A Load Balancing Algorithm Using Radio Signal Strength and Sensitivity for the Multipath Source Routing Protocol

On-demand reactive routing protocols, such as DSR, are commonly used for MANETs. Multipath Source Routing (MSR) provides DSR with multiple paths to a single destination which o#ers more network resource utilization. Applying a weighted load balancing algorithm in MSR improves the performance of routing in MANET by reducing the end-to-end delay since it depends on the round trip time (RTT) delay as a factor to distribute tra#c on available routes. However, this solution does not account for signal strength which is a major cause of packet errors in wireless networks. This thesis proposes a new route weighing mechanism that makes the MSR more adaptive to network errors resulting from loss of signal strength. The new strategy is to collect network information regarding the signal strength and the nodes' sensitivity to form a parameter called Received Power and Sensitivity (RPS), and use it as a weighing parameter with two flavors. The first evaluates the minimum RPS value among the legs between hops in each route. The second collects RPS cumulatively from all the legs in each route. Simulation results show that the new mechanism reduces the packet drop rate, and enhances the throughput with minimal additional overhead