Simultaneous estimation of multi-relay MIMO channels

This paper addresses training-based channel estimation in distributed amplify-and-forward (AF) multi-input multi-output (MIMO) multi-relay networks. To reduce channel estimation overhead and delay, a training algorithm that allows for simultaneous estimation of the entire MIMO cooperative network's channel parameters at the destination node is proposed. The exact Cramér-Rao lower bound (CRLB) for the problem is presented in closed-form. Channel estimators that are capable of estimating the overall source-relay-destination channel parameters at the destination are also derived. Numerical results show that while reducing delay, the proposed channel estimators are close to the derived CRLB over a wide range of signal-to-noise ratio values and outperform existing channel estimation methods. Finally, extensive simulations demonstrate that the proposed training method and channel estimators can be effectively deployed in combination with cooperative optimization algorithms to significantly enhance the performance of AF relaying MIMO systems in terms of average-bit-error-rate.