Adaptive Multicell 3D Beamforming in Multi-Antenna Cellular Networks

TL;DR

This paper introduces an adaptive multicell 3D beamforming method that dynamically switches between horizontal and vertical cooperation strategies to optimize cellular network performance with reduced complexity.

Contribution

It proposes a novel hybrid multicell cooperation technique that adapts based on vertical regions, improving performance while lowering complexity and signaling compared to full cooperation.

Findings

Achieves comparable performance to fully cooperative transmission

Reduces complexity and signaling requirements

Provides analytical expressions for ergodic rates

Abstract

We consider a cellular network with multi-antenna base stations (BSs) and single-antenna users, multicell cooperation, imperfect channel state information, and directional antennas each with a vertically adjustable beam. We investigate the impact of the elevation angle of the BS antenna pattern, denoted as tilt, on the performance of the considered network when employing either a conventional single-cell transmission or a fully cooperative multicell transmission. Using the results of this investigation, we propose a novel hybrid multicell cooperation technique in which the intercell interference is controlled via either cooperative beamforming in the horizontal plane or coordinated beamfroming in the vertical plane of the wireless channel, denoted as adaptive multicell 3D beamforming. The main idea is to divide the coverage area into two disjoint vertical regions and adapt the multicell cooperation strategy at the BSs when serving each region. A fair scheduler is used to share the time-slots between the vertical regions. It is shown that the proposed technique can achieve performance comparable to that of a fully cooperative transmission but with a significantly lower complexity and signaling requirements. To make the performance analysis computationally efficient, analytical expressions for the user ergodic rates under different beamforming strategies are also derived.