Optimal diversity allocation in coherent multiuser systems

Abstract

The ultimate goal of universal wireless communications on demand can only be met if a number of key issues are addressed. Foremost among these are how to deal with the multipath fading inherent to the wireless communications environment and how to allocate bandwidth to a multitude of users. The solution to the multipath fading problem is to introduce known redundancy into the data stream in the form of error control coding. The paper shows how the proper assumption on the inner code (a repetition code in the guise of direct-sequence spreading that is soft-decision decoded) allows the system to be flexible enough to encompass many of the viable wireless system implementations. A system theoretically equivalent to a coded multicarrier system arises out of the coding problem discussed. When the soft-decision inner repetition code is employed, the multicarrier system is then seen to encompass both FH/CDMA and DS/CDMA as well. A concatenated coding formulation is introduced and the throughput measure is formally defined. The binary coherent modulation scheme investigated is introduced along with the system assumptions considered. Throughput optimization over the rate of the inner code and the number of users in the system is then performed. Following this general optimization, the paper demonstrates that when the number of users in the system is fixed, throughput optimization is equivalent to minimization of the probability of error. System optimization in this special case is then considered.