(IJEECS) International Journal of Electrical, Electronics and Computer Systems. Vol: 13 Issue: 1, March 2013

Energy-Efficiency and Reliable Protocol based on Virtual Zones Infrastructure for Wireless Sensor Networks (ERVM) Amina MERBAH, Ahmed KAMIL, Hicham BELHADAOUI, Mohamed OUZZIF *GI Department, RITM-ESTC / CED-ENSEM, University Hassan II Km 7, Eljadida Street, B.P. 8012 Oasis, Casablanca, Morocco Abstract-- The field of wireless sensor networks has covered up a large part in the various aspects of scientific research, thanks to its applications that operate in most active domains, such as military purposes, medical and industrial goals… Large scale sensor networks generally consist of hundreds or even thousands of sensor nodes. Establishing a routing process to produce global coverage of a monitored area is the biggest challenge, since the implementation of energy efficiency of wireless communications in each sensor node is critical. Each routing protocol designed for WSN should be reliable, efficient and ensures the increase of the life time of the network. In this paper we present a new routing protocol that supports reliability and energy efficiency ERVM, we evaluate this protocol with a set of existing routing protocols (ZHLS and RZRP ). ERVM considers the residual battery capacity of each sensor node to establish routing paths and to offer the optimal path between several sensor nodes in some specific cases. The evaluation results show the effectiveness of the VZP protocol in terms of end to end delay, and in terms of the number of delivered packets and energy-efficiency.

the effectiveness of communications. It can also increase stability period and lifetime of the network. •

Fault tolerance: The network must be able to maintain its functionality without interruption in case of failure of one or more of its sensors. This failure can be caused by energy loss, physical damage or environmental interference. The tolerance degree depends on how critical the application and exchanged data are. These are also very important to be supported by the routing protocol to ensure proper maintenance of all of routes already executed, and declare the breakdown of these routes in the right time in order to take another decision based on the conception of the routing protocol.

The protocol that we present in this paper meets all the requirements that characterize the operating environments of WSN, and fits our proposed platform VMLI [4].

Keywords--- Virtual zones, energy-efficiency, routing protocols

I. INTRODUCTION Because of economic considerations and uses, sensor nodes are limited in terms of their small size which reduces the storage resources and computation of each entity. Thus, sensor nodes are equipped with irreplaceable batteries in harsh environments, this makes energy a crucial feature in WSN applications. Nodes in a WSN communicate through short-range radio, which requires communication through multi-hop routing to resubmit information to its destination. A routing protocol designed for WSN must satisfy a set of requirements. First, its conception should be simple in order to minimize the computational complexity at the sensor nodes and at the responsible node in some type of WSN topologies, using the following mechanisms: •

•

The network balancing: The main function of a WSN is to collect information relevant from an area of interest. Some applications, such as environmental monitoring, the WSN need work as long as possible. Thus, the extension of the lifetime of the WSN is an important objective of each routing protocol. A good routing protocol should ensure energy balancing in order to extend the lifetime of the WSN. Energy efficiency: Balancing the load between network nodes can improve the energy efficiency of sensor nodes to extend the lifetime of networks and improve

The article is organized in the following way. Section 2, presents a study of the protocols developed for wireless sensor networks. This is classified into three categories based on flat architecture, hierarchical and location-based. Section 3 and 4, describes some assumptions and our contribution developed about the new designed protocol. Section 5 presents and explains our protocol functionalities, operating on three phases: Preparatory phase, emission phase and maintenance phase. Section 6 deals with experimental work and presents related simulation results. Finally section 7 concludes the proposed work and suggests several ideas for future undertaking. II. PREVIOUS RESEARCH A. Protocols based on a flat architecture: Flat architecture protocols are often data-oriented, unlike protocols based on addressing each sensor element to have a unique identifier. In principle these protocols are developed for dense network types. Gossipping [1]. It is inspired by the principle of spreading rumors in social networks. Upon receiving a message, each node chooses one of its neighbors at random to relay the message and so on. After a moment, all nodes receive a copy of the message. Gossip protocol can limit the worst network load. However, it increases the latency of propagation of messages to all nodes in the network. The directed diffusion (DD) [2] is a model based on publication / subscription. The sinks broadcast a request

©IJEECS

(IJEECS) International Journal of Electrical, Electronics and Computer Systems. Vol: 13 Issue: 1, March 2013 (called interest) to query the network on a particular data. It operates in four phases to build routes between the sinks and the sensors involved in the broadcast interest. The four phases are: (1) interest spread, (2) establishment of gradients and propagation data (3) strengthening paths(4).

