Sivakumar, C J Jickson and R M S Parvathi , Passive Duplicate Address Detection
in DYMO Routing Protocol for MANETS, ICVCI11, 2011, pp.9-14.
[9] He Zhonglin and He Yuhua , Study on Key Technologies of MANET, ISME10,
2010, pp.112-115.
[10] Ulrich Herberg and Thomas Clausen, Yet Another Autoconf Proposal (YAAP) for
Mobile Ad hoc NETworks, MASN10, 2010, pp.20-26.
ISSN: 1694-2108 | Vol. 2, No. 1. JUNE 2013 15

IJCSBI.ORG
A New Online XML Document

Clustering Based on XCLS++
Ahmad Khodayar E Qaramaleki
Department of Computer Engineering
Shabestar Islamic Azad University of Science and Technology
Shabestar, Iran
Hassan Naderi
Department of Computer Engineering
Iran University of Science and Technology
Resalat St., Tehran, Iran
ABSTRACT
Many methods have been proposed to XML document clustering. These methods can be
divided into three categories: structure-based, content-based and hybrid methods. XCLS++
is one of the most effective and efficient algorithms to XML document clustering which fit
into the structural clustering category. Because of its efficiency, XCLS++ can be used
XML stream clustering. In this paper, we will show one of the weaknesses of this method
and then we will try to solve it by deleting a factor in the XCLS++formula. As we will
show, this factor is related to the node weight in a tree which represents a given XML
document. According to our experimentations which have been presented in this paper, the
effectiveness (in term of accuracy) and efficiency (in term of execution time) of XCLS++
can be improved once this weight factor is eliminated from the original XCLS++ formula.
Keywords
Clustering, XML documents, XCLS++, Structure similarity, Content similarity.
1. INTRODUCTION
One of the methods used to extract information from databases is data
mining which is used in the search engines. The appropriate structure to
store data in databases contributed to data mining. The easy execution of
data mining in database causes increasing of search engine efficiency. So,
using the proper structure in a database is a crucial point. One of the good
ideas for using appropriate structure in database is clustering.
During past years, various formats like HTML and XHTML are presented
for showing documents using the content and the structure. Because XML
documents are used to transfer and to search in documents and also the
usage of this technology is increased day by day, a good management on
these documents is vital.
ISSN: 1694-2108 | Vol. 2, No. 1. JUNE 2013 1

IJCSBI.ORG
In general, XML document clustering methods can be divided into three
categories: 1- Structural, 2-Content-basedand 3- Combination of structure
and content (hybrid). In the structural approach, an important criterion in
clustering is the structure of the document. In the other word, the structural
similarity between documents is a criterion for placing them in the same
class. In the content-based method, the criterion of clustering is the
similarity between the texts of both documents, and finally in the hybrid
approach, the content and the structure similarity together are criterions to
clustering two documents.
XCLS++ is one of the structured based methods to XML document
clustering [12]. Its an improved version of XCLS+ algorithm which is an
efficient method for clustering XML documents [9]. Our studies showed
that XCLS++ can be improved, because it has some problems which make it
away from optimal output. In this paper, our focus is on XCLS++ method
and new method has been proposed with better performance. In the next
section, the related articles are investigated. In the section 3 the
XCLS++method and in the Section 4 the problem of XCLS++ method are
presented. In the Section 5 our new proposed method is explained and then
it will be compared with the other algorithms in the next section. Execution
times of the algorithms are compared in Section 7, and the last section is
allocated to conclusion.
2. PREVIOUS WORKS
The criterion of clustering is based on the similarity of documents. As
mentioned above, there are three ways to find the similarity of documents:
1- structural [2][5][10][11][12]which consider only the structure of the
document2- content-based [9] which considers only the content and finally
3- hybrid (content with structure) [1][3][4][8] which consider both the
content and the structure. As we know, each XML document can be
transferred into a tree and then clustering operations can be done with those
trees. Structural methods only consider the structure and do not pay
attention to content. The XCLS++ is one of the most efficient online
algorithms to XML document clustering that fit into the structural clustering
category. Our focus in this paper is concentrated on this type of clustering.
Based on our studies, we have seen some problems in the XCLS++formula,
which makes it away from optimal value. The presented solution in this
paper can solves the problem of XCLS++ and optimizes it. Details of this
method and its evaluations will be presented in the next sections.
3. THE FORMULA OF XCLS++ METHOD
As previously mentioned XCLS++ method is an example of structural
approach and it works based on the similarity between tag names. So it
doesnt consider the content of XML documents. XCLS++ by make some
changes on XCLS+ at two stages could solve existent drawback of XCLS+
ISSN: 1694-2108 | Vol. 2, No. 1. JUNE 2013 2

IJCSBI.ORG
[11]. But in this paper we will show that those changes are not enough to
make optimization. Some trees which will be shown later will prove
inefficiency of XCLS++ in some cases. In this paper we will try to eliminate
these inefficacies. The operations of XCLS+ and XCLS++ are similar. In
this way like the XCLS+ method the incoming XML document is compared
to clustered documents. If the value of similarity is greater than are equal to
a threshold (t), the XML document will be placed in the relevant cluster.
Otherwise, new incoming document is clustered in a new cluster. This
process is continued until the last document is entered. The similarity
calculation is performed based on a formula. In what follow, the XCLS++
details will be explained and calculation of similarity value between two
trees is shown. Then problem of XCLS++ has been presented. After that a
solution to this problem will be introduced. The formula of XCLS++, which
is an improved formulation of the XCLS+ method, is as follow:
+ +

.
=( + + )

+ .
=( + + )
=
(
= ) + .
= (

+
) +
=( + )

