Abstract:
Networks formed by potentially mobile nodes, with out any support of existing infrastructure are called Mobile Adhoc Networks. These networks can be used in real word for search and rescue operations, military environments, conferences, etc. The architecture of mobile adhoc networks gives the advantage of increased speed of deployment and decreased dependency on infrastructure. Regardless of the benefits, these networks have serious limitations about bandwidth and security. Other major issues include the mobility management and routing of data packets within the adhoc network. Routing protocols for mobile adhoc networks are different than protocols for conventional networks, due to increased mobility and rapid topology change, Different international organizations and universities have proposed many protocols, but no single protocol can handle_ all the r~al world network scenarios. This masters thesis evaluates some of the proposed protocols. The evaluation is done using Network Simulator 2 from Berkeley. The simulations prove the fact that no single protocol can successfully address all the networking problems . in adhoc networks. Variant of distance vector routing algorithm like AODV do perform good in most of simulations but the jitter performance and the control over head do not match the performance of link state routing algorithm like OLSR. It is also evident that the routing layer alone cannot tackle all the problems, and support from link layer is required. MPR (Multipoint Relay) based Hybrid Routing Protocol (MBHR), is an enhancement over Optimized Link State Routing and Adhoc On-demand Distance Vector Routing. It combines both .protocols and adds certain optimizations to further enhance the performance. A hybrid approach, which incorporates both reactive and proactive route acquisition process, can give better results as compared to existing routing protocols.