Figure.1: Example of DD protocols operations. B. Hierarchical protocols: LEACH [3] is a routing protocol which optimizes the energy consumption in the network by forming clusters. Thus, the purpose of LEACH is to dynamically select nodes to act as cluster heads and form clusters in the network. Then other nodes join one of the clusters formed. These nodes transmit their data to the cluster head, which will aggregate them and send them directly to the sinks. One of the assumptions in LEACH is that all nodes can reach the sink in one hop. Cluster heads are exchanged periodically to ensure that the consumed energy through contacting the sink was distributed among all nodes. PEGASIS [5] proposes an improvement of LEACH by eliminating the additional cost generated by the created cluster operations. PEGASIS forms chains rather than clusters of sensor nodes. To build chains of nodes, each of them selects the nearest neighbor as the next hop in the chain. PEGASIS assumes that sensor nodes have a global knowledge of the network. PEGASIS forms a chain among the sensor nodes so that each node will receive from and transmit to a close neighbor. Gathered data moves from node to node, get fused, and eventually a designated node is transmitted to the BS. Nodes take turns in transmission to the BS so that the average energy spent by each node per round is reduced [1]. C. Location-based protocols : To make a routing decision, a node needs its geographical position, a destination as well as those of its neighbors. The position of the destination is often fixed in WSN, which is that of the sink. The position of neighbors is determined through an exchange of HELLO messages. The greedy algorithm [6] (the same name as its type indicates) is the simplest routing algorithm. If a node N wants to send to a destination D, it compares the position of all its neighbors D and chooses the closest one. This protocol requires neither infrastructure nor an organized graph in a certain way. Despite its simplicity, the greedy algorithm has some flaws. The most important of them is that it cannot find, in some cases, the path to a destination, even if there is one. This usually happens in areas with low densities or when there are holes in the network. An intermediate node finds itself as a local minimum on the source-destination path. GEDIR, Geographical Distance Routing, is an improvement of the greedy algorithm. When a message reaches a local

minimum instead of remove it as in the greedy algorithm, it gets another chance to escape from this minimum by transmitting it to the closest neighbor to the destination other than the node itself. GEDIR significantly improves packet delivery rate compared to the original version. However, it does not solve the problem completely. Minimum Energy Communication Network (MECN) [7] is a routing protocol that seeks to establish and maintain a minimum energy in wireless networks using low-power GPS. MECN uses a base station as destination of information, which is always the case for sensor networks. MECN identifies a relay region for each node. The main idea of MECN is to find a sub network that has fewer nodes and requires less power for transmission between any of both nodes. This is accomplished by using a localized search for each node by considering its relay region [8].Similarly, GAF (Geographic Adaptive Fidelity) [9] is a routing protocol based on the location of nodes. The location of nodes in GAF could be provided with a GPS or other positioning techniques [10]-[11]. III. ASSUMPTIONS AND CONTRIBUTION: We consider a wireless sensor network that consists of a routing layer partitioned into areas with similar surfaces, whereby each zone is located geographically in relation to the coordinates of the routing layer. A surface area varies with the nature of network and the number of devices deployed in the latter to establish global monitoring (Figure2). Each zone is composed of a number of “coordinator nodes” as FFD (Full function device), according to the IEEE 802.15.4 standard, which has the ability to have priority access to the transmission channel and a set of “simple nodes” RFD (Reduce function Device), which presents the sensor nodes with the ability to measure the information from their monitoring area and send them to the coordinators nodes. A simple node may at a certain moment become a coordinator node of a series of simple nodes. The basic protocol for access to the transmission channel is unslotted CSMA / CA proposed by IEEE 802.15.4 standard.

