Joint channel estimation in asynchronous amplify-and-forward relay networks based on OFDM signaling

In this paper, we propose a method on the training based on channel estimation issues for relay networks that employ amplify-and-forward (AF) transmission scheme. We have to point out that jointly estimating the channel from source to relay and relay to destination suffers from many drawbacks in fast fading case because the estimation of previous pilots is not suitable for current channel. In this paper we consider a new joint estimation of overall channel impulse response (CIR) using one OFDM signal without pilots. Using the maximum likelihood (ML) function, we derive a channel estimator by taking the frequency domain of transmitted signal as Gaussian and averaging the ML function over the resulting Gaussian distribution. Simulation results show that our proposed channel estimator performs a fraction of 1dB compared in high SNR region.