Formula 1. The formula of XCLS++ method

In the Formula1the value is between zero and one which is a positive feature
of this formula. The variables used in th is formula are:
1. Z is cluster size or in other words the number of documents within cluster.
2. CNi: is sum of incoming nodes which is similar to clustered document in level i.
3. CNj: is sum of incoming nodes which is similar to clustered document in level j.
4. CBi: is sum of incoming nodes which is similar&&samebrother to clustered
document in level i.
5. CBj: is sum of incoming nodes which is similar&&samebrother to clustered
document in level j.
6. CCi: is sum of incoming nodes which is similar&&samebrother&&samechild to
clustered document in level i.
7. CCj: is sum of incoming nodes which is similar&&samebrother&&samechild to
clustered document in level j.
8. l: is high of tree in the each document.
9. i, j:are related numbers of level.
10. r: is the incremental factor, which is considered number 2.
11. k: is equal to 2.
12. N: is sum of clustered nodes of related levels.
The algorithm of XCLS++ clustering method is as follow:
ISSN: 1694-2108 | Vol. 2, No. 1. JUNE 2013 3

IJCSBI.ORG
1- Represent a XML document by a relative tree.
2- Consider this tree with a tree related to a cluster.
3- Start to search same node in two trees from root node. If a node is found then do
the calculation of the Formula1.Then go to step2, otherwise go to Step3.
4- If depth of trees moves toward the lower level in the both trees. Search the same
node as step1. If there is same node, calculate the Formula1 and repeat step2,
otherwise go to step3.
5- If depth of tree is (usually in clustered document), move toward down level in the
clustered document and stay in the same level of the new incoming document.
Search again the same nodes. If there is the same node, calculate the formula and
then repeat step2, otherwise repeat step3.
3.1. An Example of XCLS++

For understanding the algorithm an example is given in this section. Our
goal is to find the similarity between tree1 and tree2 based on the XCLS++
method. It should be noted that the tree1 referred to the incoming document
and the tree2 referred to the clustered documents. Dotted arrows from left to
right indicate the order of execution of algorithm steps. In this example for
facility the variable values are calculated and placed on the dotted arrows
tree1 tree2
cn=1 cb=0 cc=1
0 0 L= 0
,
1 cn=1 cb=0 cc=2
1 1 L= 1
cn=2 cb=2 cc=2
2 3 2 3 L= 2
, ,
, , 1 1
1 1
Figure 1. An example for showing work the XCLS+ method

After obtaining above factors, the similarity value base on the XCLS++
method will be equal to:
0.5
1 + 0 + 1 22 + 1 + 0 + 2 21 + 2 + 2 + 2 20 + 0.5 1 + 0 + 1 22 + 1 + 0 + 2 21 + 2 + 2 + 2 20
1 22 +1 21 +2 20 1 + 0.5 1 22 +2 21 + 4 20 + 0.5 1 22 + 2 21 + 4 20
=
4. THE PROBLEM OF XCLS++

As was mentioned, XCLS++ method was presented to improve XCLS+
method. The primary cause of inefficiency XCLS+ is ignoring nodes
repetitions in the original formula. This reason cases hierarchy of the nodes
is changed. XCLS++ has been proposed to solve this problem. Dispute this
improvement, studies show that XCLS++ algorithm has some problems too
which can be further improved. In order to illustrate the problem, two
examples are cited. The similarity value for the first example (Figure 2)
calculated by XCLS++ method, is 1. Also the similarity value for in the
ISSN: 1694-2108 | Vol. 2, No. 1. JUNE 2013 4

IJCSBI.ORG
second example (Figure 3) with XCLS++ method is 1.4. Note that factors
without parenthesis () on the dotted arrows belong to both trees.
tree1 tree2
cn=0 cb=0 cc=0 0 L= 0
,
10 L= 1
1
,
cn(tree1)=1 cb(tree1)=0 cc(tree1)=2 10 L= 2
2 3 cn(tree2)=1 cb(tree2)=0 cc(tree2)=2
1 1 cn=2 cb=2 cc=2
1 L=3
2 3 L=4
,1 ,1
Figure 2. Same nodes of both trees are in quite different levels
tree1 tree2
cn(tree1)=1 cb(tree1)=0 cc(tree1)=2
L= 0
cn(tree2)=1 cb(tree2)=0 cc(tree2)=2 1
1
cn=2 cb=2 cc=1 2 3 L= 1
, ,
1 1
2 3
0 L= 2
, ,
1 1
0 L= 3
0 L= 4
1
Figure 3. Same nodes of both trees are in the same upper levels
These results are unreal. Because similarity values in Figure 2 and Figure 3
which only differ in placed levels, must be closed together and less than
one. But similarity values calculated with XCLS++ are very different. The
reason of this difference is related to the weighting in XCLS++ formula.
Our new formula is capable to solve this problem. In what follows, the
proposed method is discussed in more details.
5. NEW METHOD: THE OTHER CHANGE ON THE XCLS++
The main drawback of XCLS++ method which is derived from XCLS+ and
XCLS methods is its weighting factor. The weight causes high level nodes
in a tree be hardly compared to low level nodes in another tree. So we must
reduce the effect of the levels in order to minimize the difference between
two such trees. A simple change can solve the problem of the formula
ISSN: 1694-2108 | Vol. 2, No. 1. JUNE 2013 5

IJCSBI.ORG
XCLS++. As previously mentioned, the main reason of this problem is due
to the weighting factor. By removing the weight factor, this problem can be
solved. So by deleting this factor the expected results can be earned. Finally,
the proposed formula without weighting factor will be:
+ +

.
=( + + ) + .

=( + + )
=
(
= ) + .