: A simple node

: Direct link

: A coordinator node

: Wireless link

Figure.2: Example zones of our proposed topology of wireless sensor network.

(IJEECS) International Journal of Electrical, Electronics and Computer Systems. Vol: 13 Issue: 1, March 2013 • •

•

The protocol does not require maintaining all the previously established routes by the whole nodes of the network. Each sensor node maintains information concerning addresses of coordinators in its zone, in order to communicate with other sensors to tolerance in case of failures, which reduces the charge of the storage memory of sensor nodes. Each coordinator must know the coordinators addresses of its zone as well as addresses of those of adjacent neighboring areas.

The purpose of this calculation (Eq.2) is to select the best hop neighbors from a node u in order to select the relay points of this node u.Figure.3 explains the different steps to find the efficient relay points for each coordinator nodes. -

The minimum number of communication links. The maximum energy wmax(u)
#$%() = (, ) ∗

Communication between nodes is performed through sending packets to the coordinator node: •

•

If the destination node is in the same zone, the coordinator is responsible for transmitting the packet to the responsible coordinator of the destination node, which in turn will transmit it to the object. If the target node is not in the same zone, the coordinator seeks the shortest path to establish the transfer, this is carried out by searching the zone of the node and adjacent zones of the latter in order to route the packet to its destination.

The principal aim of this protocol is its simplicity, the lower cost of computational complexity, and the choice of the correct route with the least charge and maximization of the life of the network. 1. IV. PROTOCOL FUNCTIONALITIES OF ERVM : The ERVM protocol operates in three phases: - Preparatory phase. - Emission phase. - Maintenance phase. A. Preparatory phase: This phase consists of ensuring connectivity between all coordinator nodes in different areas of the routing layer to determine the set of neighbors of each coordinator node. A node is considered as neighbor node to another node if the probability of receiving the messages 'hello' is greater than a threshold p0, Each node is able to evaluate p (u,v) , to calculate the probability with no errors reception , As proposed in the work [12](Eq.1).


() = (, ) + 

() ∩ () 



( ,  )

(Eq. 1)



( '( + ') )

∗ (+(, ) ,  )

(Eq. 2)


() : Returns the nodes weight in terms of energy and links numbers, such as the node with the maximum energy has the highest weight. -./() : The set of two hop neighbors of node u. 0() : The set of one-hop neighbors of u. 0() : The set of two hop neighbors of node u. ') : The rate coefficient of energy. '( : The coefficient of link numbers (, ) : The probability of receiving a message without error in a realistic environment. +(, ) : The Euclidean distance between node u and node v. The value of coefficient β varies depending on distance between points B. Emission phase: 1) The structure of the control packet "RequestConstruct": We assume that all nodes in the network are referenced by their geographical coordinates, they will be used in the first virtual layer according to a proposed Infrastracture of WSN [4]. So, before each node joins the network it must be assigned to a zone by the location of its coordinates relative to those of the target zone. In addition, nodes are assigned to a specific area according to their function in the network. For this purpose, we define a control packet that enables to perform this task (Table 1): Parameters SrcN-ID DestN-ID SrcZ-ID DestZ-ID SrcAgr-ID DstAgr-ID

Our proposal uses the same logic, based on the MPR (Multi Point Relay), technique that ensures diffusion in all coverage sub-zones. Due to the critical factor of energy in AdHoc and sensor networks, we have improved the MPR approach by adding two computation criteria based on the weight of each neighbor node, and the number of the links between the source and the destination node. At first the computation of relay nodes is done by each coordinator node to find the relay points relating to its area. And those relating to adjacent zones, until the total coverage establishment of all zones.

|() ∩ () |

Req-ID ReqS DataTable InterTable ETE N-energy

Description Identifier of the node sending the request Identifier of the node receiving the request Identifying the zone of the transmitting node Identifying the zone of the receiving node Identifier responsible aggregator of the source node. Identifier responsible aggregator of the destination node Identifier of the sent request The subject of the request Contains all transmitted packets Composed of all adjacent zones The time of the request The rate of energy node

Table.1 Packet Structure of the "zones Requests"

(IJEECS) International Journal of Electrical, Electronics and Computer Systems. Vol: 13 Issue: 1, March 2013 •

Each table "InterTable" contains the following information: •

•

NeighboZones: Adjacent zones to the zone of transmitting node.

