Modelling Load Balancing and Carrier Aggregation in Mobile Networks

TL;DR

This paper analyzes multicarrier mobile networks, proposing two load balancing schemes for Carrier Aggregation, and evaluates their performance, showing near-ideal load balancing and insensitivity to traffic mix.

Contribution

It introduces two novel load balancing schemes, JFQ and VB, specifically designed for Carrier Aggregation in mobile networks, with analytical performance evaluation.

Findings

Proposed schemes achieve quasi-ideal load balancing.

Performance is insensitive to traffic mix of CA and non-CA users.

Analytical modeling confirms effectiveness of the schemes.

Abstract

In this paper, we study the performance of multicarrier mobile networks. Specifically, we analyze the flow-level performance of two inter-carrier load balancing schemes and the gain engendered by Carrier Aggregation (CA). CA is one of the most important features of HSPA+ and LTE-A networks; it allows devices to be served simultaneously by several carriers. We propose two load balancing schemes, namely Join the Fastest Queue (JFQ) and Volume Balancing (VB), that allow the traffic of CA and non-CA users to be distributed over the aggregated carriers. We then evaluate the performance of these schemes by means of analytical modeling. We show that the proposed schemes achieve quasi-ideal load balancing. We also investigate the impact of mixing traffic of CA and non-CA users in the same cell and show that performance is practically insensitive to the traffic mix.