Self-organizing wireless network parameter optimization through mixed integer programming

Self-organizing wireless access networks rely on key configuration parameters that guarantee quality of operation. Poorly planned network access parameters, such as channel assignments or cell identifying codes, result in increased interference on system-critical control channels, network access failures, and poor quality of service. Therefore, the optimization of such network parameters is a vital task for self-organizing wireless networks. This paper considers Long-Term Evolution (LTE) systems and identifies a number of network parameter optimization problems encountered when designing self-organizing wireless access networks. A Mixed Integer Quadratic Program (MIQP) optimization model is proposed for each problem. The optimization of frequency channel assignments, Tracking Area Codes, Physical Cell Identifiers, and LTE coordination clusters is discussed and modeled in detail. Simulation results demonstrate significant network quality gains from optimizing network parameters using the proposed integer optimization models.