Ad Hoc On Demand Distance Vector

Published Date: 02 Nov 2017

Keywords: Mobile Ad hoc Network (MANET), Dynamic Source Routing (DSR), Ad hoc On-Demand Distance Vector (AODV),Energy, Routing Protocols.

I. INTRODUCTION

According to dynamic topology of Ad hoc networks, routing and communication between the nodes in these networks have been challenging missions. To overcome this challenge, many protocols for routing in MANET have been presented to the Internet Engineering Task Force (IETF). There is a classification for these protocols as table-driven (proactive) and source initiated (routing) protocols. In proactive routing protocols, routing tables are used to keep route information from each source to every destination in network before this route is needed. On the other hand, in reactive routing protocols, a source sends a route discovery through the network, only when the route is required [5]. Dynamic Source Routing (DSR) and Ad hoc On-Demand Distance Vector (AODV) are two types of reactive protocols.

The limited battery resources consumed by nodes in a MANET must be considered as a limited resource in using a routing protocol. Moreover, designing a new routing protocol that suited for WSN based on available Ad hoc routing protocol is still a challenging issue for researchers. Power limitations in node joint by a deployment of large number of nodes have created these challenges to design a new routing protocol for WSN [2]. Thus, we need to have a protocol which can function effectively in energy consumption. According to [16] reactive routing protocols such as AODV and DSR, consume less energy than proactive routing protocols because they use route discovery only when the route is needed. Many studies have been done in performance comparison of routing protocols of MANET in terms of throughput, delay, and packet delivery ratio. There are only a few studies that evaluated the energy consumption of reactive routing protocols. Moreover, the result It�s of these studies show that the energy consumption of reactive routing is strongly related to parameters that are used in simulation. In this paper, the researcher investigates the evaluation of energy consumption of these two routing protocols in terms of routing energy consumption and average energy consumption through detailed simulation on a variety of movement and communication patters by using the Network Simulator 2 (NS2). The rest of this article is organized as follows: Section 2 presents an overview of DSR and AODV protocols. Then some related works in energy optimization and evaluation in MANET routing protocols are discussed in Section 3. The implementation of this research including simulation scenario, mobility patter and energy patter are explained in Section 4. Finally, in Section 5 and Section 6, the results and conclusion of this study are presented.

II. PROTOCOL OVERVIEW

A. DSR (Dynamic Source Routing)

DSR is an efficient routing protocol proposed specially to be used in multi-hop mobile Ad hoc networks. It has two phases including Route Discovery and also Route Maintenance. These two phases help nodes to find and maintain the best source routes to destinations. The Source Routing is a loop-fee routing that intermediate nodes do not need any up-to-date routing information by allowing nodes o cache the routing information for future use [3]. Each node in this protocol controls each packet for source-route information and forward it based on this routing information. If routing information is not found in the packet, it will provide the source route by knowing the route. When the destination is not known, node caches the packet and finds the routing information to the destination by sending route queries to all nearby nodes. Then it sends Route-Replies back to the source [20].

B. AODV (Ad hoc On-Demand Distance Vector)

On the other hand, AODV is another reactive routing protocol with quick adaptation to dynamic link conditions and low network utilization in mobile Ad hoc network. AODV does not discover and maintain any routes until two nodes need to communicate with each other. AODV uses the destination sequence number to guarantee the route freshness and loop freedom of the route [9]. Like DSR, routing method in AODV has two steps: route discovery as well as route maintenance. When a node needs to send data to another node which has no routing information, Route Discovery process is initiated. It broadcast a Route Request (REQ) packet to its neighbors. Each neighboring node either replies the REQ by sending a Route Reply (REP) packet or increasing the Hop-count feld and rebroadcasting the RREQ to its own neighbors. It means that nodes reply to the RREQ by a RREP packet only if they have a fresh route for the desired destination. The source node restarts the discovery process to make a new route to the destination if receives Route Error (RERR) control packet form nodes in case of link breakage [20].

III. RELATED WORK

A. Energy optimization in DSR and AODV

An energy efficient routing protocol decreases the power consumption of the nodes by routing data on paths that consume the least amount of energy. There are some special mechanisms to achieve this goal. Ref [11] used an efficient caching technique for storing information to propose an energy efficient routing protocol. They showed that it has a better performance in terms of energy savings compared to DSR protocol. Moreover, [14] proposed a loop-free energy conserving scheme which tries to decrease routing and storage overhead to provide optimization of resources use in large scale networks. It is based on source routing and named Energy Conserving Dynamic Source Routing (EC-DSR). They also evaluated the performance of this scheme by simulation and showed better results. Furthermore, [19] proposed a comprehensive energy optimized routing algorithm based on AODV protocol. This algorithm was created based on the combination of device runtime battery capacity and the real propagation power loss information. Moreover, [1] proposed algorithm used the AODV routing protocol to select the optimal route based on the basis of the maximum energy of each route. Furthermore, [10] proposed a new routing algorithm based on the energy level of the node. The results showed the advantages of this protocol in terms of energy consumption. In addition, [12] investigated AODV based algorithm with less energy consumption during route founding by establishing routes that are lower congested than the others. Their scheme decreased more than 20 % of total energy consumption. Ref [4] presented new routing protocol EMRP by combining the prediction of the node mobility and residual energy state. According to simulation results, EMRP can increase the lifetime of the network

