54

JOURNAL OF COMMUNICATIONS AND NETWORKS, VOL. 15, NO. 1, FEBRUARY 2013

A Tracking System Using Location Prediction and Dynamic Threshold for Minimizing SMS Delivery
Yuan-Cheng Lai, Jian-Wei Lin, Yi-Hsuan Yeh, Ching-Neng Lai, and Hui-Chuan Weng
Abstract: In this paper, a novel method called location-based delivery (LBD), which combines the short message service (SMS) and global position system (GPS), is proposed, and further, a realistic system for tracking a targets movement is developed. LBD reduces the number of short message transmissions while maintaining the location tracking accuracy within the acceptable range. The proposed approach, LBD, consists of three primary features: Short message format, location prediction, and dynamic threshold. The dened short message format is proprietary. Location prediction is performed by using the current location, moving speed, and bearing of the target to predict its next location. When the distance between the predicted location and the actual location exceeds a certain threshold, the target transmits a short message to the tracker to update its current location. The threshold is dynamically adjusted to maintain the location tracking accuracy and the number of short messages on the basis of the moving speed of the target. The experimental results show that LBD, indeed, outperforms other methods because it satisfactorily maintains the location tracking accuracy with relatively fewer messages. Index Terms: Global positioning system (GPS), location tracking, mobile phones, prediction algorithms, short message service (SMS).

I. INTRODUCTION The global position system (GPS) has become a common functionality in handheld devices, and therefore, several location-tracking applications have been developed [1][14], including continuous location-tracking of elders and children for safety reasons or to prevent them from being lost [1], [2], car monitoring and tracking [3][5], and intelligent transportation systems [6]. The GPS is used to obtain the location information of a target (e.g., a mobile device). However, most of the above-cited works used either an 802.11 wireless network or the short message service (SMS) to transmit the location information of a target to a tracker. For example, Lee et al. proposed a real-time
Manuscript received February 29, 2012; approved for publication by Xiaodai Dong, Division I Editor, November 8, 2012. J.-W. Lin is the corresponding author. Y.-C. Lai is with the Department of Information Management, National Taiwan University of Science and Technology, Taipei, Taiwan, email: laiyc@cs. ntust.edu.tw. J.-W. Lin is with the Department of International Business, Chien Hsin University, Taoyuan, Taiwan, email: jwlin@uch.edu.tw. Y.-H. Yeh is with the Department of Information Management, Oriental Institute of Technology, New Taipei, Taiwan, email: amyyeh@mail.oit.edu.tw. C.-N. Lai is with the Department of Information Technology, Hsing Wu University, New Taipei, Taiwan, email: sebastian.lai@gmail.com. H.-C. Weng is with the Department of Information Management, National Taiwan University of Science and Technology, Taipei, Taiwan, email: M9809107@mail.ntust.edu.tw. Digital Object Identier 10.1109/JCN.2013.000010

location tracking system [1] for childcare or elderly care applications. It transmits the location information of the mobile device to a central GPS application server through the 802.11 wireless networks. This application allows the server to simultaneously monitor multiple targets (e.g., elders or children). Further, Choi et al. assumed that the location information of a target is transmitted through wireless networks. Their work focused on proposing a geolocation update scheme to decrease the update frequency [7]. Lita et al. proposed an automobile localization system by using SMS [3]. The proposed system, which is interconnected with the car alarm system, transmits alerts to the owners mobile phone in the event of a car theft (e.g., activation of the car alarm, starting of the engine) or provides information for monitoring adolescent drivers (e.g., exceeding the speed limit or leaving a specic area). Hameed et al. proposed a car monitoring and tracking system that uses both SMS and GPS to prevent car theft [5]. Anderson et al. proposed a transportation information system [6]. In this system, a hardware device called StarBox, which is equipped with a global system for mobile communications (GSM) modem and a GPS unit, is installed in a vehicle to track the vehicles location. StarBox transmits short messages containing its GPS coordinates to the server at 30-s intervals. The users can send short messages to the server to determine the expected arrival time of buses at their locations. Although transmitting the geolocation information of a target via wireless networks is effective when both the target and the tracker are within Wi-Fi coverage area, the 802.11 wireless networks are not always accessible. When the target or the tracker is unable to access Wi-Fi, it is impossible to perform location tracking. Therefore, SMS is a relatively more reliable and exible solution because of its widespread use (i.e., well-structured worldwide) [6], [8]. However, SMS is a user-pay service. The objective of this study is to minimize the transmission cost of a tracking system by minimizing the number of SMS transmissions while maintaining the location tracking accuracy. In this paper, a novel method called location-based delivery (LBD), which combines SMS and GPS, is proposed, and further, a realistic system to perform precise location tracking is developed. LBD mainly applies the following two proposed techniques: Location prediction and dynamic threshold. Location prediction is performed by using the current location, moving speed, and bearing of the target to predict its next location. When the distance between the predicted location and the actual location exceeds a certain threshold, the target transmits a short message to the tracker to update its current location. The dynamic threshold maintains the location tracking accuracy and number of short messages on the basis of the moving speed of the target. The simulations performed to test the performance of LBD show that compared with other related works, the proposed LBD

