Optimizing LTE test traffic using search and expert algorithms

Abstract: This thesis investigates whether artificial intelligence techniques can replace the optimization process that is currently performed by experts for LTE/4G stability tests at Ericsson. The investigation covers the usage of search algorithms and expert systems and the thesis then implements a hybrid algorithm. This implemented solution is then tested upon three different architectures to judge how well it optimizes tests without causing further problems. The conclusion is that the implemented solution optimizes tests sufficiently but does so using only mobile users. To reflect reality better, static users would also need to be included in the optimization process.