-./() = ϕ

& -./() = 0()

Ǝ w | w ∈ -./()

Aggregator: List of responsible coordinators in adjacent zones. Metric-lst: List of hops nodes to reach the destination node.

Find isolated points in -./()

Ǝ! v ∈ 0() w ∈ 0()

if (, ) = p0

w#$%( ) = ') * E j i++, j++

Calculate the weight of each node in the zone according to ') and '(

A( ) = '( ∗ BCD( E) i++, j++

Extract from wf a number k node with the highest weight (k depending on the number of nodes)

w() G( H) /JDKLCMK(N) -./ -./() = -./ -./() ⋃w w() () = () ⋃ ()

P0 = P0 - £

Figure3. Process of choice MPR points 2) Procedure for a request establishment: When a source node needs to transmit a packet to another destination node, it first checks its data table DataTable to find a valid path, otherwise, it initializes a control packet ZonesRequest (Table.1). The detail of sending this packet is illustrated by the algorithm .1. Algorithm 1: Function: ZoneRequest (ReqId)

8: 9:

EndIF Break();

1: Case1: 2: Begin(); 3: Receive (ReqId) { 4: ETE0: = CurrentTime. 5: Check (TableData); 6: If TableData.contains (ReqId.ReqS) = = Null Then 7: TableData.add (ReqId.ReqS) ;}

14: If TableInter.Metric-lst.contains (PathOf (ReqId)) Then 15: TableInter.add (DestZId.DstIdAg); 16: SendToDestination (ReqId); 17: EndIF

10: Case2: 11: Else If (DestZId < > SrcZId) Then { 12: Send (ZoneRequestPacket); 13: Check (SrcAgrId.SrcNId) {

18: 19:

Metric-lst .add (DestZId.DstIdAg.DstNId); Update (TableInter);

(IJEECS) International Journal of Electrical, Electronics and Computer Systems. Vol: 13 Issue: 1, March 2013 20: 21: 22: 23:

Update.ETE:=CurrentTime-ETE0 Update.N-energy; } End Else IF

24: Case3: 25: Else If { 26: Process SrcAgId 27: Check.lst (Aggregator. NeighboZones) 28: Invok.MPRMethod(); 29: Send Packet.requestId; 30: Update (TableInter.Metric-lst) 31: Update.ETE:=CurrentTime-ETE0 32: Update.N-energy; 33: } 34: End Else IF 35: End The establishment of ZoneRequest packet consists of three phases: If the target zone has the same identifier as that of the recipient zone Part.1, then it suffices to forward the packet to the source node responsible for sending the request to the one responsible for the target node. Secondly, if the target node is located in another zone, the ZonesRequest packet is sent to the adjacent zones. When these latter receives requests, an initial research is done at InterTale of each zone. If the path of the request has been registered, the data are sent to their destination by the process mentioned in Part.2. The Part.3, if there is no occurrence of the request subject of sending in adjacent zones, a global broadcast with propagation is restarted in the set of all coordinators of all zones in order to find the optimal path relative to the energy and links level. If the subject of the request reqs is of a "critical" type, the processing chooses the shortest path with partial study of the energy level of gateway nodes (in terms of choice between two paths equivalent in number of hops but one way is more energetic than the other). In cases whereby the subject of the request is of a "normal" type, the process takes into account the maximization of the lifetime of the network, for this it chooses the path of least energetic (by the shortest path or through the longest one),as relay points already found. It is possible to determine relay points of each coordinator in an initial phase which results in the coverage of all sensor nodes in the architecture. This approach allows us to save time as well as computation units. Due to the factor of change topologies of WSN, either by failure of a responsible unit such as a coordinator or its disappearance due to the lifetime... All these factors may occur as a sudden change in the initial topology. For this purpose, we have chosen to relaunch the process of selecting relay points every time the need should arise. C. Maintenance phase: Maintenance of internal tables and data tables is required. After sending a packet to a DataTable or after a failure event or disappearance of a node in the whole network and this for InterTable.

