Routing performance in ad hoc networks.
An ad hoc network is a multi-hop wireless network in which mobile nodes communicate over a shared wireless channel. The network is formed cooperatively without specific user administration or configuration and is characterised by a distributed network management system and the absence of a wired backbone. Military, law enforcement, and disaster relief operations are often carried out in situations with no pre-existing network infrastructure and can benefit from such networks because base stations, which are single points of failure, are undesirable from a reliability standpoint. The rising popularity of mobile computing has also created a potentially large commercial market for multimedia applications applied over wireless ad hoc networks. This dissertation focuses on the routing aspects of ad hoc networking. The multi-hop routes between nodes constantly change as the mobile nodes migrate. Ad hoc network routing algorithms must therefore adapt to the dynamic and unpredictable topology changes, the random radio propagation conditions and portable power sources. Various routing protocols have been proposed in the literature for ad hoc networks. These protocols together with comparative simulations are discussed and a new protocol based on load balancing and signal quality determination is proposed . and the simulation results are presented. Currently the proposed routing protocols are compared using simulation packages which are often time consuming. This dissertation proposes a mathematical model for evaluating the routing protocols and the resultant end-to-end blocking probabilities. The mathematical model is based on a derivation of the reduced load approximation for analysing networks modelled as loss networks and the evaluation incorporates and adapts models that have been used for the analysis of cellular Code Division Multiple Access (CDMA) systems. While analytical methods of solving blocking probability can potentially generate results orders of magnitude faster than simulation, they are more importantly essential to network sensitivity analysis, design and optimisation.