Research And Implementation On Handover Self-optimization In LTE

Self-optimizing network (SON) technology is a key technology in LTE system, itspurpose is reduce system operating costs and improve system performance and improvethe user experience effectively. SON technology is divided into self-configuring,self-optimizing and self-healing. Mobile robust optimization (MRO) and mobility LoadBalancing (MLB), which are belong to the self-optimization techniques, are the goal of thispaper studying.The main purpose of MRO is to reduce radio link failure caused by handover and thewaste of network resources caused by unnecessary handover or unreasonable handover.The abnormal handover scenes of MRO include too early handover, too late handover,handover to wrong cell and ping-pong handover. So, MRO algorithm aims to reduce zheabnormal handover numbers according to adjust handover parameters and reselectionparameters automatically. The purpose of MLB algorithm distributes UE to the suitablecell to achieve balanced business by controling the mobile parameters (handoverparameters and reselection parameters) without affecting the quality and service(QoS). Theinnovations of this paper are the following:1. The paper presents a MRO algorithm, which is divided into global optimization andlocal optimization. Global optimization automatically adjusts the base-level parameters,and local optimization is mainly adjust the cell-level parameters. It improves the handoverperformance of LTE system through these two optimization methods.2. The paper presents a MLB algorithm, which determines an automatic neighborhoodselection mechanism based on cell load calculation, the ineractions between serving celland neighbor cell and evaluate the load of neighbor cell could receiving in advance. Themechanism arranges the order of cell according to the size of cell load. The algorithm onlyselect one neighbor cell for each optimization, and finally reduce the load of serving cellwithout increasing the load of neighbor to high load.3. The paper proposed a coordinated program between MRO and MLB, which does not affect the execution of an individual optimization algorithm. The program can avoidconflict of adjusting parameters in two algorithms by limiting the range of parameters andthe number of parameters. At the same time, the program can limit an individual algorithmoptimization time and stop the algorithm and fallback the parameters.