Asymptotic Achievable Rate Analysis for Selection Strategies in Amplify-and-Forward MIMO Two-Hop Networks With Feedback

This paper investigates relay-selection strategies in multiple-input-multiple-output (MIMO) relay networks, where joint source and relay precoding are utilized, and a single MIMO relay is adaptively selected for forwarding signals during each transmission period. Assuming that an optimal relay-selection scheme is implemented with full channel-information feedback, we study the optimal selection performance by characterizing its achievable rate with two derived bounds. Based on the rate bounds, we show that the corresponding achievable rate scales as (M/2) log log K for a fixed number of antennas M and an increasing number of relay nodes K -> infinity. Furthermore, for feedback reduction, we propose a relay-selection scheme with limited feedback while still achieving the optimal achievable rate-scaling law for a large K. Numerical results verify the effectiveness of the limited feedback-selection strategy under practical values of relay number K.