Cell-Centric Tracking Area Planning with Mobility Prediction on LTE/LTE-A(I)

Project: National Science and Technology CouncilNational Science and Technology Council Academic Grants

Project Details


Location management is one of the most critical issues at LTE/LTE-A network. The primary purpose is to utilize dynamic tracking area (TA) planning to lower location management cost. The ping-pong effect will happen when some of the related works plan static TAs. Some regard crossing rate of user equipment (UE) between cells as the primary element of TA planning. However, this kind of planning will cause much higher cost because there are various directions of UEs in the same cell. In order to solve the problem, we propose cell-centric TA planning algorithm, which utilize perching time that the UEs in the cell to calculate the relevance between cells and further plan TAs. In order to increase cell-centric performance to adapt to direction-strong environment, (for example, UEs’movement possess the direction of the law) we propose 2-order Markov Chain to construct UE’s movement model to predict the UE’s moving path and further planning lower cost for TA. However, considering specific circumstances is unable to meet the actual situation. For instance, UEs’ moving behavior is not static on the general road. Therefore, we propose variable-order Markov Chain to adapt the general situation. We set different order based on different situation and make constructed prediction model adapt to the actual situation. Last, we will femtocell into the system. In the hierarchical structure of cell and femtocell, letting femtocell as center and providing complete TA planning. This will make the cost optimization.

Project IDs

Project ID:PB9907-12651
External Project ID:NSC99-2221-E182-039
Effective start/end date01/08/1031/07/11


  • Location area planning
  • Location management
  • Automatic reconfiguration
  • Mobility management
  • Long Term Evolution (LTE) networks.


Explore the research topics touched on by this project. These labels are generated based on the underlying awards/grants. Together they form a unique fingerprint.