Improving the Performance of the Zero-Forcing Multiuser MISO Downlink Precoder Through User Grouping
Year: 2015
Type: article
Abstract: We consider the multiple input single output (MISO) Gaussian broadcast channel with Ntantennas at the base station (BS) and Nusingle-antenna users in the downlink. We propose a novel user grouping precoder which improves the sum rate performance of the zero-forcing (ZF) precoder specially when the channel is ill-conditioned. The proposed precoder partitions all the users into small groups of equal size. Downlink beamforming is then done in such a way that, at each user's receiver, the interference from the signal intended for users not in its group is nulled out. Intragroup interference still remains, and is cancelled through successive interference presubtraction at the BS using dirty paper coding (DPC). The proposed user grouping method is different from user selection, since it is a method for precoding of information to the selected (scheduled) users, and not for selecting which users are to be scheduled. The proposed precoder is a generalization of two special cases, one where each group has only one user (ZF precoder) and another where all users are in a single group (ZF-DP precoder). A larger group size helps improve the sum rate performance but at the cost of greater complexity. The proposed generalization, therefore, allows for tradeoff between performance and complexity.
Authors: Saif Khan Mohammed, Erik G. Larsson
Institutions: Indian Institute of Technology Delhi, Linköping University
Language: English
Subfield: Electrical and Electronic Engineering
Field: Engineering
Domain: Physical Sciences
Open Access status: green
OpenAlex ID: W1594064992