User Association For Load Balancing In 5g Network Based On Min-Max Optimization

The rapid increase of user equipments has stressed the heterogeneous and enormous wireless network in recent years. As an effective way to promote resource utilization and improve network performance, load balancing has becoming more and more important in the upcoming fifth generation (5G) network. In this work, a load balancing problem with limited bandwidth constraints of different cells is investigated from base stations perspective and the way to balance these loads is to design an proper user association scheme. Consequently, A 0-1 min-max linear optimization model is established to describe the problem mathematically and robustly. Moreover, based on the upper and lower bound analysis of the model, a bisection method algorithm is designed to approach the optimal solution and its time complexity is O(mnlogn), where m, n are the amount of cells and users respectively. The numerical results of simulation experiment demonstrate the reliability and rapid computational efficiency of bisection method.