1229-2370/13/$10.00 c 2013 KICS

LAI et al.: A TRACKING SYSTEM USING LOCATION PREDICTION AND DYNAMIC...

55

acters. Support for the GSM 7-bit alphabet is mandatory for all GSM handsets and network entities. SMS is the most widely used data application worldwide. The proposed system uses SMS to transmit location update messages and assumes that the message delay between the tracker and the target is negligible. B. Tracking Technique GPS is widely used for target location because of its high positioning accuracy [1][3], [6], [8]. Additionally, GPS networks are well-constructed and widely accessible. Therefore, GPS networks are used, practically, for positioning an object by using triangulation. Related works have roughly classied the location tracking methods [3], [5], [6], [8], [9] using GPS and SMS as timebased delivery and distance-based delivery. Time-based delivery is used to periodically transmit location update messages for tracking [3], [6], [10]. By contrast, distance-based delivery is used to transmit location update messages when the distance between the previously reported location and the current location exceeds a xed-distance threshold [9]. Time-based delivery is effective for tracking a target that is moving erratically. However, it exhibits a crucial aw when the target remains stationary for an extended period, that is, it continues to periodically transmit many unnecessary short messages. However, SMS is a userpay service. Therefore, the objective of this study is to minimize the number of SMS transmissions while simultaneously maintaining the location tracking accuracy. The proposed approach differs from the distance-based delivery in two major aspects. First, LBD transmits a location update message when the distance between the predicted location and the actual location exceeds a certain threshold, rather than when the distance between two continuously reporting locations exceeds the threshold, a typical mechanism adopted by distance-based delivery. Thus, LBD can deliver fewer update messages. Additionally, LBD further reduces the number of required update messages while satisfactorily maintaining the location tracking accuracy because it adopts a dynamic threshold rather than the static threshold used in distance-based delivery. III. PROPOSED APPROACH The three main features of the proposed LBD approach are a well-dened SMS format, location prediction module, and dynamic threshold module (see Fig. 2). LBD uses a proprietary SMS format. The location prediction module, which is built in both the target and the tracker side, uses the information on the current location, moving speed, and bearing of the target to predict its next location. The dynamic threshold module, which is used only on the target side, minimizes the number of short messages by dynamically adjusting the threshold TH according to the moving speed of the target. The tracker periodically updates the location of the target on the local screen according to the predicted location. However, when it receives a short message response from the target, it means that the predicted location is far from the actual location. For more accurate location tracking, the tracker updates the

Fig. 1. SMS architecture.

