Optimization of Split-And-Combine Relaying

Relays play an important role for increasing rate and reducing energy consumption of wireless networks. In this paper we consider a three-node network (source, relay, destination) in which we want to minimize total energy consumption for a given transmission rate. We analyze and optimize the Split-Combine-Relaying (SCR) protocol that for many typical parameter settings performs better than traditional decode-and- forward. SCR splits a data packet into two fragments, which are then transmitted in two phases. In the first phase of transmission, the source sends the first fragment to the relay. In the second phase of the transmission, the source sends the second fragment to the destination, while, at the same time, the relay forwards the first fragment to the destination. We provide an optimization framework to decide the amount of data in each fragment, the amount of time spent in each of the phases, and the corresponding transmission powers. We also show that an extension of SCR that employs Slepian-Wolf coding of the fragments leads to further reduction of energy consumption.