Adaptive linear-quadratic receivers for time-varying, frequency-selective code-division-multiple-access channels

We discuss a research effort focused on the design and analysis of robust, low-complexity, adaptive wireless receivers that exhibit good performance characteristics in the presence of multiple sources of complex structured interference as well as significant uncertainty regarding the exact structure of that interference. We consider problems primarily related to the code-division, multiple-access (CDMA) environment. In particular, we study adaptive one-shot linear-quadratic (LQ) receivers for time-varying, frequency-selective CDMA fading channels. We propose a novel Bayesian approach to this problem in which receivers are designed based on a probabilistic channel model that explicitly incorporates multiple sources of additive interference as well as a stochastic structure for the channel uncertainty. Under the assumption that the probabilistic structure of the channel model is known, we develop and analyze a design strategy for adaptive LQ receivers that are optimal with respect to the assumed channel model and robust with respect to uncertainty regarding the true instantaneous state of the channel. In addition, we develop and analyze an adaptive modulation scheme that works in conjunction with the proposed LQ receivers to either maximize throughput or minimize probability of error