Optimization Of Dynamic Source Routing In Ad Hoc Wireless Networks

Ad hoc wireless networks are autonomous systems composed of mobile hosts that are free to move around arbitrarily. Rather than relying on a network infrastructure to perform routing, in Ad hoc wireless networks each mobile host serves as a router to forward packets originated from other nodes. The decentralized nature of Ad hoc wireless networks makes them suitable for a variety of applications where central nodes can’t be relied on. Minimal configuration and quick deployment make Ad hoc wireless networks suitable for emergency situations like natural disasters or military conflicts. The presence of dynamic and adaptive routing protocols enables Ad hoc networks to be formed quickly.Proactive protocols are Table-Driven protocols that find routes between all source-destination pairs regardless of the actual need for such routes. On the other hand, reactive protocols are based on the reduction of the routing load by initiating new routing activities only in the presence of data packets in need of a route. The reactive protocols are also known as On-Demand protocols. One of the prominent reactive routing protocols for Ad hoc wireless networks is the Dynamic Source routing (DSR) protocol. DSR floods route requests when the route is needed. DSR is appealing for Ad hoc wireless networks as a reactive routing protocol because of the low routing overhead. However, its performance is not ideal for the high dynamic Ad hoc networks.In this thesis, I will mainly discuss the DSR protocol as an important role in Ad hoc wireless networks. As mentioned, DSR is a kind of on-demand routing protocols based on source routing and route discovery. It’s comparatively excellent in ad hoc wireless networks routing protocols, but it suffers from some shortcomings. This work deals with providing an efficient routing among the ad hoc network’s nodes. A spanning tree will be used from each node to all the other nodes with the transitive transmission range of the source node. The proposed protocol utilizes the tree structure instead of the list structure for the following reasons. First, the amount of effort to find a path between any two nodes depends on its popularity. Second, the length of the routing path between any two nodes improves with its utilization. It is always between the length of the path that includes the common ancestor of both nodes in the tree and the shortest path between these two nodes in the network. If a path is used often enough, it will eventually converge to the shortest path. Finally, the proposed protocol, using the tree, avoids flooding the entire network.The proposed optimized DSR using spanning tree is simulated. The proposed algorithm is compared with the standard DSR. Simulation results show that the proposed optimized DSR outperforms the standard DSR.