=( + ) +

=( + )
Formula 2. The new formula to use in clustering XML documents

In this formula all variables are equivalent to variables of XCLS++. The
new Nk parameter is the average of nodes clustered and by incoming trees.
N of XCLS++ method causes results dependent to clustered or incoming
tree and then unreal results. So for getting better results in the new method,
N will be the average of nodes. For proving the effectiveness of this
formula, the similarity the previous trees are calculated with new algorithm
again. Results for two groups according new algorithm are: 0.85 for Figure
1 and 0.84 for Figure 2. These results have easily obtained with replacing
variables in new formula. The calculated results show that new algorithm
has more effectiveness and these results are near to reality. Its good to
mention that by deleting the weight factor we can increment the speed of
algorithm too. So we can say the efficiency of our algorithm is greater than
XCLS++ efficiency too. If k and h be the depths of two trees, the time
complexity in the worst condition will be O[2^(h+k)] and in the best
condition will be O(2^h) if h is depth of clustered tree.
6. EVALUATING ALGORITHMS
Our justification showed that the effectiveness as well as the efficiency of
new method is better that the two other methods. For proving this sentence
in this section the new way, XCLS++and XCLS+ algorithms have been
implemented and compared. All of them were implemented with C
language in DOS environment on a machine with 2.4 GHZ Intel Celeron
CPU and 512 MB of RAM. The evolution criteria were implemented for
evaluating XML files in the same conditions too. As we will see, the results
of experiments like above examples, confirm optimality and efficiency of
the proposed algorithm is higher than the two other methods.
6.1. Dataset to evaluation
For evaluating, files have been considered from two addresses [6] and [7].
6.2. Evaluation criteria
There are three items for calculating accuracy clustering algorithms: 1-
entropy 2-purity 3-fscore
6.2.1 Entropy
ISSN: 1694-2108 | Vol. 2, No. 1. JUNE 2013 6

IJCSBI.ORG
Entropy is sum documents which located in the cluster i which are of the
class r. The entropy formula is:

=

=

( ) =

=
In above formulas , N, k, and nri are respectively ith cluster, total

number of incoming documents, number of clusters, number clustered
documents in cluster i and number clustered documents in cluster i of class
r. If the entropy value be closer to zero the efficiency is better.
6.2.2 Purity
Purity is sum maximum documents which located in the cluster i which are
of the class r. The purity formula is:

= =
as

=

If the purity value be closer to one the efficiency will be better.

6.2.3 Fscore
Fscore is another item created by combination of above two items and is:

= ,
=

as
, ,
, = =
, + , +
,

, =

,

, =

If the fscore value be closer to one the efficiency will be better.

After implementation, results are calculated and compared for analyzing. In
order to testing implemented program, incoming XML files consists of 1000
different classes, such as medical files, colleges, shops, cars, insurance,
ISSN: 1694-2108 | Vol. 2, No. 1. JUNE 2013 7

IJCSBI.ORG
etc. Files were evaluated with the algorithms. The results of algorithm are
in Table 1 and include:
Table1. Results of algorithms on XML files

ENTROPY PURITY FSCORE
ALGO XCL XCL XCL XCL XCL XC XCL XCL XC

RITHM S++ S++ S+ S++ S++ LS+ S++ S++ LS+
----------
------ impr impr impr
THRES oved oved oved
HOLD
0.7 0.22 0.26 0.30 0.95 0.90 0.75 0.88 0.80 0.80
0.8 0.18 0.25 0.26 0.80 0.80 0.78 0.90 0.80 0.79
0.9 0.20 0.30 0.30 0.96 0.83 0.80 0.95 0.90 0.88
As previous examples, the results obtained in this section shows that the
new method has higher effectiveness than both XCLS++ and XCLS+
methods. Not only optimality of new method is high but also deleting
weight factor causes decreasing execution time. This topic is discussed in
the next section.
7. COMPARING EXECUTION TIME
After comparing effectiveness of algorithms, in this section execution time
is compared too. As previously cited weighting factor has lost its efficiency
due to the fundamental change on the basic formula. With deleting weight
factor amount of calculation is deleted and running time is predictably
decreases. Algorithms have been simulated in MATLAB 6.5.1 for reaching
expected results. This simulation has been done on two same trees with
XCLS++ and new algorithms. Simulated results are:
ISSN: 1694-2108 | Vol. 2, No. 1. JUNE 2013 8

IJCSBI.ORG
Figure 4. Same nodes of two trees are in up levels

As Figure 4 shows the increasing of the depth of trees causes the
augmentation of calculations and running time for new method is lower than
XCLS++ method. We must mention that the running time for the XCLS+ is
same as XCLS++ because weighting factor is in both of the m.
8. CONCLUSION
The purpose of this paper is clustering of XML documents. Criteria for
clustering are structural, content-based, and hybrid (the structure with the
content). XCLS++ method is a clustering method in which criteria for
clustering is done based on the structure. Despite good performance for it in
comparison to the previous methods, weighting nodes in some documents
causes it to be inefficient. Therefore, new algorithm with deleting it has
been proposed. The results of entropy, purity and Fscore calculation show
that the proposed method works better than the previous method. In future
new weight of levels will be replaced for obtaining a similarity actually and
better than proposed method too. Also in future the new method will be
evaluated on much more documents and by comparing those we will be able
to obtain better results.
9. REFERENCES
[1] Ilwan Choi, Bongki Moon, Hyoung-Joo Kin, A clustering method based on path
similarities of XML data, Data & Knowledge Engineering, 2006.
[2] Andrewdn, Jag, Information systems engineering, Evaluating Structural Similarity
in XML Document, WISE07 Proceedings of the 8thinternational conference on
Web Information, 2007.
[3] Tien Tran, Richi, Peter, Data Mining, Combining Structure and Content
Similarities for XML Document Clustering, Conference 27-28November, Glenelg,
South Australia, 2008.
ISSN: 1694-2108 | Vol. 2, No. 1. JUNE 2013 9

