Dynamic Fuzzy Q-learning for Handover Parameters Optimization in 5G Multi-Tier Networks

Jin Wu,Jing Liu,Zhangpeng Huang,Shuqiang Zheng

DOI: https://doi.org/10.1109/wcsp.2015.7341220

2015-01-01

Abstract:The mobility robustness optimization can significantly enhance the quality of service in scenarios characterized by dense uncoordinated deployment of small cells, as targeted by future 5th generation (5G) radio access technology. Current solutions mostly rely on priori knowledge and rule based algorithms, these solutions do have achieved good performance. There is still, however, a lot of room for further improvements, especially when enough priori knowledge is not available. In this paper, we propose a dynamic fuzzy Q-Learning algorithm for mobility management in small-cell networks. There are no fuzzy rules initially, this algorithm gradually generates new fuzzy rules and gets the required parameters through system learning, so as to reach a balance between the signaling cost caused by handover and the user experience affected by call dropping ratio. Performances are evaluated in a LTE system level simulator and impact of UE speed is considered. Simulation results show the efficiency of the proposed algorithm in minimizing the number of handovers while maintaining call dropping ratio at a minimal level.