Sunteți pe pagina 1din 5

Research and Analysis on Routing Protocols for

Wireless Se11sar Networks


"

Shi i n Dai , Xi aorong

Jing, Le mi n Li

Key laboratory of Broadband Optical Fiber Transmission and Communication Networks


University of Electronic Science and Technology of China
Chengdu 610054, Sichuan, P.R.China

. E-MAIL*:daiphd2002@msn.com
Abstract-Wireless
products

which

sensor

networks

integrate

sensor

(WSNs)
techniques,

are

the

em be dde d

techniques,
distributed
information
and
processing
communication tec hni ques. The appearance of the wireless

sensor network is a revolution in information sensing and


detection. Recently, both academia and industries show great

interest in sensor networks. In this paper, the architecture of


the wireless sensor network and the WSNs features are
introduced.

And

then

this

paper

studies

pro toco ls for wireless sensor networks and

recent

presents

routing

a variety

of classification of them, as well as contrasts and compares the

representative routing protocols. At last, several future open


issues ofthe wireless sensor networks are put fonY[lrd.

I.
the

With

networks not only can increase the efficiency of rescue


operations but also ensure the safety of the rescue crew. On
the military side, applications of sensor networks are
numerous too. For instance, the use of networked a set of
s ensrs can limit the need for personnel involvement in the
usually dangerous reconnaissance mission.

development

of

by making them remotely controllable and target-specific in

applications of sensor networks include intrusion detection


and criminal huniing[2]. These features have motivated

intensive research in the wireless sensor networks.


Sensor nodes are constrained in energy supply and

bandwidth. Combined with a typical deployment of large


the

processor,

number of sensors have posed many challe ng es to th e design

MEMS

and management of the wireless sensor networks. Efficient


routing protocol on wireless sensor networks is one of the

radio and

memory technologies, it ' s possible to produce micro sensor

imp ortant challenges. Although there are some previous

nodes. Being characterized by their low-power, small size,

and

cheap

efforts on surveying the characteristics, applications, and

price, these nodes are capable of wireless

communication protocols in WSNs, the scope of this article

communication, sensing and computation. So, we can say

the sensor network is the

addition,

order to prevent hanning civilians and 'animals. Security

INTRODUCTION

(micro-e1ectro-mechanical-systems),

In

sensor networks can enable a more civic use of landmines

is distinguished from these surveys in many a sp ects In this

product of the combination of the

techniques,
distributed
infonnation processing and communication technique s .

paper, we present a thorough review of recent research of

A w i reless sensor network is composed by hundreds or


thousands of nodes that are densely deployed in' a large

attention on making a contrast between these protocols and

sensor

techniques,

embedded

routing protocols for

their

II.

transform these data into electric signals which can be

THE ARCHITECTURE OF THE PROTOCOL STACK

FOR

SENSOR NETWORKS

through the wireless medium, and promotes cooperative


efforts of sensor nodes. The protocol stack are made up of
physical layer, data link layer, network layer, transport layer,

high-security

fact, the applications of the wireless sensor networks are

application

quite numerous. For example, wireless sensor networks have


profound effects on military and civil application. In a

layer,

management plane,

power

and

management

task

plane,

management

mobility

plane.

The

physical layer addresses the needs of a robust modulation,


transmission and receiving techniques. Because the

great number of sensors can be dropped by

a helicopter. The wireless sensor network consisted of these

environment is noisy and sensor nodes can be mobile, the

nodes can assist rescue operations by l ocating survivors,

MAC

identifYing risky areas and making the rescue crew more

protocol must be power aware and able to minimize

collision with neighbors' broadcast. The network layer takes

aware of the overall situation. Such application of sensor

and RGC ofHK (60218002),

o7803-9015-6/05/$20.00 2005 IEEE.

our

with networking protocols, communicates power efficiently

control and guidance in automati c

This paper is supported by.NSFC

focus

stack combines power and routing awareness, in tegra te s data

smart homes, and identification and personalization[1]. In

we

wireless sensor network is shown as Fig.l. This protocol

Other applications may be environment control in such as

disaster scene,

Then

The architecture' of the protocol stack used in the

located in the area arourid these sensors. Therefore, we can


get the information about the area which is far away from us.

manufacturing environments, interactive toys,

sensor networks, includi ng

drawbacks.