minimizes the number of short message transmissions while maintaining the location prediction accuracy within the acceptable range. The remainder of this paper is organized as follows. Section II discusses the technical background for LBD and reviews the relevant literature, including SMS and tracking technologies. Section III elaborates on the design of LBD, while Section IV describes the LBD system developed on the basis of Android phones and demonstrates some of the graphic user interfaces (GUIs). Section V describes the experiment and its results. Conclusions and future works of the authors are discussed in the nal section. II. BACKGROUND AND LITERATURE REVIEW A. Short Message Service SMS is a mobile communication service for exchanging text messages between mobile devices. It originated from radio telegraphy in radio memo pagers using standardized phone protocols. In 1985, it was included in the GSM series of standards as a means for transmitting messages. Fig. 1 shows a simplied diagram of the SMS architecture. The lower bars display the protocols used by the upper-layer network entities. A short message is transmitted from the mobile station (MS) to the GSM base station (BTS) through a wireless link and is received in the backbone network of the service provider. The mobile switch center (MSC), home location register (HLR), and visitor location register (VLR) determine the appropriate short message service center (SMSC), which processes the message by applying the store and forward mechanism. If the recipient is unreachable, the SMSC queues the message for a retry at a later time. Although the system cannot guarantee successful delivery of a message to its recipient, message delays and message loss rarely occur. The messages are transmitted between an SMSC and a handset by the mobile application part (MAP) of the SS7 protocol. This signaling protocol limits the message payload length to 140 bytes. Short messages are encoded using several alphabets, including the default GSM 7-bit alphabet, 8-bit data alphabet, and 16-bit UTF-16 alphabet. The alphabet selection depends on the conguration set in the subscriber handset. The selected alphabet limits the maximum size of the individual short messages to 160, 7-bit characters; 140, 8-bit characters; or 70, 16-bit char-

56

JOURNAL OF COMMUNICATIONS AND NETWORKS, VOL. 15, NO. 1, FEBRUARY 2013

Table 2. Example of SMS. Type Query(Start) Response Query(Stop) Fig. 2. Structure of the LBD system. Example TraSMS;Q;Start TraSMS;R; 53 11 18 N ; 000 08 00 E ; 0001000; 097 30 52 TraSMS;Q;Stop

Fig. 3. Concept of LBD. Table 1. Dened SMS format. (a) SMS query (total 13 or 14 bytes) P(6) ; T(1) ; A(4 or 5) Note: Numbers in parentheses are eld lengths. (b) SMS response (total 52 bytes) P(6) ; T(1) ; La(11) ; Lo(12) ; S(7) ; B(10)

of the tracking action, respectively. In the SMS response, the latitude and longitude elds contain the GPS latitude and longitude information of a target, respectively. The speed eld contains the target speed information in meters/hour. Finally, the bearing eld stores the bearing data obtained from the GPS. All the elds listed in Table 1 are of xed lengths according to their possible values in real time. For example, the latitude eld is 11 bytes long because it stores the latitude information, for example, 53 11 18 N . For example, tracking starts after the tracker transmits the TraSMS;Q;Start SMS query informing the target of the tracking action. When the target receives the SMS query, it transmits the SMS response when necessary (e.g., when the distance exceeds a certain threshold). The SMS response contains the current location information of the target, such as latitude, longitude, speed, and bearing. Table 2 shows all the message types and their examples. For an SMS query, the start and end tracking messages are 14 and 13 characters long, respectively. By contrast, an entire SMS response is 52 characters long. The dened message format can t into a single short message because the length of a short message is limited to 140, 8-bit characters. B. Location Prediction The target, upon receiving GPS information, including coordinates (longitude, latitude) = (X1 , Y1 ), moving speed V1 , and bearing B1 (in radians, clockwise from north), at time t1, can predict its next location at time t2. However, this prediction is not straightforward because the earth is not a plane. We refer to [15] to assume this prediction on the basis of a spherical earth, although the earth is slightly ellipsoidal. Thus, the predicted location information represented in terms of longitude and latitude 2 ) can be obtained as 2, Y (X 2 = arcsin(sin(Y1 ) cos(L) + cos(Y1 ) sin(L) cos(B1 )) Y 2 = X1 + arctan X sin(B1 ) sin(L) cos(Y1 ) cos(L) sin(Y1 ) sin(Y2 ) (1)

targets location using the information encoded in the received message, rather than its prediction. Fig. 3 shows the concept of LBD. The GPS coordinates (X1 , Y1 ), moving speed V1 , and moving bearing B1 of the target at time t1, are used to predict its next location at time t2, denoted 2 ). When the distance D between the predicted loca2, Y as (X 2, Y 2 ) and the actual location (X2 , Y2 ) exceeds threshold tion (X TH, the target transmits the actual location update information to the tracker. Thus, the objective is to determine an approach that minimizes the number of total short messages while maintaining D < T H . The followings subsections further discuss the three features in detail. A. Well-Dened SMS Format As summarized in Table 1, the designed messages are of two types: SMS query and SMS response. The SMS query includes three elds: Prex (P), type (T), and action (A), as shown in Table 1(a). The SMS response includes six elds: Prex (P), type (T), latitude (La), longitude (Lo), speed (S), and bearing (B), as listed in Table 1(b). The elds are separated by a semicolon. The prex eld containing the 6-character TraSMS indicates that the message is particularly used for location tracking. The two supported types are query (denoted as Q) and response (denoted as R). The SMS query includes two commands, Start and Stop, which refer to the starting and stopping

where L is the angular distance (in radians) and is expressed as L = d/R , where d is the distance traveled and R is the radius of the earth (6371.01 km). Thus we can easily obtain the value of L as L= V1 (t2 t1) . 6371.01 km (2)

