Title
A Distributed User Association Method for LTE-U by Enabling Q-Learning in Minority Game
Abstract
To meet the continuously increasing capacity insufficiency challenges brought by the next generation mobile networks (5G), the long term evolution-unlicensed (LTE-U) has been introduced as an emerging technology to supply the limited licensed bands. As multi-radio access technology (RAT) HetNets gradually becomes the dominant form of future networks, RAT selection and user association issues arise from the extensions of mobile networks to unlicensed bands, aiming to achieve fair and harmonious coexistence with other RATs in unlicensed bands such as WiFi. In this work, we propose a distributed method of RAT selection and user association using Q-learning in minority game, where users select an RAT from two, the cellular and the WiFi networks, and self-organize to a balanced state while seeking their own interests. This reinforcement learning method only requires minimal information from the environment. Moreover, it suppresses the inherent crowd effect and the performance degradation in minority game with an arbitrary cut-off value. Numerical results demonstrate the effectiveness and the advancement of our proposed method compared with the classic algorithm of minority game.
Year
DOI
Venue
2019
10.1109/ICCChina.2019.8855939
2019 IEEE/CIC International Conference on Communications in China (ICCC)
Keywords
Field
DocType
dominant form,user association issues,unlicensed bands,fair coexistence,harmonious coexistence,RAT,distributed method,minority game,WiFi networks,reinforcement learning method,distributed user association method,LTE-U,long term evolution-unlicensed,licensed bands,multiradio access technology,Q-learning
Access technology,Unlicensed band,Minority game,Computer science,Q-learning,Computer network,Emerging technologies,Next Generation Mobile Networks,Reinforcement learning
Conference
ISSN
ISBN
Citations 
2377-8644
978-1-7281-0733-2
0
PageRank 
References 
Authors
0.34
13
3
Name
Order
Citations
PageRank
Yunjia Wang17115.63
Guanding Yu21287101.15
Rui Yin312911.38