After building our global application with its zones and its routing tables, we periodically send diagnostic messages "Hello". These messages are exchanged at the level of all coordinators units to inform themselves of the disappearance of either a coordinator node or else a son node (sensor), to ensure that the aggregator declares in state "still available" in order to accommodate other simple nodes not assigned to any responsible to establish monitoring requested. After failure detection, some update messages are sent to update the entire internal table, of the error zones and internal tables of neighbor zones. The advantage of sending messages periodically so short as that the "Hello" messages with an interval time allows us to reduce the number of exchanged data during control, compared to other protocols that exchange entire data tables between nodes for updates. This allows us to save energy consumption that would prolong the network lifetime. D. Reliability and fault tolerance: In the proposed protocol, we consider an energy level which decreases after each submitting by a node. A minimum energy threshold £. If an aggregator finds that the energy of a simple node is less than £, then it declares the latter as unable to route an information, but its role is just to inspect until the global dissemination of this information in internal tables of the internal zone as well as adjacent zones. The deactivation process of the low node is presented in (Figure.4). The sensor node in orange is a sleeping node, It is then necessary to inhibit the yellow node as critical to be replaced by the fourth node (c). Step (d) presents three sensor nodes assigned to the responsible node. In this case the coordinator gets the state “yet available” to accommodate a new simple node.

Figure 4 Steps for replacing a sensor node out of service V. EVALUATION AND RESULTS: In order to assess the effectiveness of the proposed protocol, we compare a set of performance metrics of protocol operation by simulation with both protocols RZRP[14] and which are similar in terms of communication and efficiency, as RZRP design is based on that of ZHLS[15] with the addition of the concept of sharing the

(IJEECS) International Journal of Electrical, Electronics and Computer Systems. Vol: 13 Issue: 1, March 2013 surveillance zone into a set of sub-areas to manage well the routing through RZRP. All simulations were held on J-Sim with almost the same parameters used in the evaluation of partitioning topology [4]. A. Simulation environment: We have simulated the behavior of our architecture under the J-Sim simulator. Open-source, J-Sim is built on the basis of the ACA (Autonomous Component Architecture), developed entirely in Java. The basic elements of J-Sim are components that communicate by sending and receiving data through ports. The specifications of each behavior of a component are determined by contracts. Each component can be developed and tested independently of all other components of the architecture. This makes J-Sim environment a truly platform-independent, extensible and reusable [13].

According to the function Zone Request algorithm (Algorithm.1), the end of the delay will be different from that of ZHLS, as in the first phase there needs to be a research at the level of cache to find a valid path, so it is a probability p. If p = 0 then the end-to-end delay will increase, this is through the search of the target in adjacent zones. Figure 5 shows the end to end delay of the proposed protocol ERVM compared to ZHLS and RZRP protocols.

The simulation parameters for the experiments are as follows: a. b. c. d. e.

f. g.

The dimensions of the monitored area: 1000 m /1000m The number of sub areas varies between: 20, 24, 28, 32, 36, and 44. Number of mobile nodes: 0 Channel : Wireless The topography: 800: #X dimension of topography. 800: #Y dimension of topography. The communication range of nodes is : 80 meter Number of nodes : 402,602 et 700

B. The end to end delay: The end to end delay of the ZHLS protocol is presented by the equation (Eq.3): ]^

_

` TUVWX = ∑\ [ Z[ + ∑  

^(

+ (Ja. 3)

N: the total number of nodes. M: the total number of zones. Q: the percentage of active routes in a zone. l: the average number of zones to establish a route. T: average processing time of a node. t: average propagation delay of node. e Such as Z = c + K x d x represents time of \ searching an Idzone. + = K Propagation delay.

the

So the end to end delay of ERVM is as follows (Eq.4):

Tfgh\ = ∑l(Z  × (1 − .k$ )) + ∑\ [(Z[ × (1 − _

]× ×(

.$m )) + ∑[` +[

(Ja. 4)

A = Number of aggregators in the source zone. Pca: The probability to find a valid path in the cache memory of the source zone. Pad: The probability to find a valid path in the adjacent zones of the source zone.

