Summary
International Symposium on Antennas and Propagation
2010
Session Number:3TF1
Session:
Number:3TF1-5
Optimum Decision Rule for Fixed Relay Pairing Selection Application
John F. An, Lider Pan,
pp.-
Publication Date:2010/11/23
Online ISSN:2188-5079
Summary:
Relay technologies have been recently studied and creatively incorporated in the nextgeneration mobile broadband communications systems (4G;WiMAX IEEE802.16j/m and LTEAdvanced standards) where the multi-hop radio link is to improve the transmission throughput and reduce the outage probability, especially the mobile station in the brink of cell-edge and shadowing area [1, 2, 3]. The relay selection presented in [4, 5, 6] is primarily based on the joint selection criteria of instantaneous signal-to-noise power ratio across the hops for the relay links. However, for fixed relay application, the radio links between relay nodes (R) and destination (D) of the second hop are assumed to be static fading channels which have deterministic channel statistics and same mean SNR value with amplify-and-forward (AF) protocol at relay nodes. In practice, this is true when R location is geographically fixed with a directional antenna to D (i.e. stationary sites; basestation) that also potentially reduces the transmission interference level. In the contrary, the links between source (S) and R in the first hop are usually treated as a stochastic fading channel (i.e. time-variant Rayleigh fading channel) due to its random walk mobility and unlimited extent rage from S (i.e. mobile unit). In our paper, a simple and effective relay selection method is explored for the two-hop fixed multi-relay wireless system using multiple-likelihood decision concepts. Fig.1 shows our proposed fixed M-relay network topology with Bayes optimum decision algorithm. We assume independent and identically distributed (i.i.d.) Rayleigh fadings are across all S-R and R-D relay hopes, as well as the channel gain form S-D direct link is weak enough to be negligible (no diversity gain). Meanwhile, perfect channel state information (CSI) is assumed for the channel statistics input to the optimum decision algorithm at the D. Hence, our relay selection is primarily dependent on the maximum channel power gain distribution of the first relay hop from the multiple relaying links, not on the comparative instantaneous signal-to-noise power from end-to-end link. Finally, the maximum channel power gain distribution and average Bit-Error-Rate (BER) performance are simulated against various Doppler effects on the first hop channels to validate our optimum selection criterion.