Enhancing the performance of HetNets via linear regression estimation of Range Expansion Bias

Abstract

Heterogeneous Networks (HetNets) use picocells deployed at strategic locations to fill coverage holes, improve Quality of Experience for users, and reduce blocking by supporting more users via cell splitting and spatial reuse of spectrum. Picocell Base Stations have lower transmit power compared to Macro Base Station. As a result, the observed improvements are often less than anticipated due to lower utilization and offloading. To improve picocell utilization, cell biasing attempts to offload users from macrocell by modifying cell selection/handover criteria. However, an improper bias value can increase blocking and penalize the users with higher macro interference. In this paper, we propose an efficient regression based scheme to predict a near optimal bias value that attempts to reduce blocking probability and improve load fairness index in the system. The simulation results verify that, in comparison to static bias, the proposed scheme also improves the cell edge user throughput, along with the target criteria.