Proactive Load Balancing in Heterogeneous Cellular Networks

TL;DR

This paper introduces a proactive load balancing framework for heterogeneous cellular networks that uses content caching and mobility prediction to enhance user QoE, reduce congestion, and improve downlink rates.

Contribution

It proposes a novel proactive load balancing approach combining content caching and mobility prediction, outperforming reactive schemes in HetNets.

Findings

Achieves 98.7% cell load balancing fairness

Reduces backhaul load by 20%

Improves downlink rates by 31%

Abstract

Recent exponential growth of data over cellular networks has cause the progression from conventional mobile communication networks to heterogeneous cellular networks (HetNets). Quality of experience (QoE)-aware traffic load balancing in such dense HetNets is considered a major problem. Current HetNets exploit reactive load balancing schemes that hinder in achieving desired QoE gain. In this paper, we propose a novel proactive load balancing framework which leverages content caching and mobility prediction to improve user QoE. We use Semi-Markov model to predict users' future cells and a novel proactive caching algorithm is proposed to pre-fetch user future demands at these cells. This allows us to reduce cell congestion and offers better average downlink rates. To validate the effectiveness of the proposed framework, system level simulations are performed and compared with state of the art approaches. Encouragingly, our system achieves up-to $98.7\%$ cell load balancing in terms of Jain's fairness index, reduces $20\%$ backhaul load, and provides $31\%$ improved downlink rates at congested cells in contrast to reactive approaches.