Equalization and interference cancellation in linear multiuser systems based on second-order statistics

Potential applications of blind channel identification and equalization in data communication systems have recently been explored. For multiuser systems that are irreducible and column-reduced, second-order statistical methods normally can identify channel dynamics up to a unitary mixing matrix. Additional user separation (equalization) can rely on higher order statistics and other prior information. In this paper, we investigate the equalizability of user signals and the cancellation of unwanted interfering signals based only on second-order output statistics. We show that a user channel can be equalized if it has the longest memory. Furthermore, interfering user signals can be cancelled under a more relaxed multiuser channel condition.