A Modified Superimposed Training Scheme for Individual Channel Estimation for Amplify-and-Forward Relay Network

Abstract

In this letter, we deal with the problem of individual channel estimation in amplify-and-forward (AF) relaying systems. A novel superimposed training (ST) scheme is proposed where the relay superimposes its own training sequence directly on top of the received data signal without bandwidth expansion. As a result, the training sequences from the source and relay nodes are independent of each other and can be viewed as a time-multiplexed (TM) mode in the proposed scheme, thus making it more flexible and robust in relay-training design. To remove the data-induced interference and relaying-propagated noise during channel estimation, a modified ST scheme is designed by discarding some relaying data to accommodate the relay-training sequence. Simulation results are presented to assess the performances of the proposed scheme and to obtain the optimal power allocation.