Low Energy Geographical Routing Protocol for Wireless Multimedia ...

Low Energy Geographical Routing Protocol for Wireless Multimedia Sensor Networks I.Bennis∗‡ , O.Zytoune∗§ , D.Aboutajdine∗ , H.Fouchal‡ ∗ LRIT,

unité associée au CNRST (URAC29) ,Faculté des Sciences, Université Mohammed V - Agdal , Rabat, Maroc. § ENCG, Kenitra, Maroc. ‡ CReSTIC,EA 3804,Université de Reims Champagne-Ardenne France Email: [email protected], [email protected], [email protected], [email protected]

Abstract—The field of wireless multimedia sensor networks(WMSN) is attracting more and more research community as an interdisciplinary field of interest. This type of network is low-cost, multifunctional due to advances in microelectromechanical systems, and the proliferation and progression of wireless communications. However transmit multimedia information captured must satisfy the criteria of QoS , which increases energy consumption, fact that should be taken into consideration in the design of any routing protocols for WMSNs. In this paper we present routing protocol which we call an Energy Aware TPGF (EA-TPGF), that improves the Two Phase geographical Greedy Forwarding (TPGF). The basic idea is to take into account the residual energy of the nodes in the identification of pathways. Simulations and statistics produced by the simulator NeTtopo showing a significant contribution with regard to the extension of the life of the network. Keywords: WMSN, TPGF, EA-TPGF, Energy, Routing protocol, QoS. This is the accepted version of the following article: Low energy geographical routing protocol for wireless multimedia sensor networks, which has been published in final form in [1].

I. I NTRODUCTION

I

N recent years, Wireless Sensor Networks Wireless Multimedia (WMSNs) have aroused much attention. Including features and applications they offer. The constraints related to deployment, topology, and energy motivates much of the research on the issues WMSNs. The majority of routing protocols developed for this type of network focused on meeting the criteria for Quality of Service (QoS). But the multimedia convergence also requires satisfaction side energy management view that the media stream is greedy in that. Therefore, we found it interesting to study a protocol, based on QoS and take into account the real-time constraint in order to study its advantages and drawbacks and thus provide a possible improvement. The remainder of the paper is organized as follows; in section I, we present a brief description of WMSNs, its architecture and design challenges, in section III we introduce relative works, and in section IV we propose the EA-TPGF ,our work to enhance the TPGF routing protocol, the results of extensive simulations are shown in section V. Finally section VI, concludes the paper. 978-1-4673-2480-9/13/$31.00 © 2013 IEEE

II. W IRELESS S ENSOR N ETWORKS W IRELESS M ULTIMEDIA The WMSN is an improvement of the traditional sensor network, given that new features have emerged. The development of this type of network was solicited by the availability of inexpensive CMOS cameras and microphones. In addition to the ability to retrieve multimedia data, the WMSN is also capable of storing, processing in real time, correlate and merge multimedia data from heterogeneous sources of traffic. These networks promising a wide range of potential applications in the civil and military fields that require visual and audio information such as sensor networks monitoring, traffic control systems, automated support for the elderly ... etc. To reduce energy consumption many works have been done, like in [2] where athors handle the paradigme in the MAC layer, or in [3] where the Slot scheduling mechanism is presented, which can be able to guarantee communications over all sensors and to save a high amount of energy. In [4] authors presente Energy efficient approch combined with the QoS criteria. A. Architecture of WMSNs In the WMSNs, different types of traffic can be generated by the sensors. To address this variety, flexible and heterogeneous architecture must be designed. There are two types of basic architecture [5]: • Single-Tier: There are two types, either a set of homogeneous sensors with distributed processing and centralized storage, or a set of heterogeneous sensors with distributed processing and storage. • Multi-Tier: This hierarchical structure provides the flexibility to use network resources adaptively, it offers considerable advantages in terms of scalability, cost minimization, better coverage, and better reliability. Figure 1 is a summary diagram of the two types of architectures Defining an architecture for WMSNs also depends on the coverage. An important challenge in this area follows from different coverage properties of multimedia sensors, such as cameras and microphones, relative to scalar sensors such as temperature sensors and humidity, as shown in Figure 2.

