Load & Backhaul Aware Decoupled Downlink/Uplink Access in 5G Systems

TL;DR

This paper proposes a load and backhaul aware cell association algorithm for 5G systems that improves uplink and downlink performance by considering network load and backhaul capacity, validated through realistic simulations.

Contribution

It introduces a novel association algorithm that explicitly accounts for cell load and backhaul capacity, outperforming previous methods in 5G heterogeneous networks.

Findings

The proposed algorithm outperforms prior art in simulations.

Considering load and backhaul improves association efficiency.

Power control settings influence algorithm performance.

Abstract

Until the 4th Generation (4G) cellular 3GPP systems, a user equipment's (UE) cell association has been based on the downlink received power from the strongest base station. Recent work has shown that - with an increasing degree of heterogeneity in emerging 5G systems - such an approach is dramatically suboptimal, advocating for an independent association of the downlink and uplink where the downlink is served by the macro cell and the uplink by the nearest small cell. In this paper, we advance prior art by explicitly considering the cell-load as well as the available backhaul capacity during the association process. We introduce a novel association algorithm and prove its superiority w.r.t. prior art by means of simulations that are based on Vodafone's small cell trial network and employing a high resolution pathloss prediction and realistic user distributions. We also study the effect that different power control settings have on the performance of our algorithm.