Channel training and decoding for MIMO relay networks

Abstract

To accomplish coherent relaying schemes in cooperative relay networks, accurate channel state information (CSI) is essential. To get such CSI, channel training is employed in practice. In this thesis project, we perform theoretical analysis on channel training design and training-based decoding for multiple-input-multiple-output (MIMO) relay networks, which is in general very challenging for relay networks. The objective of channel training for MIMO relay networks is to obtain global CSI at the receiver. To perform training, training scheme design, training code design, training time design, and power allocation, are discussed respectively. Employing obtained channel estimations, two coherent training-based decodings are studied for distributed space-time coding (DSTC) MIMO relay networks: mismatched decoding and matched decoding. For both decodings, the diversity and complexity performance are investigated and compared with each other.