IV. MATERIALS AND METHODS

A. Simulation Tool

Simulation research tool is being used by the majority of MANET community that estimates how event might occur in the real world. Commonly, this method is used to evaluate the performance of network in terms of different metrics. Discrete Event Simulation is a software-based method to employ the models of real environment to draw a conclusion from the output. Therefore, a Discrete Event Simulator is used in this study. There are many Discrete Event Simulators available for MANET community; but the research shows that Network Simulator (NS2) is the most widely used Discrete Event Network Simulator in the MANET research [15]. The most important reasons for using NS2 are software availability, large community of developer, and also supporting energy model.

B. Simulation parameters

A random traffic patter with a TCP

Table 1: Simulation parameters

Connections between mobile nodes is used in the simulation. The starting time of various connections are generated randomly by the simulation [13]. Other parameters are shown in Table I.

C. Energy Performance Metrics

Routing energy consumption and average energy consumption are energy performance metrics in this study. In routing energy consumption, protocols are evaluated in term of energy consumption only in network layer, the average energy consumption is simply the total consumed energy over the number of nodes. In Ad hoc network, energy consumption is sum of transmit, receive, idle and sleep power in all layers.

V. RESULTS

A. Routing energy consumption

To compare the routing energy consumption, four varying parameters are chosen. These parameters are:

- The traffic patter

- The node's mobility pater

- The mobile nodes number

- The simulation area size

B. Varying Traffic Patterns

In varying traffic patterns, AODV and DSR protocols are evaluated by number of traffic source and rate of source sending. According to Fig. 1, when the number of sources grows, an increase of routing packets can be seen in DSR and AODV protocols. The result shows that the energy consumption of DSR has a slower trend compared to AODV. It shows that when the traffic sources numbers increase from 10 sources to 20 sources, routing energy consumption grows 88% in AODV and 28% in DSR. However, when this factor moves from 20 sources to 30sources, routing energy consumption grows 20% in AODV and 46% in DSR. So, the increase of energy consumption of AODV is more than DSR in low traffic and DSR is more than AODV in high traffic. Source routing characteristic of DSR and caching may be the main reasons for this behavior.

Figure: 1 Routing energy consumption versus number of traffic source

Figure2: Routing consumption versus source sending rate.

C.Varying Area Size

Fig.3 shows that routing energy for RP�s

Figure 3.Routing energy consumption versus area sizes.

Consumption of DSR and the AODV protocols are increased by incrementing the area. Again, in this scenario DSR is more efficient than AODV protocol.

VI. CONCLUSION AND DISCUSSION

Usually pure reactive protocols including AODV and DSR perform better than proactive protocols in mobile Ad hoc network. In this study, an energy performance comparison of DSR and AODV routing protocols for mobile Ad hoc network was presented. DSR and AODV have different routing mechanisms. However, DSR and AODV have the same on-demand behavior. AODV applies routing tables with one route for each destination. In contast, DSR applies route caches and uses source routing without using any periodic transmission. It also uses caching and keeps more than one route for each destination. Type of routing protocol affects the energy consumption due to the different routing overhead used for sending and receiving the routing packets.

AODV is efficient with some mobility scenarios by eliminating source routing overhead of the DSR protocol. But in AODV, discovery route requires more overhead and actually is more expensive than DSR. The overall results shows a better performance of DSR rather than AODV except in static networks while AODV uses hop-by-hop routing and DSR uses source routing with longer header. The reason is that DSR uses caching mechanisms to reduce the discovery routes overhead. It also shows that DSR resulted in the least energy consumption for low density networks and AODV generated higher volume of energy than the DSR in high density networks. They have a similar behavior in static network. The reason for this behavior can be less overhead in DSR due to source routing. Also, the results demonstrate that DSR performs better than AODV in low and high loads. However, AODV is found effective for low loads. Therefore, as an overall conclusion, routing protocols used currently in MANET may require some

effort to minimize the energy cost of interface in the network. A comparison of results of frst section and second section shows that the cost of sending packets in AODV protocol is very significant. So, energy consumption is increased mostly due to the increase in the routing packet

Overhead like RREQ and RREP packets. AODV is more efficient when the cost of application layer and transport layer are added in second section. So, by considering the routing overhead of AODV protocol and reducing the number of control packets, energy consumption can be decreased and the lifetime of the network can be increased.