2, Y 2 ) and the Distance D between the predicted location (X actual location (X2 , Y2 ) can be obtained by using haversine formula [15] as

LAI et al.: A TRACKING SYSTEM USING LOCATION PREDICTION AND DYNAMIC...

57

2 Y2 Y 2 ) sin2 + cos(Y2 ) cos(Y 2 P Q = 2 arctan 1P P = sin2 D = RQ = 6371.01 (km)Q.

2 X2 X 2

(3)

When the distance exceeds the threshold, i.e., D > T H , the target transmits a short message to the tracker to indicate its actual location, moving speed, and bearing. The tracker, upon receiving the message, updates the target location and renews the prediction process according to the information stored in the short message. However, if the distance is less than the threshold, the target does not send a short message to the tracker. The tracker continues to use its predicted location information to mark the target location because the difference between the actual location and the predicted location is within the acceptable range. C. Dynamic Threshold Threshold TH affects both the number of transmitted short messages and the location accuracy. A large threshold reduces the number of short messages as well as the location accuracy; that is, there is a large difference between the predicted location and the actual location. By contrast, a small threshold requires relatively an increased number of short messages; however, it increases the location tracking accuracy. We observed that the user tolerance of the perceived location accuracy mainly depends on the moving speed. When the speed of the target is high, that is, while driving, the tracker can tolerate a higher location tracking inaccuracy. On the other hand, when the speed of the target is low, that is, while walking, the tracker requires a higher target location tracking accuracy. Therefore, a feasible mechanism is to dynamically adjust the threshold according to the current moving speed of the target. This solution satisfactorily maintains the number of transmitted short messages and the location tracking accuracy. Thus, according to the moving speed of the target, the dynamic threshold can be set as T H = V C , where V is the current moving speed and C is a constant (e.g., 50 is used as the value of C in subsequent experiments). Fig. 3 depicts the application of the dynamic threshold concept. The gure clearly shows that TH is larger at t2 than at t3 because the moving speed of the target is higher at t1 than at t2. IV. DEVELOPED LBD SYSTEM A. System Architecture The developed LBD system mainly comprises the tracker and target sides and is built on the GPS and Android application framework, as shown in Fig. 4. The tracker side comprises the following components: SMS Switch, SMS Receiver, Predictor, and Ofine Pap. The operation procedure is as follows. The tracker rst uses the SMS switch to deliver the TraSMS;null;null;Q;Start Track; message to inform the target that it should start transmitting its loca-

Fig. 4. System architecture of location tracking system.

tion information as required to launch a tracking task. The Ofine Map periodically updates the target location information on the local tracker phone according to the predicted location information obtained from the Predictor. However, when the tracker receives a response message from the target, it means that the accuracy of the predicted location is too low. Therefore, the Ofine Map updates the target location information according to the received message rather than according to its prediction. Particularly, the messages from the target are received by the SMS Receiver on the tracker side. The SMS Receiver extracts the location information (e.g., coordinate, speed, and bearing) from the received message and passes it to the Ofine Map, which in turn displays and marks the target location on a map. After completing the tracking action, the tracker uses the SMS Switch to transmit the TraSMS;null;null;Q;End Track message to inform the target to stop transmitting its location information. Similarly, the target side comprises the following components: SMS Locator, SMS Receiver, Predictor, and Comparator. The operation procedure is as follows. When a start tracking message is received by the SMS Receiver, the target starts predicting its next location by using the Predictor, which uses the same algorithm as that used by the tracker to predict the current location. Thus, the predicted location information obtained from the tracker and that obtained from the target are the same. Further, the Comparator compares the predicted coordinates with the current actual coordinates obtained by the GPS to verify whether the distance between the two exceeds the threshold. When the distance exceeds the threshold, the SMS Locator transmits a response message to the tracker. The above procedure is performed iteratively until the track process is terminated. Fig. 5 shows the GUI of a mobile phone on the tracker side. The start tracking and stop tracking buttons, as shown in Fig. 5(a), are pressed to start and stop tracking a moving object, respectively. The map view button, as shown in Fig. 5(a), is pressed to display the tracking information of the target on an ofine map (Fig. 6). Fig. 5(b) shows the GUI of the inputted phone number of the target. B. Ofine Map Electronic maps are stored in the trackers mobile phone in advance to avoid a massive increase in the transmission cost for obtaining online maps. The SQLite (http://www.sqlite.org/) database is embedded to save the map information in the mo-