WIRELESS

p rocessed to reveal some characteristics about phenomena

robot

wirel e ss

and

highlight some future aspects of research.

ge og raphi cal area. These sensors


measure ambient
c on ditions in the environment surrounding them and then

office buildings,

advantages

407

care of routing the data supplied by the transport layer. The


transport layer helps to maintain the flow of the data if the
wireless sensor networks application requires it. Different
types of application software can be set up and used on the

sensing

task

and lower

Unlike a node in a MANET, sensor node may not

application layer depending on the different sensing tasks. In

have a unique ID.

addition, the power, mobility, and task management planes


monitor the power, movement, and task distribution among
the sensor nodes. These planes help the sensor nodes
coordinate the
consumption.

The number of sensor nodes in w i re less sensor


networks can be several orders of magnitude higher
than that in MANETs.

Sensor nodes are much cheaper than nodes in a

MANET and are usually deployed in thousands.

overall power

Power resource of sensor nodes could be very


limited; however, MANET's nodes can be
re-changed.

The power management plane manages how a sensor


node uses its power. For instance, the sensor node can shut
down its receiver after receiving the data from one of its

Sensor nodes are more limited in their computation

and communication capabilities than their MANET

neighbors. It is to avoid getting duplicated messages. Also,


when the power level of the sensor node is low, the sensor
node broadcasts to its neighbors that it is low in power and

counterparts because oftheir low cost.

can't take part in transmitting messages. The remaining


power is reserved for sensing. The mobility management
plane detects and registers the mobility of sensor nodes, so a
route back to the user is always kept, and the nodes can keep
track of who their neighbor sensor nodes are. Therefore, the

Sensor nodes are prone to failures.

The topology of a sensor network changes very


frequently.

Sensor
nodes
mainly
use
a
broadcast
communication paradigm, whereas most Ad Hoc

nodes can balance their power and task usage by knowing


this situation. The task management plane balances and
schedules the sensing tasks given to specific area. Not all of

networks

the sensor nodes in that region are required to perform the

./"

AppJir.nlitm Layer

TraJ'upon {.nyc,.

....

,/

1)aw Ljllk U"",r

for the problem of routing data in sensor networks. We


classify the routing protocols for sensor networks first, and
then analyze the existing routing protocols.

/'
e'

A.

.'

Ph)in,ll...ay.c-r

""

Almost all of the routing protocols can be classified as flat,


hierarchical or location-based, according to the network
structure. Furthermore, these protocols can also be classified
into
multipath-based,
query-based,
negotiation-based,

'
"V
V

quality ofservice (QoS)-based, or coherent-based depending


on the protocol operation. In flat networks all nodes play the
same role, while hierarchical protocols aim at routing

Fig. l. The wirelcss sensor nctworks protocol


sUlck
III.

Classification ofRouting Protocols


There are many ways to classifY the routing protocols.

I
ES I

point-to-point

of energy-efficient, low-cost, secure and fault-tolerant sensor


networks. Now, many new algorithms have been proposed

Nm\l/r)'-/'; 1..n)"Y

on

wireless sensor networks. Both academia and industries


have shown great interest in the wireless sensor networks
and have focused on the issues involved in the development

management planes are need so that sensor nodes can work


together in an energy-efficient way, route data in a mobile
wireless sensor network, and share with the resources
between nodes.

__n'_/

based

So it is important to study new routing protocols for

sensing task at the same time. So, some nodes perfonn the
task more than others depending on their power level. These

/"

are

communications.

techniques clustering the nodes so that cluster heads can do


some aggregation and reduction of data in order to save
energy. Location-based protocols utilize the position
information to relay the data to the desired regions rather

ROUTING PROTOCOLS

Among the existing network models, the mobile Ad Hoc


Networks (MANETs) are the closest to sensor networks.

than the whole network[4].


B.

MANETs and wireless sensor networks have many similar


characteristics. For example, network topology is not fixed;

Routing protocols

/) Flooding is an old technique that can also be used in


the sensor network. In flooding, each node received a data

power is an expensive resource; nodes in the network are


connected to each other by wireless communication links.
And yet, the protocols and techniques developed for
MANETs can't be directly applied to wireless sensor

and then sent them to the neighbors by broadcasting, unless


a maximum number of hops for the packet are reached or the
destination of the packet is arrived. However, it has several
deficiencies such as[5]:

networks because there are many difference between these


two types of networks. The two networks vary in the

following respects[3].

Implosion: Implosion is a situation where duplicated


data are sentto the same node. For example, ifnode

408

Directed diffusion[7] is an important milestone in the

A has N neighb or nodes which are also the


neighbors of the node B, node B w i ll receive N

data-centric routing p rotoc ol s research in the wireless sensor

ne tworks. The algorithm aims at diffusing data through


sensor nodes by using a naming scheme for the data. The

copies of the message sent from node A.

