Self-optimizing load balancing with backhaul-constrained radio access networks

TL;DR

This paper enhances Self-Organizing Network algorithms by incorporating backhaul constraints into load definitions, leading to improved autonomous load balancing in heterogeneous radio access networks.

Contribution

It introduces a new load definition considering backhaul state and integrates it into a self-optimized load balancing algorithm for better network performance.

Findings

Load estimator effectively captures backhaul effects.

Enhanced SON algorithm maintains proper operation under backhaul constraints.

Simulation confirms improved load balancing in heterogeneous networks.

Abstract

Self-Organizing Network (SON) technology aims at autonomously deploying, optimizing and repairing the Radio Access Networks (RAN). SON algorithms typically use Key Performance Indicators (KPIs) from the RAN. It is shown that in certain cases, it is essential to take into account the impact of the backhaul state in the design of the SON algorithm. We revisit the Base Station (BS) load definition taking into account the backhaul state. We provide an analytical formula for the load along with a simple estimator for both elastic and guaranteed bit-rate (GBR) traffic. We incorporate the proposed load estimator in a self-optimized load balancing algorithm. Simulation results for a backhaul constrained heterogeneous network illustrate how the correct load definition can guarantee a proper operation of the SON algorithm.