Fig. 1.

The reference architecture [5]

multimedia streaming and do not satisfy the constraints of new sensor network. There are several family of routing protocols for WMSNs in particular the gradient routing and geographic routing. In this work we will focus on the laste one. The notion of geographic routing is such that a node is assumed to know its geographical position, that of its neighbors and of the destination node. Knowledge of the neighborhood is periodically updated using HELLO messages exchanged between neighboring nodes. For WMSNs, several protocols based on geographic coordinates were defined as GPSR, MPMPS and TPGF, this last one was our topic of study and going to be more detailed in the following section. GPSR

Fig. 2.

Illustrated the cover of scalar sensors (a) and video sensor (b) [6]

GPSR (Greedy Perimeter Stateless Routing) [7] was originally designed for mobile ad-hoc networks, but it was quickly adapted for sensor networks. This protocol uses two different strategies for routing: Greedy Forwarding and Perimeter Forwarding. MPMPS

B. Design challenges of WMSNs There are several criteria that influence the design of WMSN, among them: 1) Multimedia source coding: Production, processing and transmission of multimedia traffic such as audio, video, or simple images is a great challenge. Therefore, the raw data should be compressed by sophisticated coding exploiting the redundancy in the images or video. 2) The high demand for bandwidth : Although multimedia coding techniques significantly reduce the transmitted information, the information compressed still exceeds the current capacity of wireless sensor nodes. More specifically, the video stream requires a transmission bandwidth that is higher than what is supported by currently available sensors. 3) Application requiring specific QoS : The routing techniques and communication proposed so far for sensor networks typically follow a ”best effort” service approach. In other words, no guarantees in terms of energy consumption, delay, jitter, or bandwith are provided. On the other hand, multimedia applications have special need for these types of guarantees for efficient transmission of the detected phenomenon. 4) Energy consumption : Energy consumption is a major concern in traditional sensor networks. This is even more pronounced in WMSNs, given that the sensors have batteries with a limited lifetime, whereas multimedia applications produce high volumes of data that solicit high transmission rates and thus intensive energy consumption. III. R ELATED WORK Efficiently transmit multimedia streams in WMSNs is an important problem to solve, due to the limited bandwidth and energy resources of sensor nodes. Routing protocols designed for traditional sensor networks are inappropriate for WMSNs since they do not support the transmission characteristics of

MPMPS (Multi-Priority Multi-Path Selection) [8] is a extension of the protocol TPGF. It highlights the fact that the paths found by TPGF may not all be used for video transmission seen that long paths are not desirable for this type of transmission. In addition, the audio / video stream may play different roles depending upon the application, and the importance of each stream can be different, it is therefore preferable to separate them, thus, we can give the highest priority to the most important flow according to the application, which ensures the use of good paths for different flows. IV. TPGF PROTOCOL The routing protocole Two Phase geographical Greedy Forwarding (TPGF) was the first to introduce the concept of multi-paths in the field of WMSNs [9]. This algorithm focuses on the exploration and establishment of a maximum number of best disjoint routes in terms of delay end-to-end. The first phase of the algorithm is responsible for finding a routing path that offers guaranteed delivery, while avoiding the holes. This phase consists of two methods: greedy forwarding, where the node transmitting the information, always chooses the next hop node that is closest to the base station among all neighboring nodes(figures 3 and 4). The second method, step back & mark which handles the problem of blocking nodes (figure 5). The second phase of the routing protocol is responsible for optimization of paths found by reducing the number of hops. This optimization includes only one method: label based optimization, which is responsible for removing the circles that can appear in a path routing(figure 6). TPGF algorithm can be executed repeatedly to discover several disjoint paths.

Fig. 6. Fig. 3.

Fig. 4.

Exemple of path circle

Exemple of greedy forwarding

Exemple of greedy forwarding -case of hole

Fig. 7.

score calculation of a neighbor

V. EA-TPGF PROTOCOL

Fig. 5.