Figure.5 The end to end delay The proposed protocol reduces the end to end delay in a remarkable manner according to the ZHLS and RZRP protocols. This is due to the structure of the request "PaquetRequest" containing two tables, an interzone table "InterTable" and an internal table contains the set of transmitted data by the local zone "DataTable". These tables are used respectively to maintain all paths already established between two different zones, thus providing valid paths within the same zone. These tables result the decrease of the end to end delay of transmission of a request when the internal or external path is already established, without going through for each transmission, the set of sub-zones of the routing layer. And the updates made in the zones to detect broken paths and find valid paths, allows at the moment of sending a request to minimize the search procedure of new routes. The factors we have just mentioned can produce better results of end to end delay compared to RZRP and ZHLS protocols C. Transmitted packets: The parameters selected to calculate the average delivered packets in the network to different constitutions of the routing layer are: 1. The total number of packets sent by the nodes. 2. The total number of packets received at the base station. Scenarios that we have considered to calculate the two parameters listed below for 402, 602 and 700 nodes by

(IJEECS) International Journal of Electrical, Electronics and Computer Systems. Vol: 13 Issue: 1, March 2013 choosing different constitutions such as 20, 24, 28, 32, 36,

Sent packets for 402 nodes (a):

Sent packets for 602 nodes (c):

Sent packets for 700 nodes (e):

44

sub-zones

of

routing

layer.

Received packets for 402 nodes (b):

Received packets 602 nodes (d):

Received packets for 700 nodes (f):

Figure.6 Sent and received packets in different networks

For each network constituted in sequence 402, 602 and 700 nodes both parameters (1) and (2) are averaged. For example, if the total number of sent packets per 100 nodes is 2000 and the number of packets that have been received at base station is 800, then the average delivered packets will be 800/2000 = 0.4. These averages give an idea of the number of lost packets during the routing process, due to congestion, a wrong choice of route for a packet and exhaustion of the energy of an intermediate node (during the execution time of the change process of failed nodes).

From the figures.6 a, c and e, we can see that the number of sent packets in different networks, when ZHLS RZRP are used, the number of sent packets increases relative to the use of the ERVM protocol. And Figures b, d, f show that when using ZHLS and ERVM the number of received packets is less than the number of sent packets in a proportional manner, but the use of ERVM is much better than ZHLS, and the difference between the number of packets received over the use of ERVM and ZHLS increases between the two when the number of nodes increases from 402 to 700 nodes. Other than the RZRP protocol that behaves differently than the two other, such as

(IJEECS) International Journal of Electrical, Electronics and Computer Systems. Vol: 13 Issue: 1, March 2013 the number of received packets in a network made up of a minimum number of nodes, otherwise this rate tends to zero. ERVM

ZHLS

RZRP

20

0,47018565

0,19802279

0,21688486

24

0,48450054

0,18236373

0,19918402

28

0,51325935

0,17843161

0,17048836

32

0,47626218

0,19160568

0,12531729

26

0,5289723

0,14768495

0,07686022

44

0,5216436

0,1338407

0,04898224

Jk : Energy consumption for each transmission. vJk : The total number of transmitted packets by each node (the number of nodes involved in the network construction). N: Total number of nodes.

