Broadband beamforming for joint interference cancellation and low-complexity turbo equalization

We propose a low-complexity turbo space-time equalizer for signals transmitted over broadband wireless channels. The transmitter employs trellis coded modulation (TCM) for its bandwidth and power efficiency, and the broadband transmission is severely impaired by both intersymbol interference (ISI), due to multipath dispersion of the wireless channel, and cochannel interference (CCI), due to the presence of adjacent users. The proposed turbo equalizer consists of a broadband beamformer, which performs soft interference cancellation and space-time minimum mean-square error (MMSE) equalization, followed by a soft-in, soft-out (SISO) TCM decoder. A priori expected symbol values are used as soft inputs in the soft interference cancellation stage and then, after beamforming, the elements of the output sequence are mapped onto extrinsic log-likelihoods or probabilities using a Gaussian assumption. Because these two operations make the beamformer a SISO processor, the front-end of the space-time receiver is suitable for iterative processing. The proposed structure is implemented with small antenna arrays and simulated for turbo equalization of 8-PSK TCM signals, and simulation results indicate a relatively close performance to "no interference" bounds after only a small number of turbo iterations.