Example of node blocking and blocking situations

Characteristic of TPGF TPGF supports three features which generally must be included in any routing protocole for multimedia transmission in WMSNs, these requirements are: • The multi-path transmission : the multimedia data packets are usually large in size and transmission requirements can be several times higher than the maximum transmission capacity (bandwidth) of sensor nodes. This requires that the multipath transmission be used to increase the performance of transmission in sensor networks. • Bypassing holes: static or dynamics holes can occur if multiple sensor nodes in a small area are overloaded due to multimedia transmission. So to ensure communication, the routing protocol must be able to avoid these holes. • Transmission with the shortest path: multimedia applications usually have a delay constraint, which requires that the multimedia streaming in wireless sensor networks should always use the shortest path routing which has the minimum transmission delay from end to end. We note that the multi-path exploration is only used in the case where the amount of data to be transmitted exceeds the capacity of the bandwidth of the sensor, otherwise, only one path will be sufficient.

After seeing the concept of TPGF and the strengths it brings, namely, the holes bypassing, choosing the shortest path and use the notion of multi-path, TPGF appears as an interesting protocol. Nevertheless, there sulfur of some weaknesses, especially the fact that it is based only on the distance in the selection of next hop. So if we consider that there is no change in the topology, the protocol will always choose the same path for fixed source and destination. This means an overload of using some nodes compared to others, a fact that leads to an early death of these nodes and thus a shortest network lifetimes. Our approach is to improve this weakness, by modifying the manner in which the protocol chooses paths, and adding energy as a metric. We noted the new version of the protocol by Energy Aware - Two Phase geographical Greedy Forwarding (EA-TPGF). Distance-energy formula The choice of energy as a metric to determine the routing path was designed to balance the use of sensor nodes in a network in order to extend much as possible their life. Now it is clear that if we base ourselves only on the distance to calculate the paths we have the advantage of having the best in terms of delay, thus we were to have a formula between energy and the distance that satisfies our approach while respecting in the best possible way, the constrained of delay. Let j the sensor source and i1 , i2 , i3 ...in the adjacent sensors of j (Figure 7). The formula is therefore :

TABLE I M AIN CONFIGURATION PARAMETRES

α scorej (i) = Dis ∗ Ei −β (1)

Parameters Network size Number of sensor nodes Number of source node Transmission range (TR) bandwidth Packet size Delay of one hop α, β

with D is the distance between the sensor i to the destination s which is the sink, and E is the residual energy of the sensor i at the time of routing decision. This formula represents the score of a neighbor according to its distance to the destination and its energy, then the sensor j selected from its neighbors the one with the minimum score. Constants α et β are two weighting parameters.

value 400m*400m 100 ¨ı¿½ 500 1 60m 128 Kbits/s 8 Kbit 20 ms 0.6, 0.4

Energy cost of transmission When a node A sends a packet with a length of n bits to ¯ another node B, the energy of A decreases by ET X (n, AB) while energy of B decreases by ERX (n), assuming that the packets transmitted over the network are all the same size, ¯ + ERX (n). The the cost of transmission is ET X (n, AB) ¯ are based on the two functions ERX (n) and ET X (n, AB) model of energy consumption proposed by Heinzelman [10]. According to this model, we have :

ET X (K, D) = k.(EELEC + amp .D2 )(2) ERX (K) = k.EELEC (3) With : K is the packet size in bits, D is the distance of transmission in meters. EELEC is the energy consumed by the transmitter, amp is the energy consumed by the amplifier. As in [10], we took EELEC = 5µJ/bit and amp = 1ηJ/bit. VI. S IMULATIONS AND ANALYSIS OF RESULTS In this section we present various experiments conducted with the simulator NetTopo [11], to compare our approach with the results announced by the TPGF protocol [9]. We will first present the simulation parameters, then we will proceed with the different scenarios performed with an analysis of the results. Simulation environment Setting used during these simulations are summarized in table I : The aim of this simulation is to calculate the number of periods completed until the exhaustion of the energy of one node in the network. A period is defined as a successful transfer of a batch of n packets. Figure 8 shows clearly that for a given topology the EA-TPGF makes higher number of periods that TPGF which means better management of energy resources in the case of EA-TPGF, we also note that the more the number of nodes increases, the more number of periods increases too. Another thing that we can obviously see regarding the protocol TPGF is that the number of periods

Fig. 8.

