Relaying energy allocation in training-based amplify and forward relay communications

We consider relay-assisted communication in a training-based transmission scheme. Each transmission block consists of a training phase and a data transmission phase. The relay node employs the amplify-and-forward protocol during all transmissions. We focus on the relay signaling design and investigate the benefit of allowing for different relaying power during the training phase and the data transmission phase. Specifically, the relaying energy allocation between the two phases is optimized for maximizing the average received signal-to-noise ratio at the destination node. We study this optimization problem for both single-antenna relay and multi-antenna relay and derive a simple closed-form relaying energy allocation strategy that achieves near-optimal performance. This closed-form strategy depends only on the length of the data transmission phase but not on other system parameters such as the relaying energy budget, the number of antennas at the relay, and the distances between the source, relay and destination nodes.