58

JOURNAL OF COMMUNICATIONS AND NETWORKS, VOL. 15, NO. 1, FEBRUARY 2013

Fig. 5. Graphic user interfaces on the tracker side: (a) Buttons for starting and stopping tracking and viewing map and (b) inputting the phone number of the target.

bile phone. SQLite is sufcient when the amount of processing data is small or when the data is only used by a single user. The SQLite database has many features in common with other public database systems, including the use of SQL-92 language syntax and atomic, consistent, isolated, and durable (ACID) transactions. The SQLite database is in the form of a single le, thereby avoiding a database system installation. Thus, it is quite simple to copy or create the database on a mobile device. A programming language is used to call the provided functions through the SQLite library. In this study, we use the SQLite database as a single le to store and retrieve the map information. Furthermore, a program, called RMap (http://robertdeveloper.blogspot.com/), is installed for displaying the ofine map on an Android mobile phone. The RMap uses the concept of overlays to draw the map and further marks the targets location on the map on the basis of the information retrieved from the received short messages or from the predictor. Fig. 6 illustrates the result of marking the targets location on the map, where the user can manually zoom in or zoom out for desired granularity. C. Implementation on Android The provided LocationManager class is used to access the GPS services on the Android system. The location manager service allows the upper-layer applications to obtain periodic updates of the geographical location of the device or to trigger an application-specic task when the device approaches the determined geographical location. Continuous monitoring and updating of an objects movement can be performed by using both LocationUpdates and LocationListener. Similarly, the provided SmsManager class is used to access SMS-related functions on an Android system, for example, by calling the sendTextMessage class to transmit a short message. Android adopts a permission-based policy, indicating that the permissions required by an upper-layer application should be specied in the AndroidManifest.xml le in advance. Thus, before using SmsManager, the two permissions SEND SMS and RECEIVE SMS must be added to that le.

Fig. 6. Ofine map GUI.

V. EVALUATION A. Comparison of Performances of Delivery Methods The three delivery methods, time-based delivery [3], [6], distance-based delivery [9], and proposed LBD, are evaluated using a given moving pattern of the target. In the time-based delivery, a short message is delivered every 30 s, as reported by Anderson [6]. In the distance-based delivery, the xed threshold is set to 50 m, as reported by Civilis [9], among other xed threshold values. The target is assumed to be moving at varying speeds and bearings or stationary for an unpredictable period to characterize a moving object. That is, the target used in this experiment is alternately moving and stationary. The experiment lasts for 3 h to obtain reliable results. The moving pattern is such that the target moves for 1 min and remains stationary for x min. The speed of the target ranges from 1 to 5 km/h during the 1 min movement period, which is generated by a uniform distribution. For example, when x = 1, the target rst moves at 3 km/h (randomly generated) for 1 min and remains stationary for 1 min. It again moves at 5 km/h (randomly generated) for 1 min, followed by remaining stationary for 1 min. The movements are repeated until all the experiments are completed. Fig. 7(a) shows the total number of short messages for timebased delivery, distance-based delivery, and LBD approach, denoted as TBD, DBD, and LBD, respectively. This gure shows that the number of short messages for TBD is xed because TBD short messages are periodically transmitted regardless of whether the target is moving or is stationary. The number of

LAI et al.: A TRACKING SYSTEM USING LOCATION PREDICTION AND DYNAMIC...

59

Fig. 7. Comparison of performances of three delivering methods: (a) number of short messages and (b) average location gap.

