Low Complexity Selection Cooperation Techniques Using Information Accumulation in Dual-Hop Relaying Networks

Three low complexity, single-parameter, selection cooperation protocols, called P-n, P-γ, and P-t, are introduced for dual-hop relaying networks. The protocols are based on information accumulation, and can potentially be implemented using rateless codes. The expected transmission time is analyzed for the three protocols for block fading channels. As a baseline for performance comparison, a rate optimal protocol, called P-o, is proposed. In the single-relay case, all the protocols, if used with their optimized parameter which is trivially obtained, coincide with P-o. Large signal-to-noise ratio (SNR) approximations to the optimal parameters in the multirelay case, which have good accuracy in large SNR regimes, and satisfactory accuracy for smaller values of SNR, are derived. The dependence of the optimal parameters on the network parameters is numerically studied for Rayleigh fading. The average rate and average source transmission time (ASTT) of the schemes are compared with one another and other comparable previous relaying schemes through numerical examples. It is observed that the suboptimal schemes exhibit near-optimal rate and ASTT performances. As a general rule, P-o and P-γ have the best performances. The P-n scheme has a larger rate, but a larger or smaller ASTT, compared to P-t. Increasing the number of relays ultimately causes diminishing returns in the average rate and ASTT.