Kalman filter-based channel estimation for amplify and forward relay communications

We propose an autoregressive model for the combined amplify and forward time-varying relay channel and derive a causal iterative channel estimation method using Kalman filter. This formulation enables us to study and compare two widely-used pilot transmission strategies in terms of the channel estimation errors. We provide a single-letter formula for the power allocation between the source and relay to achieve near optimal bit error rate (BER) performance in dual-hop communications. For cooperative communications, we show that the relay speed has a significant impact on the BER performance, and hence is important to be considered in practical system design.