Optimal dynamic periodic location update for UMTS mobile networks - a simulation approach

Location management is essential for delivering services to mobile stations (MSs) in mobile networks. In 3GPP technical specifications for the circuit-switched domain of Universal Mobile Telecommunications System, both normal location update and periodic location update (PLAU) are adopted to track locations of MSs. PLAU is also used for detecting detachment of an MS from the network. In this paper, we analyze the cost of location management under the presence of abnormal detachments. Based on our cost analysis, we propose a simulation approach to find the optimal parameter to minimize the cost for the dynamic PLAU. A gradient descent algorithm is also implemented to find the optimal parameter for a different cost function. Results show that the cost is significantly deducted with the proposed optimal approach. Another contribution is that we show how to solve an optimization problem through a simulation approach.