Performance Analysis of Profile Based Paging for Addis Ababa UMTS Networks

Paging is an important mobile network procedure that is performed to locate and connect mobile users when they receive calls/sessions. It is needed as users are mobile, and continuously updating their serving cells is not bandwidth efficient. In traditional broadcast paging, all cells of a network are grouped into clusters that are identified by Location Area Code (LAC). Users update their LAC and network look for a paged user by sending paging messages for all cells within the user's LAC. A decision on cluster size is an important factor that determines the paging overhead and connection delay. The larger cluster size, the smaller LAC update but the larger capacity for the paging message. To minimize the capacity required for paging messages while maintaining a larger cluster size, profile-based paging has been proposed and studied. Yet, such a paging approach has not been investigated and applied for Addis Ababa third-generation network. In this thesis work, we analyzed the performance of profile-based paging algorithm for Addis Ababa third-generation network based on call detail record data collected from the network management system and compared it with conventional one. The algorithms considered were static and dynamic profile based paging. We analyzed the paging algorithms for users with less, moderate and highly mobile. A user's mobility is modeled using a Markov model. Achieved results show that the profile-based algorithm considerably increases the efficiency of utilization with slight or no effect in the overall paging success rate.