LAUSR

The performance of load balancing algorithms in HetNets is investigated in this paper. We performed a comparative analysis of three load balancing algorithms, including our previously proposed algorithm based on local optimization. We evaluated the offloading capabilities, fairness of the distribution of the load across base stations and spectral efficiency gains in the downlink. As opposed to previous studies were the evaluation of such algorithms was carried out under static conditions, in this study we performed the analysis in a dynamic and more realistic scenario. We quantified the effects of user mobility and the triggering of handovers during the load balancing process on the gains provided by the algorithms. We were able to show that the resulting gains in spectral efficiency after balancing the load are not as high as previously reported under static conditions. Our study shows that user mobility does affect the benefits of load balancing algorithms, particularly in the case of iterative algorithms. However, balancing the load in a HetNet still leads to faster downlink data rates, up to 80% gain for cell edge users was achieved according to our study. Our results show that our load balancing algorithm was able to provide similar benefits compared to the other two algorithms, however our approach is substantially less complex. Additionally, our load balancing algorithm was superior in terms of its practicality due to a minimum level of coordination and exchange of information among base stations.