A study of algorithms for single-user and multiuser detection in wireless communication systems

This dissertation is focused on developing new algorithms for single-user and multiuser signal detection in wireless communications. Low complexity algorithms are of particular interest. Some focus is given to cases with multiple geometrically dispersed receivers where only a limited amount of data from each receiver will be sent to a central location called the fusion center.; First, we demonstrate the amount of information that should be transmitted from each receiver to the fusion center is limited by the amount of information transmitted from the other receivers. These results provide important guidelines for the optimum way to allocate a fixed overall number of bits across receivers.; Second, we analyze the performance of distributed receiver networks employing multiuser detection techniques. We show that distributed systems achieve near centralized performance using only one or two bit preliminary decisions. Moreover, distributed systems can be helpful in lessening the severity of the near-far problem when full multiuser detection is impractical.; Next, we propose a new iterative multiuser detection algorithm, the gradient guided search algorithm. This algorithm achieves near-optimum performance while its implementation complexity is linear in the number of active users. In addition, we find that the multiuser decorrelating detector is a good choice for the initialization of the proposed algorithm.; Finally, we present an approximate MMSE multiuser detection algorithm for DS-CDMA communications in slowly varying, flat Rayleigh fading channels. The performance of the approximate MMSE algorithm is shown to be very close to that of the exact MMSE detector so that it can be used as a tool for the performance analysis of the exact MMSE detector. The approximate algorithm also provides considerable complexity reduction.