SUMMARY
In recent years, combining multiuser detection and intelligence computer scheme have received considerable attention. In this paper, adaptive fuzzy‐inference multistage matrix wiener filtering (FI‐MMWF) techniques, based on the minimum mean‐square error criterion, are proposed for ultra‐wideband (UWB) impulse radio communication systems. These FI‐MMWF‐based algorithms employ a time‐varying fuzzy‐inference‐controlled filter stage. Consequently, the proposed approaches accomplish a substantial saving in complexity without trading off the system performance and dynamic‐tracking characteristic. In addition, the fuzzy‐logic‐controlled matrix conjugate gradient algorithm is adopted to reduce the system complexity without trading off the bit‐error‐rate (BER). Simulations are conducted to evaluate the convergence and tracking behavior of the proposed MMWF algorithm, and the BER of the time‐hopping‐UWB system in a realistic UWB channel is investigated. Copyright © 2011 John Wiley & Sons, Ltd.