Automatic Replanning of Tracking Areas in Cellular Networks
Matias Toril, Salvador Luna-Ramírez, Volker Wille · IEEE Transactions on Vehicular Technology · 2013
In future cellular networks, tracking areas (TAs) that are defined for location management purposes will be frequently updated to cope with changes in user trends. In this paper, an automatic method for defining when and how to change an existing TA plan to minimize network signaling is presented. The method first detects periods of similar user trends that could share the same TA plan. For this purpose, a series of graphs showing the evolution of user mobility and traffic is built from handover and paging statistics in the network management system (NMS). Then, a clustering algorithm is used to identify periods of similar user trends based on graph correlation. Thus, a TA replan is performed at the end of each period. To build the new TA plan, the TA replanning problem is formulated as a classical graph partitioning problem, which is then solved by a sophisticated graph partitioning algorithm combining multilevel and evolutionary techniques. Assessment is based on graphs taken from a live cellular network. Results show that the proposed method can significantly decrease the total number of location update (LU) and paging requests in the network by only changing the TA plan twice a week.