SAGE-based estimation algorithms for time-varying channels in Amplify-and-forward cooperative networks

Cooperative communication is a technique that achieves spatial diversity by exploiting the presence of other nodes in the network. Most analyses of such networks are conducted under the simplifying assumption of perfect channel knowledge. In this paper we focus on the popular Amplify-and-forward (AF) cooperative protocol. We propose several SAGE-based iterative algorithms with different complexities for estimating the channel gain and noise variance in the case of time-varying channels. Computer simulations are provided to evaluate their performance over Rice-fading channels. We point out a low-complexity estimation algorithm yielding an error performance that (for Rayleigh fading) is only about 0.5 dB worse than in the case of perfect estimation, while outperforming a pilot-based estimation algorithm by about 1.5 dB.