An Almost Blank Subframe Allocation Algorithm for 5G New Radio in Unlicensed Bands

The 5G new radio communications in unlicensed bands (5G NR-U) can meet the growing data traffic requirements and make better use of the unlicensed bands. However, the harmonic coexistence of NR-U and WiFi in unlicensed bands is a challenging problem as NR-U and WiFi employ different access technologies. To address this problem, this paper proposes an almost blank subframe (ABS) optimization mechanism by jointly considering the data transmission of WiFi users and the positions of ABS, which can achieve fair coexistence of NR-U and WiFi and effectively improve the system throughput. Specifically, we first investigate the optimal number of ABS according the data transmission of WiFi users. Then, we use the Q-learning algorithm to learn the data transmission rules of WiFi users to solve the problem of matching data transmission positions of WiFi users with ABS positions. Simulation results show that the algorithm can effectively improve the total throughput compared with the traditional ABS allocation algorithm.