A novel approach to processing speed enhancement for dynamic CDMA network simulations

After the first 3G WCDMA networks were put into operation in Japan and Europe, network operators discovered unexpected problems during busy minutes of packet data traffic bursts. The solution for these problems is beyond the limits of conventional CDMA planning software. In contrast, dynamic network simulations are the perfect tool for troubleshooting and optimization of these issues. However, these tools still suffer from an unacceptably high computation time for day-to-day use. While parallelization strategies to overcome the performance problem have not led to a broadly usable solution in the past, the paper presents alternative approaches that reduce the computational load of dynamic mobile network simulations by introducing a semi-dynamic traffic load bias and compile-time scheduling. Thereby, the speed of dynamic simulations can be improved by up to 400% without a significant loss in the correctness of the simulation results. For the first time, this simulation technique allows the application of dynamic simulations to everyday troubleshooting and radio resource management (RRM) optimization of CDMA networks.