Interference cancellation with Bayesian channel models and application to TDOA/IPDL mobile positioning

Abstract

In wireless communications, dealing with multiple access interference is a common problem. Interference, like the signal of interest, is characterized by a transmitted signal and a channel response. In recent years, improved channel estimation techniques have been proposed based on Bayesian channel estimation. The exploitation of priori knowledge such as the power delay profile allows to reduce the effective number of channel parameters. Whereas the direct use of the Bayesian channel estimate for the user of interest may lead to bias problems, its use for an interfering signal allows to better suppress interference. In this paper we introduce the concept of Component-wise Conditionally Unbiased (CCU-) LMMSE channel estimation and consider its application to the problem of TDOA-based mobile positioning in 3G systems exploiting IPDL frames.