Optimum Receivers Under Complexity Constraints for Radio CDMA Communications

点击翻译按钮获取中文摘要

9406069 Varanasi Several civilian nondirectional communication applications such as mobile radio telephony and personal communication services require bandwidth-efficient multiple-access radio communication as well as a tremendous flexibility in system design. The dynamic channel sharing strategy of Code-Division Multiple-Access (CDMA) is particularly well-suited and its feasibility for such applications and is currently being actively investigated. This project will advance the theory of optimum receivers under complexity constraints for CDMA communication in fading multipath radio channels through research on several aspects of this broad subject. The asymptotic performance measures of near-far resistance and fading resistance will be defined in order to succinctly characterize the effects of estimable and inestimable interference in multiuser fading channels. The design and analysis of optimum linear detection rules for slowly fading and time-varying but memoryless multipath fading CDMA channels is proposed. Linear adaptive receivers that are near-far and fading resistant will be developed for dynamically time-varying multipath fading channels with memory. The concept of group detection will be studied in the context of Gaussian as well as time-varying multipath fading CDMA channels in order to address the need for strategies with complexities ranging between those of linear and optimal detectors that bridge the wide gap in performance between them in applications requiring bandwidth efficient communication. Receiver optimality is also addressed through the design of signature waveforms that optimize the performance of fading channel detectors under bandwidth constraints. The results of this research effort will yield optimum centralized CDMA receivers under various complexity constraints that are well-suited for implementation at base stations and that will serve as ideals in both function and performance to be emulated by decentralized signal reception schemes constrained by even lower complexities. ***