Fig. 8. Comparison of dynamic threshold and static threshold: (a) number of short messages and (b) average gap ratio.

short messages for DBD signicantly decreases as the stationary time increases. The reason is that DBD short messages are not transmitted when the target is stationary. Moreover, in LBD, fewer messages are transmitted as compared with DBD, because the former uses location prediction. However, the difference between the number of messages transmitted by using the two methods decreases when the stationary time is long, because neither LBD nor DBD short messages are transmitted when the target is stationary. Fig. 7(b) shows another metric, the average location gap, which is dened as the distance between the location displayed in the tracker map and the actual location. This gure clearly shows that the average location gaps for all the methods decrease as the stationary time increases. TBD obtained the smallest location gap, among all the three methods. In TBD, the target transmits the same location information to the tracker even when it is stationary; that is, the location gap between the location displayed in the tracker map and the actual location is zero. However, in this condition, the high accuracy of TBD for the tracker is meaningless. Meanwhile, LBD achieves a better location tracking accuracy than DBD. B. Effect of Dynamic Threshold The next experiment compares the dynamic-threshold LBD with the xed-threshold LBD. In this experiment, the same mean speed with varying coefcients of variation (CoV) is used to determine whether the dynamic threshold technique reduces the frequency of short message transmissions when the location tracking accuracy is within the acceptable range. Fig. 8(a) shows that the number of short messages correlates with the CoV of speed, regardless of whether the thresholds are xed or dynamic. The reason is that a high CoV of speed can increase the probability that the distance between the predicted location and the actual location exceeds a certain threshold. However, the dynamic threshold signicantly reduces the number of short messages delivered because the threshold is dynamically adjusted according to the varying speed of the target. Further experiments evaluated whether the average location gap remains within an acceptable range under varying CoV of speed. A new performance metric, average gap ratio, is required

because the accuracy tolerance is affected by speed of the target (e.g., an object moving at a lower speed requires more accurate location tracking than that moving at a higher speed). This ratio is dened as the distance between the location displayed on the tracker map and the actual location divided by the acceptable distance. Fig. 8(b) shows that the dynamic threshold has a better gap ratio than the xed threshold, and therefore, dynamic threshold, indeed, achieves a more acceptable location tracking accuracy than the xed threshold. Fig. 8 shows that, as the CoV of speed increases, the number of short messages increases and the average gap ratio decreases because the sudden changes in the moving speed of the target increase the probability of exceeding the threshold. VI. CONCLUSIONS AND FUTURE WORKS A handful of studies have developed location tracking applications through SMS. However, SMS is a user-pay service. The number of SMS transmissions must be minimized while maintaining the location tracking accuracy within the acceptable range to reduce the transmission cost. This study proposes a novel solution, LBD, to this problem, and further develops a realistic system for tracking the target location. In addition to dening the short message format, LBD uses the current location, speed, and bearing of the target to predict its next location. In LBD, the moving pattern information of the target is transmitted only when the distance between the predicted location and the actual location exceeds a certain threshold, which is dynamically adjusted according to the speed of the target. The experiment shows that, in LBD, the number of short messages required is signicantly reduced as compared with TBD and DBD. In addition, LBD achieves an acceptable location tracking accuracy. Finally, the use of a dynamic threshold reduces the required number of short message transmissions compared with the xed threshold. In this evaluation, we have assumed that a target moves erratically at low speed. Thus, the proposed LBD nds potential applications for elderly care and childcare. In addition, LBD is used in car monitoring and tracking applications because

