Permutation Optimization in QRD Based Multi-Relay Systems

QR decomposition (QRD) based relay protocols can achieve both the distributed array gain and the intra-node array gain in multi-relay systems. Because the QRD is not unique if column permutation is utilized, the traditional QRD based relay protocol can be further improved. Row permutation of the forward channel and column permutation of the backward channel are proposed in this paper. Exhaustive and iterative search algorithms are presented to solve the permutation optimization problem. The iterative search algorithm with modified QRD is a good trade-off between capacity performance and complexity.