Relay assignment in Decode-and-Forward cooperative networks based on order-statistics

This paper offers the statistical analysis of relay assignment in cooperative Decode-and-Forward (DF) networks based on order statistics. By assuming a set of N independent Rayleigh fading channels, first the channel with r-th biggest SNR is analyzed. The diversity order and average error probability for this channel are calculated. Then it is assumed that there are two clusters of terminals and the terminals in each cluster are relatively close to each other, so they have equivalent average SNR to the destination. The cluster with better average SNR is assumed to be the relay cluster. The problem is to find best pairing of source-relays. By assuming the SNR of direct channels to be i.i.d. random variables, it's more reasonable to work with their order statistics, because order statistics reveal the relative SNR of the channels. This is the basic distinction of this paper from similar papers. Using this approach, two different scenarios are proposed to assign the best relay to each weak-channeled terminal based on the instantaneous SNR information. In the first scenario, only the SNR of direct links between the terminals and the destination are known; whereas in the second scenario we assume that the SNR of all channels are available. It's shown that in the second scenario, a much higher performance can be achieved through cooperation.