Overlap: If two nodes share the same measuring


region, both of them may sense the same data at the
same time. As a result, neighb or nodes receive
duplicated messages.

main reason is to get rid of unnecessary operation of


network layer routing in order to save energy In this
protocol, the sink sends out interest, which is a task
description, to all of the sensors. The task descriptors are
named by assigning attribute-value pairs. Each node then
stores the interest entry in its cache. The interest contains a
.

Reso urc e blindness: The flooding protocol does not


take i nto account the available energy resources. An
energy resource aw are protocol must take into

timestamp field and several gradien t fields. As the interest is

ac count the amount of energy available to them at


all times.

propagated througho ut the sensor network, the gradients


from the source back to the sink are established. When the

G ossip ing protocol is the derivation of flooding. In this


al gorithm, nodes do not use b r oadcast but send the incoming
packets to a randomly selected neighbor nod e. Once the

data alon g the interest's gradient path. The interest and data

source has data for the interest, the source node sends the
propagation and aggrega tio n are determined locally. Also,
the sink must refresh and reinforce the interest when it starts
to receive data ITom the source. Since Directed diffusion
networks are application aware, they can achieve energy
saving by s electin g good paths by c aching and processing
data in-networks. Although the protocol has advantage of
saving energy, it also has prob lems . For example, to

neighbor node receives the data, it randomly selects another


sensor node. It can avoid the implosion problem, but the cost

is the lo ng time p rop agation for sending messages to all


sen sor nodes.
2)

SPIN

(Sensor

Protocols

for

Informatio n

via

implement
data
aggregation,
it
employs
time
synchronization techniques, which is not easy to realize in a
wireless sensor network. One other problem in data

Negotiation)[6] is a mong the early work to p ursu e a


data-centric routing mechanism . The idea behind SPIN is to
name the data using meta-data that highly describes the

characteristics of the data, which

aggregation is overhead involved in recording information.


All of these may lead to incre a s ing the cost ofa sen so r node.

is the key feature of SPIN.

SPIN has three types of messages, that is, ADV, REQ, and
DATA.

3)
LEACH
(Low
Energy
Adaptive
C luste ri ng
Hierarchy)[8] is a clustering-based protocol that utilizes
randomized rotation of the cluster-he a ds to evenly distribute
the energy load among the sensor nodes in the network. It is
one of the most popular hierarchical routing algorithms for
sensor networks. The idea is to form clusters of the sensor
nodes. based on the received signal strength and use local
cluster heads as router to the sink. This w ill save energy
since the transmission wil l only be done by cluster heads
rather than all of the nodes. All the data processing such as
propagation and aggregation are local to the cl u ster. Cluster
heads hange randomly over time in order to balance the
energy' di ssipation ofthe nodes. This decision is made by the
node choosing a random number between a and 1. The node
becomes a cluster head for the current round if the number is

ADV-When a node has data to send, it advertises


this message containing m eta-data .

REQ-A node sends this message when


it wishes to
.
receive some data.

DATA - Data message

contains

the data with a

meta-data header.
Before se ndi ng a DATA message, the s ensor node
broadcasts an ADV message c onta ining a descriptor (i.e.
meta-data) of the DATA. If a neigh bo r is interested in the
data, it sends a REQ m ess age for the DATA, and then
DATA is sent to this neighbor node. Respectively, the
neighbor node repeats the same process untilthe data i s sent

less than the following threshold:

to the sink (or BS). SPIN's meta-data negotiatio n and


resource ad ap t i ve solves the cl assic proble ms of flooding
such as im plosion , overlap and resource blindness, achieving
a lot of energy efficiency. The semantics of the me1a*data
format is application-based and not sp ecified in SPIN. One
of the advantages of SPIN is that topological changes are

T(n)=

(P(fomOd;) i/neG
1-/

others

localized since each node needs to know only its single-hop


neighbors. However, SPIN's disadv ant a ges are clear. First of
all, it is not scalable. Secondly, the nodes around a sink

where p .is the desired percentage of cluster head s (e.g.


is the current round, and G is the set of nodes that

0.05), r

have not be en cluster heads in the last lip rounds. LEACH is


organized into rounds, where e ach of them begins with a
set-up phase, and is fo llowed by a steady-state phase.
Usually, the latter phase is longer than the former phase. fn
cluster set-up phase. each non-cluster-head node tells its
cluster-head its decision by using CSMA MAC protocol.

could deplete their energy if t he sink is interested in too


many events . Finally, SPIN's data advertisement mechanism
can't guarantee the delivery of data. For example, if the
nodes that are interested in the data are far away from the
source node and the nodes between source and destination
are not interested in that data, such data will not be transport

Then

to the destination at all.

