Optimization of algorithms for mobility in cellular systems

Sammanfattning: Inter-frequency measurements are needed to determine when user equipment should make a handover to the best available base station in a cellular mobile network. The measurements are expensive from a resource point of view and therefore there is a need for optimization of this kind of measurements. In this work both this optimization and a way to predict the future of the measurements have been evaluated in a simulated Long-Term Evolution network. Both fast and slowly moving user equipment have been tested and the prediction was made by storing old measurements and calculating the gradient of the signal power. Depending on the resulting signal's gradient, different decisions on what to do were made. The results were then compared to the traditional way of controlling the inter-frequency measurements, in terms of throughput, handover failures and other quality factors. The results from the simulations show that there is some optimization that can be made without compromising the connection. More specifically, the time spent measuring inter-frequency has been successfully improved (lowered). Handover failures have been harder to control and the throughput has more or less been unchanged throughout the simulations. The speed of the user equipment influenced the results a lot and no setup was found that works best with all UE speeds.