Joint interference suppression and power allocation techniques for multiuser multiantenna relay broadcast systems

Abstract

In this work, we study the downlink of multiuser multiantenna systems which employ one multiple-input multipleoutput (MIMO) relay and the amplify-and-forward (AF) relay strategy. We propose joint constrained optimization frameworks that consider the allocation of power levels between the base station (BS) and the relay station (RS) subject to a total power constraint and the joint design of BS precoders, relay beamformers and mobile station (MS) receivers. In terms of the perfect channel state information (CSI), we derive constrained minimum mean square error (MMSE) expressions for the BS precoding and RS beamforming matrices that determine the optimal power levels across the BS and the RS, and the MS receiver matrices. Simulation results show that the performance of the proposed techniques is significantly better than prior art in terms of perfect CSI.

A Unified Approach to Optimal Transceiver Design for Non-Regenerative MIMO Relaying

“…3) BC (1S-1R-MD): The MMSE design for BC was considered in [19] with single-antenna users, and in [20], [23], [25], [27] with multi-antenna users. The latter case usually requires complex algorithms that iterate multiple times through the precoder, the relaying matrix and every equalizers.…”

“…VII-A, different initial points almost lead to the same performance. Algorithm 2 is different from a conventional alternating approach widely used to solve similar problems [19], [20], [27], [50]. The latter needs to alternate between all three matrices B, F and Q.…”

Linear transceiver design for relay-assisted broadcast systems with diagonal scaling

“…These procedures are summarized in Algorithm 1. Algorithm 1 is different from a conventional iterative approach widely used to solve similar problems [6,7,16]. The latter needs to iterate through all three matrices B, F and Q: if F and Q are fixed, B is updated as the optimal precoder for the equivalent multipleinput multiple-output (MIMO) channel [15]; if B and Q are held constant, F is optimized according to (11); if B and F are preserved, each diagonal entry of Q is the optimal MMSE receiver for the scalar channel of the corresponding user.…”

“…This is because the mobile users are not collocated and hence cannot jointly process their received signals. In general, the problems of transceiver design for relay-assisted BC fall into two categories: minimizing the weighted sum power of the base station (BS) and the relay subject to quality of service (QoS) constraints [4], or optimizing a selected performance criterion such as the sum rate or mean square error (MSE) subject to power constraints [5][6][7][8][9][10][11]. The maximum sum-rate design was studied in [5] with special structures, such as zero forcing (ZF), dirty paper coding (DPC) and QR transceiver, and in [8] using quadratic programming.…”

“…The maximum sum-rate design was studied in [5] with special structures, such as zero forcing (ZF), dirty paper coding (DPC) and QR transceiver, and in [8] using quadratic programming. The minimum mean square error (MMSE) design was considered in [6] with single-antenna users, and in [7,9,11] with multi-antenna users. The latter case usually requires This work was funded by a CRD grand from NSERC of Canada with industrial support from InterDigital, Inc. complicated algorithms that iterate through the precoder, the relaying matrix and every equalizers multiple times.…”