Abstract

Relay network was introduced to realize high-data rate transmission and high capacity with limited transmit power. However, channel state information (CSI) is needed due to the requirement from coherent data detection and the self-data removal at terminals in the two-way relay networks (TWRN). Traditional linear probing techniques are able to acquire the accurate CSI by using enough training sequence. However, they may lead to low bandwidth efficiency due to that the implicit assumption of a rich underlying multipath. However, in many cases, the multipath channel has a sparse structure. Unlike the previous methods, we propose a compressed channel estimation method which exploit the sparse structure in multipath TWRN and hence provide significant improvements in MSE performance when compared with conventional LS-based linear channel probing strategies in the TWRN. Simulation results confirm performance of the proposed method.

Key words: Two-way relay networks (TWRN), channel state information (CSI), compressive sensing, sparse multipath channel, compressed channel estimation.