IJCSBI.ORG
[4] Woosaeng Kim, Computer Engineering and Applications, XML document
similarity measure in terms of the structure and contents, CEA'08 Proceedings of
the 2nd WSEAS International Conference, 2008.
[5] G. R. Nayak, Fast and effective clustering of XML data using structural
information knowledge. Information System, 2008.
[6] The Wisconisns XML data bank. Accessed
from:http://www.cs.wisc.edu/hiagara/data.html Cited2012.
[7] The XML data repository. Accessed from:
http://www.cs.washington.edu/research/xmldatasets/. Cited 2012.
[8] Waraporn Viyanon, Sanjay K. Madria, Sourav S. Bhowmick, Management of
Data, XML Data Integration Based on Content and Structure Similarity Using
Keys, 2008.
[9] Aptarshi Ghosh and Pabitra Mitra, Pattern recognition, ICPR Combining Content
and Structure Similarity for XML Document Classification using Composite SVM
Kernels, 19th International Conference, 2008.
[10] Jing Peng Dong Qing Yang Shi Wei Tang et al, similarity in chinese text
processing, A New Similarity competing method based on concept, series F:
Information science, 51(9): p1212-1230, 2008.
[11] Mohamad Alishahi, Mohmoud Naghibzadeh and Baharak Shakeri Aski, Tag
Name Structure-based Clustering of XML Documents, International Journal of
Computer and Electrical Engineering Vol. 2, No. 1, February, 2010.
[12] Ahmad Khodayar and Hassan Naderi, XCLS++: A new algorithm to improve
XCLS+ for clustering XML documents, International Journal of Information
Technology, Control and Automation (IJITCA) Vol.2, No.4, 2012.
ISSN: 1694-2108 | Vol. 2, No. 1. JUNE 2013 10

IJCSBI.ORG
Secured and Energy Based QoS

Routing in MANETs
S. Sridhar
S.A.Engineering College
Poonamallee-Avadi Road,Chennai -77, India
R. Baskaran
CEG Guindy, Anna University
Chennai 600025, India
ABSTRACT
A Mobile ad-hoc network (MANET) is a wireless network, self-configuring, capable of
self-directed operation, hastily deployable and operates without infrastructure. MANET
operates without centralized administration. The nodes are self configuring, independent,
quickly deployable. Nodes are movable since topology is very vibrant and they have
restricted energy and computing resources. Routing protocols should incorporate QoS
metrics in route finding and maintenance, to support end-to-end QoS. General AODV
routing faces problems like long route, time delay, mobility and many other while routing.
The nodes low in energy level will not be in a position to complete the routing. The QoS
parameters like throughput, PDR and delay are affected directly. The proposed protocol
EN-AODV, announces energy and based on nodes sending and receiving rates and the sizes
of the data to be transmitted it justifies whether its energy level is maintained or decreased.
It calculates the energy levels of the nodes before they are selected for routing path. A
threshold value is defined and nodes are considered for routing only if its energy level is
above this threshold value. The transmissions are made secure by introducing message
digest algorithm. The nodes perform routing and all these routing are secured using MD5
algorithm which increases the reliability of routing. The work is implemented and
simulated on NS-2. The simulation results have shown an increase in PDR, decrease in
delay and throughput is maintained. The proposed EN-AODV provides more consistent and
reliable data transfer compared to general AODV.
Keywords
Ad-hoc, MANET, AODV, EN-AODV, QoS, MD5.
1. INTRODUCTION
Mobile ad-hoc network is an extraordinarily testing vibrant network. They
do not rely on existing infrastructure to support communication. Each
mobile node acts as an end node when it is the source or destination of a
communication and forwards packets for other nodes when it is an
intermediate node of the route Mobile Ad-Hoc network [1] is a system of
wireless mobile nodes that self-organizes itself in dynamic and temporary
network topologies. Ad hoc networks are easier to organize than wired
networks and are used in many applications, such as in human or nature
induced disasters, battlefields, meeting rooms where either a wired network
is unavailable or deploying a wired network is inconvenient. MANETs are
characterized by self-configured, dynamic changes of network topology,
ISSN: 1694-2108 | Vol. 1, No. 1. MAY 2013 1

IJCSBI.ORG
limited bandwidth, instability of link capacity and other resource constraints.
This dynamic nature of MANET makes it enormously complicated to obtain
accurate knowledge of the network state and thats why the consistency of
data transmission in this network cannot be guaranteed.
There have been many MANET routing protocols, which fall into several
categories: proactive routing protocols such as dynamic Destination-
Sequenced Distance-Vector routing (DSDV), Optimized Link State Routing
(OLSR), Topology Broadcast based on Reverse Path Forwarding (TBRPF),
on-demand routing protocols such as Dynamic Source Routing (DSR),
Adhoc on demand distance vector (AODV), Signal Stability-based Adaptive
routing (SSA). Proactive routing protocols have little delay for route
discovery and are robust enough to link breaks and obtain a global optimal
route for each destination. However, their routing overhead is also high. On-
demand routing protocols are easy to realize and their overhead is low. But
routes in on-demand routing protocols are easy to break in the case of
topology variations. In AODV [2] node doesnt have any information about
other nodes until a communication is needed. By broadcasting HELLO
packets in a regular interval, local connectivity information is maintained by
each node. Local connectivity maintains information about all the
neighbours.
In ensuring QoS provisioning, a network is expected to guarantee a set of
measurable pre-specified service attributes to the users in terms of end-to-
end performance, such as challenging task to ensure QoS provisioning
including routing in ad-hoc networks due to the mobile and dynamic nature
of the nodes. Recent QoS solutions are planned to operate on trusted
environments and totally assume the participating nodes to be cooperative
and well behaved [3, 4]. The major drawback of conventional AODV
protocol is the absence of the Quality of Service (QoS) provision that make
routing protocols which requiring applications of QoS lower efficiency.
MANETS usually consist of mobile battery operated devices that
communicate over the wireless medium. These devices are battery operated
and therefore need to be energy conserving so that the battery life of each
individual node can be extended. To make the most of the lifetime of an ad
hoc network, it is essential to lengthen each individual node life through
minimizing the total transmission energy consumption for each
communication request. Therefore, an efficient routing protocol must satisfy
that the energy consumption rate at each node is evenly distributed and at
the same time the total transmission energy for each request is minimized.
The transaction made by nodes in MANET should be a secured transaction.
To provide security for all transactions Message digest algorithm in
ISSN: 1694-2108 | Vol. 1, No. 1. MAY 2013 2