409

the cluster-heads

create

TDMA

sched ules

and

broadcast them back to their members in schedule creation


phase. In data transmission phase, each node waits for its
turn to send data if needed.

sensor networks, discussed above, with respect to a few


metrics we identified. Table I giv es the comparison.
Based on the analysis of the above protocols, we believe
that a good routing protocol for wireless sensor networks
should have some desirable features, such as:

LEACH provided many good features to the sensor

network ,

such
as
clustering
architecture, .localized
coordination and randomized rotation of cluster-heads;
however, it suffers from the following problems:

'
It can not be app lied to tirl!e critical applications

extends the network's lifetime.

The nodes on the route from a hot spot to the sink


might drain their energy quickly, which is known as

PEGASIS

(Power-Efficient

Systems)[9]

is

Gathering

in

helps in efficient query processing, and decreases


network overhead dramatically. Hence saves energy.

Sensor

chain-based power efficient

protocol based on LEACH. Because each node has global


knowledge of the network, the chain can be constructed
easily by using a greedy algorithm. PEGAS]S outperfonns
LEACH by eliminating the overhead of dynamic cluster
formation, minimizing the sum of distances that non leader

load and tolerating the failure of nodes.

must

transmit, and limiting the number of


transmissions. Ho wever . PEGAS]S .has the same problems
as LEACH does. Furthermore, it requjres global information
,

knowledge is not easy to get.

Thresholds f o sensor nodes to transfer sensed data.

Chosen good thresholds, it may solve "hot spot"


problem and save energy by limiting unnecessary
transmissions. It will be helpful to extend the
.
lifetime of the sensor

of the network known by each noe. It does not scale well

and is not suitable for sensor networks where such global

network.

Thresholds for sensor nodes to relay data.


Determining appropriate thresholds of energy and

time delay to relay data would help in elongating

4) GEAR (Geographic and Energy Aware Routing)[IO]


uses energy aware and geographically informed neighbor
selection heuristics to route a packet .towards the target

nodes' Iifetime[l I].

TABLE!.

region. ]n GEAR, each node keeps an estimated cost and a


learning cost of reaching the destination through its

neighbors. The estimated cost is a combination of residl)al


energy. and distance to destination. The learned cost .is a
refinement of the estimated cost that accounts for routing
around holes in the network. A hole occurs when a node
d oes not have any closer neighbor to the target region than
itself.,Ifthere are no holes, the estimated cost is equal to the
learned cost. The learned cost is propagated one hop back

Scalabilily

limiled

Limiled

Good

No
Long

Long
Ye,

No

Mola""'''

No

Yo,

No

Y..

Power
required

High

LimiLed

High

No

No

No

Yo>

No

No

No

No

y",

Yo.

No

Yo,

Fbi

Dilla-ccnlric

Hierarchical

Locil.Li()nb

multihop

Cb,r.ti

energy consumption for the route set up, but also performs
better than GPRS in terms of packet delivery.

GEAR

No

roolC

LEACH

Long

Optimal

protocols in geographic routing, GEAR not only reduces

SPIN

No

...wan::ss

target region; the other is the forwarding the packets within


the region Compared to GPSR, which is one of the earlier

Flooding

Shon

Location

in the. a lgorithm : One is the forwarding packets towards the

COMPARISON OF VARIOUS ROUTING PROTOCOLS

Datil
diffusion

Lifetime

every time a packet reaches the destination so. that route


setup for next packet will be adjusted. There are two phases

IV.

Randomizing path choice. If a routing algorithm can

. support m l ti ple paths to a destination with low

overhead, it could help in balanCing the network

nodes

Data aggregation. If the data classification and

fusion can be completed quickly in sensor nodes, it

"hot spot" problem.


Infonnation

Dynamic clustering architecture. It prevents cluster


heads from depleting their power quickly, and hence

on

V.

CONTRAST ANDCOMPARlSON

Wireless

sensor

, Lirnil;::d

CONCLUSION

networ s have become popula r due to

the progress made i sensing, communication, and


computing areas. To make wireless sensor networks more
practical, we need to develop effective routing protoco ls for
them that meet several unique requirements. and constraints.

