Joint Design of Transceivers for Multiple Access Channels using MMSE Decision Feedback Detection

Abstract

In this thesis, we consider the joint design of transceivers for a multiple access Multiple Input and Multiple Output (MIMO) system having Inter-Symbol Interference (ISI) channels. The system we consider is equipped with the Minimum Mean Square Error (MMSE) Decision-Feedback (DF) detector. Traditionally, transmitter designs for this system have been based on constraints of either the transmission power or the signal-to-interference-and-noise ratio (SINR) for each user. Here, we explore a novel perspective and examine a transceiver design which is under a fixed sum Gaussian mutual information constraint and minimizes the arithmetic mean square error of the MMSE-decision feedback detection. For this optimization problem, a closed-form solution is obtained. We prove that the optimal solution is achieved if and only if the sum mutual information is uniformly distributed over each individual user per the number of its active subchannels; i.e., user mutual information uniform distribution. Meanwhile, the Gaussian mutual information of the current user under perfect feedback for all the previous users is uniformly distributed over each individual symbol within the block signal of the user; i.e., symbol mutual information uniform distribution. The user mutual information uniform distribution is attained by successively solving a series of inverse (dual) problems of maximizing single user throughput, while the symbol mutual information uniform distribution is maintained by using the equal diagonal QRS decomposition. We also show that such uniform decomposition, in addition to minimizing the arithmetic MSE of MMSE-decision feedback detection, also has another two optimality properties: (a) Both the optimal user-detection order and symbol-detection order are natural orders in terms of signal to interference and noise ratios. (b) The free-distance for the Maximum Likelihood (ML) detector has an asymptotic behavior when the sum Gaussian mutual information tends to large.