An Iterative (Turbo) Channel Estimation and Symbol Detection Technique for Doubly Selective Channels

Reliable communication over a doubly selective wireless channel is a challenging problem, mainly due to poor estimation of the channel parameters in the presence of limited training. We therefore propose turbo equalization in conjunction with Doppler domain channel estimation techniques for communication over a channel that is selective over both the time and frequency dimensions. It is shown that joint techniques to address the problems of channel estimation and symbol detection along with iterative equalization and decoding constitute a powerful technique with overwhelming performance gains over separate channel estimation and decoding in doubly selective environments. Such techniques can provide the flexibility required to improve the estimate with limited training.