IJCSBI.ORG
introduced during transmission. All transmissions are secured using MD5
Algorithm. Thus it increases the reliability of routing.
Therefore, energy for nodes needs to be considered while routing since
nodes may drain out of energy levels. Though a node is providing its
complete support for routing it can perform well only if it has sufficient
energy. MD5 algorithm is also introduced to secure transmissions and
increases the reliability in routing. Traditional AODV does consider the
energy levels of nodes before routing. Energy is announced by the proposed
AODV protocol that checks for energy levels of nodes before taking part in
routing in order to make the MANET routing efficient and effective and also
ensure QoS.
2. LITERATURE SURVEY
An energy efficient routing protocol for maximizing lifetime in MANET [5]
is introduced. If the network is divided into more than two, and one of the
nodes consumes all the energy, that node can no longer participate in the
network. In recent years, more works has been under taken to not only
improve the energy storage but also to lengthen the networks lifetime. A
enhanced AODV routing protocol is presented which is modified to
improve the networks lifetime in MANET .One improvement for the
AODV protocol is to maximize the networks lifetime by applying
an Energy Mean Value algorithm which considerate node energy-aware.
An energy consumption analysis based on mobility models [6] is discussed
to know which protocol is better than another in different mobile network
scenarios, four mobility models are proposed for simulating different
scenarios of mobile ad hoc networks. Also a byte-based energy consumption
evaluation methodology is introduced for the protocol assessment. The
experiment built upon mobility models show that it is fit for the mobile ad
hoc network with low node mobility, while AODV, DSR, and especially
DSDV perform well on energy consumption for the mobile ad hoc network
with high node mobility.
A novel cross layered energy based AODV protocol [7] is proposed. A
dynamic energy conscious routing algorithm ECL-AODV where cross layer
interaction is provided to utilize the energy related information from
physical and MAC layers. This algorithm avoids the nodes which are having
low residual energy. By maximizing the lifetime of mobile
nodes routing algorithm selects a best path from the viewpoint of high
residual energy path as part of route stability. The RTS/CTS transmission is
a crucial step towards saving the energy of mobile nodes. In this scheme, the
RTS/CTS transmission occurs after route discovery and route reply process.
ISSN: 1694-2108 | Vol. 1, No. 1. MAY 2013 3

IJCSBI.ORG
The path is reserved for further transmissions. The receiving power of
sender, intermediate nodes and receiver are also another part
of route stability. The protocol is implemented for achieving quality of
service (QoS) in terms of average energy consumption, packet delivery
ratio, end-to-end delay and throughput
An energy level based routing protocols-ELBRP [8] that not only makes the
system energy consumption down but also prolongs the system lifetime and
improves the delay characteristic. The proof of correctness and complexity
analysis of ELBRP are presented and also compares the performance of
existing protocols. The studies show that ELBRP has a better delay
performance, and lower energy consumption and longer network lifetime.
The analysis are based on the comparison of two energy-based mechanisms
called E-AODV, an energy consumption rate-based routing protocol, and F-
AODV, a cross-layer-based routing protocol [9]. The trends and the
challenges on designing cross-layer communication protocols for MANETs
are investigated. The results show that the performance of the layer
cooperation paradigm depends on the network characteristics and the
application constraints.
An energy efficient integrated routing protocol (E2IRP) [10] for mobile ad
hoc networks used in remote surveillance systems is presented. The
integration of MAC and routing layers can effectively reduce the amount of
control information being exchanged for discovery and maintenance of the
route in the network. This in turn reduces the energy and time consumed for
the processing of these packets. Though the number of packets and
processing is less, the protocol provides a better reliability and throughput.
The nodes are organized in concentric tiers around the gateway. The event
reports are routed towards the gateway from one tier to another and the
response is routed back to the source, in the same manner. The proposed
E2IRP outperforms traditional AODV routing protocol in terms of battery
power consumption and also the throughput.
A new routing protocol called energy-aware grid multipath routing
(EAGMR) [11] protocol is proposed. The proposed protocol can conserve
energy and provide the best path to route according to probability.
Simulation results indicate that this new energy-aware protocol can save
energy of mobile hosts and improve data packet delivery ratio
A novel energy saving energy routing protocol: ES-AODV [12] is
presented. Nodes made use of the HELLO message mechanism in AODV
and reduced energy consumed by inserting intermediate node iteratively.
From performance analysis and simulation results, it could be found that
ISSN: 1694-2108 | Vol. 1, No. 1. MAY 2013 4

IJCSBI.ORG
ES-AODV had many advantages. Compared to the AODV protocol, ES-
AODV prolonged nodes' lifetime and substantially improved the saving
energy performance.
3. PROPOSED WORK
Many energy management schemes have been proposed to evaluate energy
values and most of the energy based protocols for calculated values based
on the energy consumed by nodes during the transmission. Routing in
mobile ad hoc networks is pretentious due to the dynamic nature of nodes,
which are not stable and keep moving. But still nodes communicate with
each other and exchange data within the available nodes on the network.
The architecture of the proposed work is presented in Figure 1. The node
energy level also plays a very crucial role in MANET routing. Focus is on
identifying the nodes energy level consumed so far and energy level left
over and higher than the threshold value assumed to be half the initial value
of the nodes energy assumed, which should be sufficient for performing the
upcoming transmission. If energy level not sufficient the proposed protocol
selects an alternate path to carry on routing successfully using reliable
nodes.
Figure 1. Architecture of proposed EN-AODV routing in MANETs
The proposed work concentrates on identifying these unreliable nodes

