Particle Swarm Optimization of Fuzzy Logic Controller for high quality RRM Auto-Tuning of UMTS Networks
Hervé Dubreil, Zwi Altman, Vincent Diascorn, Joseph S. Picard, Maurice Clerc · 2005
Auto-tuning of radio resource management (RRM) parameters using fuzzy logic controllers (FLCs) has recently been introduced as a means to enhance UMTS network performance and adapt it to traffic variations. This paper presents a methodology for optimizing the performance of FLC-based auto-tuning tasks. The particle swarm (PS) approach is utilized to perform the optimization. The feasibility and effectiveness of optimizing and auto-tuning both admission control and macrodiversity parameters are illustrated. The scalability of FLC designs optimized for small network size and utilized for auto-tuning of large networks is demonstrated.