Adaptive Random Linear Network Coding with Controlled Forwarding for Wireless Broadcast
Carleton University, Ontario, Canada. December 2010.

Multicasting and broadcasting are important communication techniques in wireless adhoc networks. Recently, Network Coding (NC), which has emerged as a promising technique for various applications, has been applied to multicast and broadcast in wireless adhoc networks. It is however observed that the performance using NC is strongly dependent upon the topology, node density and the kind of coding algorithm. The algorithms that are proposed are mostly dealing with single source multicasting or broadcasting.

In this thesis I propose an adaptive multi-source broadcasting protocol using Random Linear Network Coding (RLNC). The key features of this protocol include its multi-source operation, cross-session generations, controlling the number of re-transmissions effectively based on neighbourhood information and earlier decoding. Our simulations with and without cross-session generations show that cross-session generations result in improved Packet Delivery Ratio as well as lower latency. We also investigate its adaptive performance compared to packet forwarding schemes, including a simple flooding protocol, a probabilistic flooding protocol, BCAST and Simplified Multicast Forwarding. We observe the steady performance of our protocol under different node densities and rates.