Abstract:

Ad hoc networks are known by much identification, like a multi-hop wireless connectivity, dynamic wireless and multi hop network, a self-organized networks consisting of a mass of sensor nodes,   each mobile node is responsible to execute a local broadcast to identify its neighbors. Over the last several years, Ad Hoc Networks  provide a real opportunity to design flexible networks, above all, with the appearance of communication interaction are transformed from human-to-machine  to machine-to-machine (M2M), and more specifically things-to-things  (IoT), the frequent change of the network topology, the need for efficient dynamic routing protocols which play an important role in guaranteeing the quality of service of the Ad Hoc network, routing protocol becomes more and more critical. Due to this, several studies focuses their works on Ad Hoc network, especially in wireless applications like sensor networks, which present a challenge in this area of search. In this context, The particularity of this type of network is that each station can communicate with any other node in the network. Indeed, if a node 1 wants to communicate with a node 2 which is not to be radio ported, then it will go through a series of intermediate nodes which will play the role of relay between the source and the destination. But if two stations wish to exchange packets, which may not be able to communicate directly? This situation becomes more complicated if more nodes are added within the network. When transmitting a packet from a source to a destination, it is necessary to use a routing protocol that will correctly route the packet by the "best" path. Several protocols have been proposed at the ad hoc level. In order to understand their behavior in mobile networks, we were therefore interested in carrying out a theoretical study on some protocols, for this, the first step to do was to choose the protocols on which to base ourselves. Our choice fell on AODV and OLSR. Indeed, these protocols show better qualification in last researches [1]. So this paper presents a performance comparison between two various techniques and are the most advanced on the road to standardization [4]. They each belong to a family, the first, reactive, is called AODV (Ad hoc On demand Distance Vector Routing). It uses a broadcast mechanism in the network to discover valid routes. The second is proactive, and is called OLSR (Optimized Link State Routing Protocol). It uses a mechanism which makes it possible to designate a subset of its neighborhood responsible for the dissemination of topology control information in networks at a lower cost [2]. Both protocols were simulated using the ns-2 package and were compared in terms of the packet loss ratio, routing overhead, and average throughput, while varying the number of nodes and speed. The simulation revealed that although OLSR perfectly scales to small networks with low node speeds, AODV is preferred due to its more efficient use of bandwidth.