Abstract
This thesis addresses the minimum transmission broadcast problem for the many-to-all scenario in wireless sensor networks and presents how a network-coding broadcast protocol with priority-based deadlock prevention can reduce the number of transmissions effectively. Our main contributions are as follows: First, we formulate the MTB problem for the many-to-all with network coding scenario as an integer linear programming (ILP) problem. The solutions obtained by ILP can serve as a lower bound for any protocol. Second, we propose a distributed network-coding broadcast protocol, which constructs efficient broadcast trees and dictates nodes to transmit packets in a network coding manner according to the constructed trees. In addition, we present the priority-based deadlock prevention mechanism to avoid deadlocks. Simulation results confirm that compared with existing protocols in the literature and the performance bounds obtained by our ILP technique, our proposed network-coding broadcast protocol performs very well in terms of the number of transmissions needed.