(running low in energy level) using the energy level values calculated for
each node. The energy level value calculation is based on the parameters
shown in the Table 1.
ISSN: 1694-2108 | Vol. 1, No. 1. MAY 2013 5

IJCSBI.ORG
Table 1. Energy value calculation parameters
Parameters Description
The initial energy of each node in the MANET set to
Initial energy
default value 100 J.
Final energy The maximum energy is set to 0.
Nodes The number of nodes that are part of MANET.
Node id Unique Id of each node in MANET.
Energy consumption based on the various events like
Event
R Received, D Dropped, S Sent, F Failed.
Time Time consumed for the event.
The energy consumed by a node to complete the
Consumed energy
transmission successfully.
The total energy consumed by all nodes in the
Total energy
network.
Average energy Average energy consumed by a node.
Energy calculation is based on nodes sending and receiving rate. If a node is

selected for transmission then it should concentrate more on the
corresponding transmission in order to save energy and not to drain out by
involving in unnecessary transmissions. To identify energy level the nodes
are evaluated where sender to increase radio frequencies to identify best
nodes with more energy levels. Current Energy level of node can be
calculated by the initial energy level and the consumed energy level of a
node. Drawback in energy based work is that the source itself may drain out.
In such cases introduce external energy to source node by introducing
Virtual energy concepts. Other nodes have to store energy for future
transmissions.
Energy value calculation procedure:

Step 1: Set initial parameters values as initialenergy = 100, maxenergy=100,
nodes= 50 and Nodeid (unique id for each node)
Step 2: Calculate Intermedenergy based on event, time where events can be

(event="r" || event="d" || event="s"|| event="f")
Step 3: Compute consumed energy for each node;

for (i in Intermednergy) {
consumenergy[i]=initialenergy-Intermedenergy[i]
totalenergy +=consumenergy[i]
ISSN: 1694-2108 | Vol. 1, No. 1. MAY 2013 6

IJCSBI.ORG
if (maxenergy<consumenergy[i]){
maxenergy=consumenergy[i]
nodeid=i}}
Step 4: Compute average energy

averagenergy=totalenergy/nodes
MD5 algorithm for secured transmissions:

The critical job in routing is to identify the attackers in the path. To identify
the attackers we initially set (flag) all nodes as true nodes. Nodes change
their nature only after performing transmissions. Nodes properties
considered are IP Address (IP), Nodes identification (ID), MAC Address
and msg. If any one of these property of a node is altered or changed, we
conclude the node is an attacker. We propose that change in ip address
concludes the node as attacker. Initially ip is set in a sequence. Example set
ip_node0 192.26.2.0; set ip_node1 192.26.2.1; set ip_node2 192.26.2.2; set
ip_node3 192.26.2.3; set ip_node4 192.26.2.4; set ip_node5 192.26.2.5
If any node uses an another IP which is already existing then we conclude
that node as attacker. IP addresses are set in sequence. For instance node 5
and node 10 has same IP address, then check nodes flag whether it is true
or false. It would be true for node 5 but false for node 10 since the IP is
replicating for node 10. Since we have set IP addresses in sequence it is
clear that the current IP address of node originally belongs to node 5.
These algorithms operate on a message 512 bit at a time. Pad the message to
a multiple of 512 bits. Digest calculation begins with digest value initialized
to a constant. This value is combined with first 512 bits of message to
produce a new value for the digest; using a complex transformation. New
value is combined with next 512 bits of message using same transformation
and so on until final value of digest is produced. The main ingredient of
MD5 alg is the transformation that takes input as current value of the 128 bit
digest, plus 512 bits of message and outputs a new 128-bit digest. MD5
operates on 32 bit quantities. Current digest value can be thought of as four
32-bit words(d0, d1, d2, d3) and piece of message currently being digested
(512) as sixteen 32 bit words (M0 through M15).
First pass- New digest is produced from old value and the 16 message
words using 16 steps. Process continues until all 16 words (till M16) have
been digested.
Second pass--same as first pass with following difference. F is replaced by
a slightly diff function G. Constant T1 through T16 are replaced by another
set (T17 through T32) .Amount of left rotation is {5, 9, 14, 20, 59..} at each
ISSN: 1694-2108 | Vol. 1, No. 1. MAY 2013 7

IJCSBI.ORG
step . Instead of taking bytes of message in order M0 through M15, the
message byte that is used at stage i is M (5i+1) mod 16.
Third pass- G is replaced by function H (XOR of its arguments), another
set of constants (T33 thru T48), amount of left rotation{4,11,16,23, 4, 11..} at
each step and message byte used at stage I is M(3i+5) mod16.
Fourth pass- H replaced by function I(combination of bitwise XOR, OR
and NOT), another set of constants (T49 thru T64), mount of left rotation{6,
10, 16, 21, 6, 10...} and message byte used at stage is M7i mod 16 .
4. RESULTS
The proposed EN-AODV protocols performance is analyzed using NS-2
simulator. The network is planned and implemented using network
simulator with maximum of 50 nodes and other parameters based on which
the network is shaped are given in Table 2. The simulator is applied with
traditional AODV and with proposed energy based EN-AODV and results
are obtained for assessment. The proposed EN-AODV protocol has shown
good progress over the QoS parameters like PDR and Delay and throughput
is maintained. PDR is increased and delay is reduced compared to the
traditional AODV. The performance of the proposed protocol is also
represented graphically where it clearly shows the betterment of the QoS
parameters. The consumed energy levels of each node are also shown
graphically.
Table 2. Simulation Parameter Values
Parameter Value
Network size 1600 x 1600
Number of nodes 50
Movement speed 100 kbps
Transmission range 250 meters.
Packet size 5000
Traffic type CBR
Simulation time 30 minutes.
Maximum speed 100 kbps
MAC layer protocol IEEE 802.11
Protocol AODV
NS2 version 2.34
ISSN: 1694-2108 | Vol. 1, No. 1. MAY 2013 8

