Throughput maximization in Wireless Networks by Scheduling End-to-End Flows

Due to the hard bandwidth-constraints in wireless ad hoc networks, maximization of achievable throughput is the most important way of enhancing system performance. Existing works have attempted to maximize system throughput by increasing the number of concurrent packet transmissions within the network using multiple radio channels or by using power control techniques. The approach taken in this work is based on the fact that the overall system throughput in an ad hoc network comprising multiple multi-hop flows (IP data paths) varies considerably with the direction and number of hops involved in the flows. When multiple flows cross each other, or are close enough, then packets belonging to those flows contend with each other for channel access, thereby reducing the overall system throughput. If these flows are properly scheduled based on the number of hops involved or their direction of flow, the overall system throughput can be significantly increased, In this work, certain scheduling strategies are proposed and evaluated that prioritize the different flows in the network and aims to improve the overall system throughput. Simulation results indicate that significant improvement can be achieved with the proposed scheduling methods.