A Q-learning-Based Dynamic Spectrum Allocation Algorithm
Chunying Lv, Jiyang Wang, Fei Richard Yu, Hui Dai · Proceedings of the 2nd International Conference on Computer Science and Electronics Engineering (ICCSEE 2013) · 2013
Contraposing the status of spectrum resources is tension and spectrum utilization is low, the dynamic spectrum allocation technology is studied.Due to lack of central control entity in distributed cognitive networks, cognitive users need to select channel and power for communication intellective.Based on the intelligence of Q-learning, we proposed an improved dynamic spectrum allocation algorithm.The state space, action space and reward function of the algorithm are built in the paper, and the agents are guided to perform actions through designing the reward function.The algorithm gives fully autonomy selective to cognitive users.This strengthened the dynamic management of spectrum resources.We also did many experiments.Numerical simulation results show that the proposed algorithm can not only realize the autonomy of channel and power allocation, but also improve system throughput compared to others.It boosts spectrum utilization effectively.