Block transmission equalizers using constrained minimum variance filters with application to MC-CDM

Equalizers for general block transmissions are derived based on two concepts. Firstly, filters are designed to minimize the noise variance constrained to deliver unbiased estimates; this enables a good performance with arbitrary quadrature amplitude modulation (QAM). Secondly, for derivation of decision feedback equalizers, we propose the principle of modeling the uncertainty of subtracting hypothetical interference, derived from previous estimates, as a virtual noise. The equalization algorithm is described by an iterative procedure, covering both the linear equalizer (without iterations) and decision feedback equalizers. Different variants for decision feedback are presented, including hard interference cancellation and different types of soft interference cancellation. For the special case of an orthogonally spread transmission over Rayleigh fading channels, e.g., multi-carrier code division multiplexing (MC-CDM), an efficient implementation using fast transforms is described. The performance of the proposed equalizers is assessed by computer simulation