A SON-Based Algorithm for the Optimization of Inter-RAT Handover Parameters

First, the deployment of the Long-Term Evolution (LTE) system will be concentrated on areas with high user traffic overlaying with the legacy second-generation (2G) or third-generation (3G) mobile system. Consequently, the limited LTE coverage will result in many inter-radio access technology (RAT) handovers from LTE to 3G systems and vice versa. Trouble-free operation of inter-RAT handovers requires the optimization of the handover parameters of each cell in both RATs. The current network planning and optimization methods provide a fixed network-wide setting for all the handover parameters of the cells. Cells that later show considerable mobility problems in operation mode are manually optimized with the aid of drive tests and expert knowledge. This manual optimization of the handover parameters requires permanent human intervention and increases the operational expenditure (OPEX) of the mobile operators. Moreover, the interoperability of several RATs increases further the parameter space of the handover parameters, which makes the manual optimization difficult and almost impracticable. To reduce OPEX and to achieve a better network performance, we propose in this paper a self-optimizing algorithm where each cell in a RAT updates its handover parameters in an autonomous and automated manner depending on its traffic and mobility conditions. The proposed algorithm uses a feedback controller to update the handover parameters as a means to providing a steady improvement in the network performance. In the context of control theory, the feedback controller consists of a proportional control block, which regulates the change in the magnitude of each handover parameter, and a gain scheduler, which modifies the parameters of the proportional control block depending on the mobility conditions in each cell. To benchmark the design of the proposed algorithm, we apply two general and nonself-optimization algorithms: Taguchi's method and simulated annealing to optimize the handover parameters. Simulation results show that the proposed self-optimizing algorithm reaches a stable optimized operation point with cell-specific handover parameter settings, which considerably reduce the number of mobility failure events in the network, compared with three fixed settings for the handover parameters. Moreover, it is presented that the proposed self-optimizing algorithm outperforms Taguchi's method and simulated annealing when applied to a mobility robustness optimization (MRO) problem.