Number of period done VS Number of sensors

does not varied when the number of nodes exceeds 400, it can be explained by the death of some sensors which formed the paths due to the hard use of the same paths and thereby the same nodes. Figure 9 leds also to the same idea, it represents the residual energy of nodes that are used by TPGF and also by EA-TPGF after stopping the transfer (after depletion of the energy of the first node in the topology) , for display purposes this figure is just about the case where the number of sensor is 100. so we find that the energy of these sensors is sharply reduced implying a reduction of paths found and thus limitation of the number of periods carried. Figure 10 Figure shows in some way, the disadvantage of the protocol EA-TPGF, which resides in the point that it has a longer delay than the TPGF. This weakness can be neglected if the scope of the EA-TPGF is not critical to the time factor, especially as delays difference between the two protocols is only of a few ms. VII. C ONCLUSION TPGF presented as a protocol that satisfies the constraints relating to the nature of the multimedia sensor networks, such as delay and energy consumption. However this protocol suffers from some drawbacks and especially the fact that it always uses the same discovered paths for other later transmissions , and thus decreases the lifetime of network. Our approach is based on this problem in order to find a solution, thing that we have conducted by a compromise between two metrics: energy and distance. The results found, show much better exploitation of the network. For future work, we will

[4] Jalel Ben-Othman and Bashir Yahya, Energy efficient and QoS based routing protocol for wireless sensor networks, J. Parallel Distrib. Comput., 70, 8, 2010, 849-857, http://dx.doi.org/10.1016/j.jpdc.2010.02.010, DBLP, http://dblp.uni-trier.de [5] Ian F. Akyildiz and Tommaso Melodia and Kaushik R. Chowdhury , A Survey on Wireless Multimedia Sensor Networks , Computer Networks,2007,51,921-960 [6] Ian F. Akyildiz, Mehmet Can Vuran ,Wireless Sensor Networks ,373397,2010 [7] B. KARP AND H. T. KUNG. GPSR: Greedy perimeter stateless routing for wireless networks , in Mobile Computing and Networking ,243-254 ,2000. [8] L. Zhang, M. Hauswirth, L. Shu, Z. Zhou, V. Reynolds AND G. Han, Multi-priority Multi-path Selection for Video Streaming in Wireless Multimedia Sensor Networks,Lecture Notes in Computer Science, vol. 5061,2008, 439-452. [9] Lei Shu , Yan Zhang , Laurence T.Yang , Yu Wang , Manfred Hauswirth and Naixue Xiong, TPGF: geographic routing in wireless multimedia sensor networks, Springer Science+Business Media, LLC,2009. [10] W. Heinzelman, A. Chandrakasan AND H. Balakrishnan.EnergyEfficient Communication Protocol for Wireless Microsensor Networks ,In Proceedings of the 33rd Hawaii International Conference on System Sciences (HICSS),2000, 4-10. [11] http://www.semanticreality.org/nettopo/index.htm [12] VINT, The network simulator ns-2,2012,http://www.isi.edu/nsnam/ns.

Fig. 9.

Fig. 10.

Residual energy of TPGF VS EA-TPGF

Average delays VS Number of sensors

look to reduce the delay in the proposed protocol, also we are planning to move to a simulator which is more robust and better known as NS2 [12] which offers better envirenement to simulate networks. R EFERENCES [1] I. Bennis, O. Zytoune, D. Aboutajdine and H. Fouchal, ”Low energy geographical routing protocol for wireless multimedia sensor networks,” Wireless Communications and Mobile Computing Conference (IWCMC), 2013 9th International, Sardinia, 2013, pp. 585-589. doi: 10.1109/IWCMC.2013.6583623 [2] Thibault Bernard and Hacène Fouchal, A low energy consumption MAC protocol for WSN, ICC, 2012, 533-537, http://dx.doi.org/10.1109/ICC.2012.6363918, DBLP, http://dblp.unitrier.de [3] Thibault Bernard and Hacène Fouchal,Slot scheduling for wireless sensor networks, J. Comput. Meth. in Science and Engineering,12, Supplement-1,2012,1-12, http://dx.doi.org//10.3233/JCM-20120432, DBLP, http://dblp.uni-trier.de