Collaborative communications: Joint relay and protocol selection

We consider a decode and forward collaborative network and investigate the relay selection problem. Users are able to collaborate with each other; decode messages of each other, re-encode and forward along with their own messages to the destination (collaborative protocol). We study the performance obtained from collaboration in terms of 1) increasing the achievable rate, 2) saving the transmit energy, and 3) reducing the resource requirement. To ensure fairness among the users, we assume that the transmit energy to the rate ratio is fixed for all users. We allocate resource optimally for the proposed collaborative protocol (CP) with the imposed fairness constraint, and compare the result with the non-collaborative protocol (NCP) where users transmits their messages directly to the destination. We show that a considerable gain can be obtained if the direct source-destination channel gain is significantly smaller than those of alternative involved links. We demonstrate that a rate and energy improvement of up to (1 + etaradic(k/k+1))^eta can be obtained, where eta is the environment path loss exponent and k is the ratio of the rates of involved users. The maximum gain can be obtained for low transmit-energy-to-received-noise-ratio (TERN). In contrast, in a high TERN environment the NCP is preferred.