60

JOURNAL OF COMMUNICATIONS AND NETWORKS, VOL. 15, NO. 1, FEBRUARY 2013

it works under the condition that the target moves at a high speed. However, further studies are required to verify these applications. A notable limitation is that LBD can only track one target at a time. We extend this work for future studies on monitoring multiple targets simultaneously by taking into account additional value-added services. REFERENCES
[1] [2] [3] [4] [5] [6] [7] [8] H. H. Lee, I. K. Park, and K. S. Hong, Design and implementation of a mobile devices-based real-time location tracking, in Proc. UBICOMM, 2008, pp. 178183. Z. Tian, J. Yang, and J. Zhang, Location-based services applied to an electric wheelchair based on the GPS and GSM networks, in Proc. ISA, 2009, pp. 14. I. Lita, I. B. Cioc, and D. A. Visan, A new approach of automobile localization system using GPS and GSM/GPRS transmission, in Proc. ISSE, 2006, pp. 115119. P. Perugu, An innovative method using GPS tracking, WINS technologies for border security and tracking of vehicles, in Proc. RSTSCC, 2010, pp. 130133. S. A. Hameed, O. Khalifa, M. Ershad, F. Zahudi, B. Sheyaa, and W. Asender, Car monitoring, alerting, and tracking model: Enhancement with mobility and database facilities, in Proc. ICCCE, 2010, pp. 15. R. E. Anderson, A. Poon, C. Lustig, W. Brunette, G. Borriello, and B. E. Kolko, Building a transportation information system using only GPS and basic SMS infrastructure, in Proc. ICTD, 2009, pp. 233242. W. J. Choi and S. Tekinay, Location-based services for next-generation wireless mobile networks, in Proc. IEEE VTC, 2003, pp. 19881992. R. E. Anderson, W. Brunette, E. Johnson, C. Lustig, A. Poon, C. Putnam, O. Salihbaeva, B. E. Kolko, and G. Borrielllo, Experiences with a transportation information system that uses only GPS and SMS, in Proc. ICTD, 2010. A. Civilis, C. S. Jensen, and S. Pakalnis, Techniques for efcient roadnetwork-based tracking of moving objects, IEEE Trans. Knowl. Data Eng., vol. 17, no. 5, pp. 698712, 2005. M. Zahaby, P. Gaonjur, and S. Farajian, Location tracking in GPS using Kalman lter through SMS, in Proc. IEEE EUROCON, 2009, pp. 1707 1711. A. Civilis, C. S. Jensen, J. Nenortaite, and S. Pakalnis, Efcient tracking of moving objects with precision guarantees, in Proc. MOBIQUITOUS, 2004, pp. 164173. Y. Y. Xiao, H. Zhang, and H. Y. Wang, Location prediction for tracking moving objects based on grey theory, in Proc. FSKD, 2007, pp. 390394. P. H. Tseng, K. T. Feng, Y. C. Lin, and C. L. Chen, Wireless location tracking algorithms for environments with insufcient signal sources, IEEE Trans. Mobile Comput., vol. 8, no. 12, pp. 16761689, 2009. R. Bajaj, S. L. Ranaweera, and D. P. Agrawal, GPS: Location-tracking technology, Computer, vol. 35, no. 4, pp. 9294, 2002. Movable Type Scripts. (2012 June). [Online]. Available: http://www.mova ble-type.co.uk/scripts/latlong.html

Jian-Wei Lin received his Ph.D. degree in the Department of Information Management at National Taiwan University of Science and Technology in 2008. He then joined the Faculty of the Department of International Business at Chien Hsin University in 2009. His research interests are on E-Learning, Web-based applications, and network protocol.

Yi-Hsuan Yeh received her B.S., M.S., and Ph.D. degrees in Information Management from National Taiwan University of Science and Technology, Taiwan, in 1995, 1997 and 2012, respectively. She is currently an Assistant Professor in the Department of Information Management at Oriental Institute of Technology, Taiwan. Her research interests are in the areas of location-based social networking, positioning technologies, and design and implementation of the smartphone application.

[9] [10] [11] [12] [13] [14] [15]

Ching-Neng Lai received his B.S. and M.S. degrees in Power and Energy division of Electrical Engineering from National Taiwan University of Science and Technology in 1995 and 1998, respectively, and his Ph.D. degree in Information Management from National Taiwan University of Science and Technology in 2009. He joined the Faculty of the Department of Information Technology at Hsing Wu University as an Assistant Professor in August 2009. His research interests include performance analysis and protocol design on networks.

Hui-Chuan Weng received her B.S. and M.S. degrees in Information Management from National Central University in 2008 and National Taiwan University of Science and Technology in 2010. She joined the lab of High Speed Networks at National Taiwan University of Science and Technology in August 2008.

Yaun-Cheng Lai received his Ph.D. degree in the Department of Computer and Information Science from National Chiao Tung University in 1997. He joined the Faculty of the Department of Computer Science and Information Science at National Cheng Kung University in August 1998. He then joined the Faculty of the Department of Information Management at National Taiwan University of Science and Technology in August 2001 and has been a Professor since February 2008. His research interests include performance analysis, protocol design, and wireless networks.