IJCSBI.ORG
Security scheme scenarios:
The nodes are marked initially as true and as transmission starts they start
changing.
Flag for node(1)--------> true
Flag for node(2)--------> true
Flag for node(9)--------> false
Flag for node(28)-------->false
Routing Path : N23 , N1 , N40 , N41 , N48

Mis-behaving Routing : N23 , N9 , N28 , N48
Alternative Routing : N23, N1,N2, N40, N41,N48
Mis-behaving Nodes : Node 9 , Node 28
If node is replicating the IP addresses of another node then same message to

be created for both nodes by MD5. Hence both nodes will be tested with
their flags where node 1 and 2 will be true and node 9 and 23 will be false.
Node 9 and 28 are defined as attackers since they replicate ip address shown
clearly with same message been created using MD5
Routing Node Signature

N1 c4dfd145e649849eb4a66f83c052a8de - Trusted Node
N9 c4dfd145e649849eb4a66f83c052a8de - Replicated as N1
N28 a9913d1a1eaccaa08606200dc92faaac - Replicated as N2
N2 a9913d1a1eaccaa08606200dc92faaac - Trusted Node
Figure 2 shows the snapshot of the simulation where node 9 and node 28 are
marked as attackers because they replicate ip address. Figure 3 shows the
snapshot where transmission goes on between source and destination while
network identifies attackers. Figure 4 shows the routing has taken an
alternate path thus avoiding attackers.
ISSN: 1694-2108 | Vol. 1, No. 1. MAY 2013 9

IJCSBI.ORG
Figure 2. Snapshot showing attackers
Figure 3. Snapshot showing routing and identification of attackers
Figure 4. Snapshot of alternate routing
ISSN: 1694-2108 | Vol. 1, No. 1. MAY 2013 10

IJCSBI.ORG
The values obtained using traditional AODV and proposed EN-AODV at

different node sizes are listed in table 3. The traditional AODV doesnt
provide reliable routing since the nodes energy is not checked which may
result in inconsistency.
Table 3. Result comparison with different node sizes
Node Traditional AODV Proposed EN-AODV

PDR Delay PDR Delay
Size
25 64.32 0.33567 86.18 0.18567
50 72.56 0.22496 92.93 0.12404
100 74.73 0.18624 88.75 0.13993
The QoS parameter values are showing better improvement when the
routing takes place with the proposed EN-AODV protocol which works
using energy levels of each node that identifies nodes with low energy
levels in the route and immediately take an alternate path to provide reliable
routing. The results shown in the above table clearly shows the PDR and
delay of the proposed EN-AODV protocol is superior compared to
traditional AODV protocol at different node sizes.
Figure 5 specifies the increase in PDR by implementing the proposed

energy based EN-AODV protocol compared to the traditional AODV
protocol. Figure 6 specifies the decrease in delay while using the proposed
energy based EN-AODV compared to traditional AODV. Figure 7 specifies
the energy consumed by each node in the MANET during the transmission.
Figure 5. Comparison of general AODV PDR and EN-AODV PDR
ISSN: 1694-2108 | Vol. 1, No. 1. MAY 2013 11

IJCSBI.ORG
Figure 6. Comparison of general AODV Delay and EN-AODV Delay
Energy consumption by nodes

100
90
80
Energy Level
70
60
50
40
30
20
10
0
Nodes
Figure 7. Comparison of energy consumed by nodes in MANETs
5. CONCLUSIONS
In this paper, an energy based EN-AODV protocol is proposed that
identifies the nodes that drain out of energy level during data transmission.
Energy value for each node is calculated to spot the unreliable nodes in the
path during routing. A node which has sufficient energy level for the
transmission is selected for routing. MD5 algorithm in used to make
transmissions secured which add more reliability and also identifies
attackers and eliminates them. This proposed scheme has shown a good
development over QoS parameters like PDR and delay and has also
provided reliable routing. The same scheme can also be implemented on
other MANET routing protocols and check the performance with respect to
QoS parameters. The future work may provide an encryption scheme for
secured packet transmission and also to provide virtual energy for source
nodes participating in the routing to enhance reliability in MANET routing.
ISSN: 1694-2108 | Vol. 1, No. 1. MAY 2013 12

