Regularized transceiver designs for multi-user MIMO interference channels

Document Type

Article

Publication Date

8-7-2012

Abstract

For multi-user interference channels (IC), an altruistic approach based on the zero-forcing (ZF) criterion shows the near-optimal performance at high signal-to-noise ratio (SNR), whereas its performance at low SNR becomes poor compared to a simple egoistic algorithm (selfish beamforming). Thus, balancing between the egoism and the altruism has been an important issue to achieve good sum-rate performance at overall SNR regime. In this paper, we propose a new approach for enhancing the performance by regularizing the ZF based transceivers. To this end, we start with investigating efficient ZF transceivers for 2-user and 3-user ICs. First, coordinated spatial multiplexing (CSM) is proposed for 2-user IC. For the 3-user case, it is shown that the enhanced interference alignment (E-IA) introduced in our previous work is the optimal ZF transceivers in terms of the sum-rate performance. Next, to improve the performance of the CSM and E-IA schemes at low SNR, we propose a non-iterative regularization method under the high SNR approximation. The distributed implementation of the proposed regularization method is also presented where each node is able to compute its own precoding or decoding matrix using local channel state information. From simulations, it is observed that the proposed regularized design outperforms the conventional schemes in overall SNR regime. Also, we confirm that our distributed approach provides a substantial performance gain over the conventional distributed scheme with reduced computational complexity. © 2012 IEEE.

Identifier

84866735451 (Scopus)

Publication Title

IEEE Transactions on Communications

External Full Text Location

https://doi.org/10.1109/TCOMM.2012.072612.100666

ISSN

00906778

First Page

2571

Last Page

2579

Issue

9

Volume

60

Grant

2010-0017909

Fund Ref

National Research Foundation of Korea

This document is currently not available here.

Share

COinS