Table.5 Average transmitted packets for 402 nodes ERVM

ZHLS 0,15975938

RZRP

20

0,60375517

0,17525778

24

0,68503577

0,18523217

0,17420089

28

0,7677194

0,26238888

0,12905949

32

0,83560056

0,32330169

0,04731465

36

0,90809862

0,40855858

0,02840571

44

0,97864168

0,49045462

0,01222073

Figure.10. Average Energy Consumption for 402 nodes

Table.6 Average transmitted packets for 602 nodes ERVM

ZHLS

RZRP

20

0,65932442

0,05906378

0,0361834

24

0,81129194

0,04541476

0,03592703

28

0,94279575

0,06308656

0,02901744

32

0,70446172

0,07821455

0,02934648

36

0,85520401

0,12393114

0,01668722

44

0,88254168

0,14430067

0,00455204

Figure.11. Average Energy Consumption for 602 nodes

Table.7 Average transmitted packets for 700 nodes To summarize, tables 5, 6 and 7 show the average rate for delivered packets in different simulation scenarios with the constitution of different zones. We can clearly observe that the best delivery rate is the ERVM protocols, which reach 0,97864168 for 602 nodes and 0,97864168 for 700 nodes, with 44 surveillance subareas: this is a very satisfying value for routing in a sensor network, and safe in terms of reliability of data transmission We will provide a set of algorithms and proposals for the evaluation of a sensor network in a realistic environment that is heavily dependent on natural requirements. D. Energy consumption: Energy consumption in a network in general is influenced by the number of packets transmitted over the network. As we have presented in this work, the evaluation of the first part of the construction, energy consumption rate is very reduced Figures.10, 11 and 12. So the average energy consumption is calculated using equation (5): ∑(,r) rfs tu e

o (Jk ) = Jk × pq

(5)

Figure.12 Average Energy Consumption for 700 nodes The energy consumed by a node depends necessarily both on the number of transmission and reception packets and on the number of calculations performed by the node. When there is comparison between the energy levels of each node, we see that the energy consumed by the act of transmission is considerably greater than that of the reception and calculation. Through simulations that we have performed, we assume that the average energy consumption is relatively proportional to the number of transmitted packets

(IJEECS) International Journal of Electrical, Electronics and Computer Systems. Vol: 13 Issue: 1, March 2013 in the network. Such is the case that the number of transmitted packets automatically increases the average of the energy lost increases. The simulation results presented by the figures 6,7. Once the ZHLS and RZRP protocols send more packets over ERVM for the same scenarios. The average energy spent by ERVMs strongly minimum than this of ZHLS and RZRP. VI. CONCLUSION In this paper we propose a routing protocol adapted to the VMLI infrastructure [4]. The design adopted is simple to deliver the packets to their destination with the minimum delay. The principal aim of this protocol is its simplicity, the lower cost of computational complexity, and the choice of the correct route with the least charges and with the maximization of the life time of the network. The ERVM protocol was inspired for making a simple solution to routing in wireless sensor networks, which has very less complexity and is easy to implement, while making it energy-efficient.

Acknowledgment The author wishes to thank Mr. Lhoussain SIMOUR for his proofreading and for his constructive comments which have successfully completed this work.

References [1] Sandra M. Hedetniemi, Stephen T. Hedetniemi et Arthur L. Liestman. A survey of gossiping and broadcasting in communication networks.Networks, vol. 18, no. 4, pages 319–349, 1988. [2] Chalermek Intanagonwiwat, Ramesh Govindan et Deborah Estrin. Directed diffusion : a scalable and robust communication paradigm for sensor networks. In Proceedings of the 6th annual international conference on Mobile computing and networking, MobiCom ’00, pages 56–67, New York, NY, USA, 2000. ACM. [3] W.R. Heinzelman, A. Chandrakasan et H. Balakrishnan. Energyefficient communication protocol for wireless microsensor networks. In System Sciences, 2000. Proceedings of the 33rd Annual Hawaii International Conference on, page 10 pp. vol.2, 2000. [4] A.Merbah, A.Kamil, O.Malassé, H.Belhadaoui, M.Ouzzif ,” Virtual Zones and Virtual Coordinates on a Multi Layer Infrastructure for Wireless Sensor Networks (VMLI)”,International Journal of Computer Science and Information Security,