IJCSBI.ORG
6. REFERENCES
[1] Kortuem.G., Schneider. J., Preuitt. D, Thompson .T.G.C, Fickas. S. Segall. Z.
(2001), When Peer to-Peer comes Face-to-Face: Collaborative Peer-to-Peer
Computing in Mobile Ad hoc Networks. 1st International Conference on Peer-to-
Peer Computing, August, Inkoping, Sweden, 75-91.
[2] C. Perkins, E. Royer and S. Das, Ad hoc on-demand Distance Vector Routing,
RFC-3651.
[3] Hu, Y., (2003), Enabling Secure High-Performance Wireless Ad Hoc Networking,
PhD Thesis, Carnegie Mellon University (CMU).
[4] IIyas M., (2003), The Handbook Of Wireless Ad Hoc Network, CRC
[5] Jin-Man Kim, Jong-Wook Jang (2006), International Conference on internet and
web Applications and services.
[6] Jun-Hu Zhang, Hui Peng, Feng-Jing Shao (2011), Eighth International
Conference on Fuzzy Systems and Knowledge Discovery (FSKD) Volume: 4,
2275 2280
[7] Muthumayil, K., Rajamani, V., Manikandan,S (2011), Third International
Conference on Advanced Computing (ICoAC), 276 281.
[8] Li Layuan, Li Chunlin, Yuan Peiyan (2006), International Conference on
Intelligent agent technology, IEEE/WIC/ACM, 306 312
[9] Romdhani, L, Bonnet, C (2007), 9th IFIP International Conference on Mobile
Wireless Communications Networks, 96 100.
[10] Kathiravan, K., Divya, V., Selvi, S.T (2009), 5th International Conference on
Mobile Ad-hoc and Sensor Networks, 340 346.
[11] Wu, Zhengyu, Song, Hantao, Jiang, Shaofeng, Xu, Xiaomei (2007), 1st Asia
International Conference on Modelling and Simulation, 36 41.
[12] Xinsheng Wang, Qing Liu, Nan Xu (2008), Fourth International Conference on
Natural Computation, Volume: 5, 276 280.
ISSN: 1694-2108 | Vol. 1, No. 1. MAY 2013 13

Vol 2 No 1 - June 2013

Încărcat de

Informații document

Drepturi de autor

Formate disponibile

Partajați acest document

Partajați sau inserați document

Opțiuni de partajare

Vi se pare util acest document?

Este necorespunzător acest conținut?

Drepturi de autor:

Formate disponibile

Vol 2 No 1 - June 2013

Încărcat de

Drepturi de autor:

Formate disponibile

International Journal of Computer Science

and Business Informatics

ISSN: 1694-2507 (Print)

Unstructured Data Integration through Automata-Driven Information Extraction ............................... 1

Upcoming Trends of Virtual Experiments for Laboratories................................................................... 1

Negotiation Based Resource Allocation for Distributed Environments ................................................. 1

Digital Image Tamper Detection Techniques - A Comprehensive Study ............................................... 1

A New Online XML Document Clustering Based on XCLS++ .................................................................. 1

Analytical Hierarchy Process Based

ISSN: 1694-2108 | Vol. 2, No. 1. JUNE 2013 1

Prioritizing requirements is one of the tasks of the Requirements Phase. It

Analytical Hierarchy Process (AHP) is used to estimate the priorities among

ISSN: 1694-2108 | Vol. 2, No. 1. JUNE 2013 2

A goal model is effective in concisely capturing large number of alternative

ISSN: 1694-2108 | Vol. 2, No. 1. JUNE 2013 3

In modeling mandatory requirements, only tasks and goals are considered,

The basic concerns that are involved in the stakeholders communication

ISSN: 1694-2108 | Vol. 2, No. 1. JUNE 2013 4

ISSN: 1694-2108 | Vol. 2, No. 1. JUNE 2013 5

Figure 1 Goal model

3.2 Identification of Goal Preferences and Priorities

ISSN: 1694-2108 | Vol. 2, No. 1. JUNE 2013 6

3.3 AHP Method

ISSN: 1694-2108 | Vol. 2, No. 1. JUNE 2013 7

3.4 Determination of Alternative Solution

ISSN: 1694-2108 | Vol. 2, No. 1. JUNE 2013 8

Table 4. Relative intensity table

7 Very strong value A goal is strongly favoured and its dominance is

9 Extreme value The evidence favouring one over another is of the

ISSN: 1694-2108 | Vol. 2, No. 1. JUNE 2013 9

4. EXPERIMENT AND RESULTS

Figure 2. Task priority graph

ISSN: 1694-2108 | Vol. 2, No. 1. JUNE 2013 10

ISSN: 1694-2108 | Vol. 2, No. 1. JUNE 2013 11

Unstructured Data Integration

ISSN: 1694-2108 | Vol. 2, No. 1. JUNE 2013 1

ISSN: 1694-2108 | Vol. 2, No. 1. JUNE 2013 2

ISSN: 1694-2108 | Vol. 2, No. 1. JUNE 2013 3

ISSN: 1694-2108 | Vol. 2, No. 1. JUNE 2013 4

ISSN: 1694-2108 | Vol. 2, No. 1. JUNE 2013 5

Decrees the following:

Article 1: To give the Ministry of Finance treasury advance worth / 500.000.000.000 /

Baabda, the 18th of November 2011

Table 1. Relevant Information

Preceding Wanted Entity

Insert a new record in

ISSN: 1694-2108 | Vol. 2, No. 1. JUNE 2013 6

Preceding Wanted Entity

ISSN: 1694-2108 | Vol. 2, No. 1. JUNE 2013 7

Figure 2. Type Finite State Automaton

q0 No. q1 0-9 a-z|A-Z

Figure 3. Number Finite State Automaton

D= { Update(qi, 'decree', 'number'), Concatenate(qi, c) }

ISSN: 1694-2108 | Vol. 2, No. 1. JUNE 2013 8

Figure 4. Title Finite State Automaton

D = { Update(qi, 'decree', 'title'), Concatenate(qi, c) }

Based on And after

D = {Insert(qi, 'basedon', 'value'), Concatenate(qi, c),Clear(qi)}

ISSN: 1694-2108 | Vol. 2, No. 1. JUNE 2013 9

D = {Insert(qi, 'andafter', 'value'),Concatenate(qi, c), Clear(qi)}

ISSN: 1694-2108 | Vol. 2, No. 1. JUNE 2013 10

Figure 8. Location Finite State Automaton

D = { Update(qi, 'decree', 'location'), Concatenate(qi, c) }

ISSN: 1694-2108 | Vol. 2, No. 1. JUNE 2013 11