A Study on Programming Model and Heuristic Optimization Algorithms for WCDMA Radio Networks

The 3^rd generation (3G) cellular networks, such as WCDMA (wideband code-division multiple access) networks, require accurate network planning and optimization. However, the planning and optimization of WCDMA radio network, which is a highly dynamic and inter-dependent system, are often carried out based on the static snapshot simulation due to simplicity and time limitation. Therefore, importing the link-level performance with consideration to some essential factors, such as power control and soft handover, has been a trend. Due to the complexity and inter-dependency, these characteristics have not been considered together in the previous works. In this paper, we give a brief introduction to our programming models with consideration to these characteristics, and present optimisation strategies base on three major meta-heuristics, namely genetic algorithms (GA), simulated annealing (SA) and variable neighborhood search (VNS), are presented. Extensive experiment results are provided and the performances of different algorithms are compared.