ISSN: 1947-5500 Vol. 10, No. 11,pp:6-15, November 2012 [5] S. Lindsey et C.S. Raghavendra. PEGASIS : Powerefficient gathering in sensor information systems. In Aerospace Conference Proceedings, 2002. IEEE, volume 3, pages 3–1125 – 3–1130 vol.3, 2002. [6] G. G. Finn. Routing and Addressing Problems in Large Metropolitan- Scale Internetworks. Rapport technique ISI/RR-87-180, Information Sciences Institute, Mars 1987. [7] Rodoplu, V., Ming, T.H., 1999. “Minimum energy mobile wireless networks”. IEEE Journal of Selected Areas in Communications, Vol. 17, No. 8, pp. 1333-1344. [8] Akkaya , K., Younis , M., 2005. “A Survey on Routing Protocols for Wireless Sensor Networks”. Journal of Ad Hoc Networks, Vol. 3, No. 3, May, pp. 325-349. [9] Xu ,Y., Heidemann , J., Estrin, D., 2001. “Geographyinformed Energy Conservation for Ad hoc Routing”. In Proceedings of the 7th Annual ACM/IEEE International Conference on Mobile Computing and Networking (MOBICOM’01), Rome, Italy, July, pp.70-84. [10] Doherty, L., Pister, K.S.J., El Ghaoui, L., 2001. “Convex position estimation in wireless sensor networks”. In Proceedings of the IEEE INFOCOM, vol.3, Alaska, , pp.1655-1663. [11] Intanagonwiwat , C., Govindan, R., Estrin, D., 2000. “Directed diffusion: A scalable and robust communication paradigm for sensor networks”. In the Proceedings of the 6th Annual ACM/IEEE International Conference on Mobile Computing and Networking (MobiCom'00), Boston, MA, August. [12] M.LEHSAINI, H.GUYENNET, M.FEHAM, “ MPRbased broadcasting in ad hoc and wireless sensor networks with a realistic environment”, IJCSNS International Journal of Computer Science and Network Security, VOL.7 No.10, pp: 82-89, October 2007 [13] Ahmed,S., Chen ,W.P., Jennifer, C.H, Kung, L-C., Li, Ning., Lim, Hyuk., Tyan, H-Y., Zhang, H., 2005. J-Sim: A Simulation and Emulation Environment for Wireless Sensor Networks, To appear in Proceedings of the Annual Simulation Symposium(ANSS 2005), April. [14] W, chen., K, Yang., 2008.“Zone-based Two-level Routing Protocol for Wireless Mobile Ad Hoc Networks: A Fully Reactive Approach”. IWCMC,August 6-8. 2008 [15] Joa-Ng , M., Lu, I.T., 1999. “A Peer-to-Peer ZoneBased Two-Levl Link State Routing for Mobile Ad Hoc Networks”. IEEE Journal on Selected Areas in Communication, 17(8):1415-1425.

(IJEECS) International Journal of Electrical, Electronics and Computer Systems. Vol: 13 Issue: 1, March 2013 AUTHORS PROFILE Amina MERBAH Currently a PhD student ,University Hassan II /ENSEM Casablanca Morroco. Received a master’s degree in 2010 at University of Sciences Semlalia–Cadi Ayyad, Marrakech, Morroco. Research: Reliability in Wirelss Sensor Networks. Ahmed KAMIL Currently a PhD student, University Hassan II /ENSEM Casablanca Morroco. Received his engineering degree from The National School of Mineral Industry, Rabat. Morroco. Research :Domain of Wireless sensor networks . Hicham BELHADAOUI Currently working as a Assistant Professor in University Hassan II /ESTC, Casablanca Morroco. Received his Phd degree at National Polytechnic Institute of Lorraine/France. Research : Reliability, Automatic Signal Processing and Computer Engineering. Mohamed OUZZIF Presently working as a Professor Ability in University Hassan II /ESTC. Received his Phd degree from University Hassan II /ENSEM, Casablanca Morroco. Research : Distributed Systems and Computer Engineering .