A Basic Unified Context for Evaluating the Beam Forming and MIMO Options in a Wireless Link

TL;DR

This paper compares beamforming and MIMO techniques in wireless links, analyzing their performance based on channel properties and array configurations to determine when each method is advantageous.

Contribution

It provides a unified framework for evaluating beamforming and MIMO, highlighting how their effectiveness depends on eigenmode spectra and array size.

Findings

MIMO outperforms beamforming for large array apertures.

Beamforming performs nearly as well as MIMO for smaller arrays.

Performance varies with eigenmode power distribution and array parameters.

Abstract

For one isolated wireless link we take a unified look at simple beamforming (BF) as contrasted with MIMO to see how both emerge and under which conditions advantage goes to one or the other. Communication is from a high base array to a user in clutter. The channel propagation model is derived from fundamentals. The base knows the power angular spectrum, but not the channel instantiation. Eigenstates of the field spatial autocorrelation are the preferred apodizations (APODs) which are drivers of the natural modes for exciting lectric fields. Preference for MIMO or BF depends on APOD spectra which are surveyed pointing to various asymptotic effects, including the maximum BF gain. Performance is studied under varying eigenmode power settings at 10% outage. We focus on (1,4) driving the strongest mode for BF and (4,4) driving the 4 strongest for MIMO. Results are obtained under representative parameter settings, e.g. an angular spread of 8 deg, 2 GHz carrier, 0 dB SNR and an array aperture of 1.68m (4 field decorrelation lengths) with antenna elements spaced as close as lambda/2. We find MIMO excelling for array apertures much larger than the decorrelation length; BF does almost as well for smaller apertures.