A hybrid framework for capacity and coverage optimization in self-organizing LTE networks

In this paper, we address the capacity and coverage optimization (CCO) problem for LTE networks by means of self-organizing network (SON) techniques. A novel hybrid two-layer optimization framework is proposed to enhance the network capacity and coverage, where on the top layer a network entity of eCoordinator is implemented to ensure overall network coverage by optimizing the antenna tilt and capacity-coverage weight of each cell in a centralized manner, and on the bottom layer individual eNB optimizes cell-specific capacity and coverage by tuning its pilot power in a distributed manner. A heuristic algorithm is developed for the eCoordinator operation at large time granularity and the Genetic Programming (GP) approach is exploited for the eNB operation at small time granularity, for the purpose of tracking overall network performance as well as adapting to network dynamics. Our simulation results have demonstrated the usefulness of the proposed algorithms by enhancing network capacity and coverage performance under various system requirements.