Routing research in wireless sensor networks has


attracted a l ot of attention in these years and brought unique
challenges compared to traditional data routing in wired
networks. From the statement above, we can make
conclu sion every protocol has relationship with others, such
as LWACH, PEGASIS has si m il a r features designed with
the same idea, although many new features are added to the

In this paper, we have enumerated some research results on


data routi ng protocols in wireless sensor networks and
classified the appoaches. Then w make' a contrast between

these algorithms. Although extensive efforts have been


exerted so far on the routing problem in wireless sensor
networks; there are still some challenges that confront
.

later one. So, it is hard to say this protocol is better than

a noth er one because sensor networks are application specific.

Now, we compare and contrast the routing protocols for

410

effective solutions to the routing problem. As our study


reveals, it is not possible to design a routing algorithm which
will have good performance under all scenarios and for all
applications.
The further research would be needed to address issues
such as Quality of Service (QoS) posed by video and
imaging sensors and real-time applications. Currently, there
is very little research that looks at handling QoS

International Conference on Mobile Computing


(MobiCom '00), Boston, MA, 2000.

[8]

[9]
-

ACKNOWLEDGMENT

This work is supported by NSFC and RGC of HK


(60218002).
REFERENCES
R. C, Shah and

Wireless Sensor Networks," Elsevier Ad Hoc Network Journal, vol.

3,

[3]

L F. Akyildiz, W. Su, Y. Sankarasubramaniam, and E, Cayirci, "A


Survey on Sensor Networks," in IEEE Communications Magazine,
vol. 40,2002, pp. 102-114.

[4]

J. N. AIKaraki and A. E. Kamal, "Routing techniques in wireless

[5]

], Kulik, W. Rabiner, and H. Balakrishnan, "Adaptive Protocols for

[6]

1.

sensor networks: a survey," Wireless Communications, IEEE [see


also IEEE Personal Communications], vol.

II, pp. 6 - 28,2004.

Information Dissemination in Wireless Sensor Networks," presented


at Proceedings of the 5' Annual ACMfIEEE Conference on Mobile
Computing and Networking (MobiCom '99), Sealtle, WA, 1999.
Kulik,

W,

"Negotiation-based

R.

Heinzelman,

protocols

for

and

H.

disseminating

wireless sensor networks," Wireless Networks, vol.


2002.
[7J

Balakrishnan,

information

2002.

Qiangfeng and D. Manivannan, "Routing protocols for sensor


networks," presented at Consumer Communications and Networking

1.

Conference, CCNC 2004. First IEEE, 2004.

wired networks (i.e., the Internet). Most 'of the applications


in security and environmental monitoring require the data
collected from sensor nodes to be transmitted to a server so
that further analysis can. be done. On the other hand, the
requests from the user should be made to the BS through the
Internet. Since the routing requirements of each environment
are different, further research is necssary for handling these
kinds of situations[2].

pp. 325349,2005.

of the IEEE Aerospace Conference, Montana,

UCLA-CSD TR -01-0023, May 2001.

and propagation of that information through the network


may excessively drain the battery of nodes. New routing
algorithms are needed to handle the overhead of mobility
and topology changes in such an energy-constrained
environment. Other possible future research for routing
protocols includes the integration of sensor networks with

K. Akkaya and M. Younis, "A Survey on Routing Protocols for

S. Lindsey and C. S. Raghavendra, "PEGASIS: Power-Efficient


Gathering in Sensor Information Systems," presented at Proceedings

[10] Y.Yu, D.Estrin, and R.Govindan, "Geographical and Energy-Aware

[II]

possibly need to be mobile. In such cases, frequently update


the information of the position of the sink and sensor nodes

[2]

W. R, Heinzelman, A. Chandrakasan, and H. Balakrishnan,


'Energy-Efficient Communication Protocol for Wireless Microsensor

Routing: A Recursive Data Dissemination Protocol for Wireless


Sensor Networks,"
UCLA
Computer
Science
Department

protocols is the consideration of the node mobility. Most of


the current research assume that the node and the sink are
both stationary. However, there might be some situation
such as in hospital scenario where the nodes and the sink

[I]

Ner.orkmg

Networks," presented at Proceedings of the 33' Hawaii International


Conference on System Sciences, Hawaii, 2000.

requirements in a very energy constrained environment like


sensor networks. Another interesting issue for routing

J. M. Rabaey, "Energy Aware Routing for Low


Energy Ad Hoc Sensor NetworkS," presented at Proc. IEEE Wireless
Communications and Networking Conference (WCNC), Orlando. FL.
2002

and

in

8, pp> 169-185,

C Intanagonwiwat. R. Govindan, and D. Estrin, "Directed Diffusion:

A Scalable and Robust Communication Paradigm for Sensor


Networks," presented at Proceedings of the 6' Annual ACMfIEEE

411

S-ar putea să vă placă și