Power Allocation Algorithm Based on Machine Learning for Device-to-Device Communication in Cellular Network

With the development of the Internet, more and more mobile user equipment access to the cellular network, so the shortage of wireless spectrum resources has become increasingly prominent. Device-to-device (D2D) communication, as a key technology to solve this problem, can greatly improve the spectrum utilization rate and reduce the load of the base station. However, in the communication process of cellular users, D2D users occupying the same channel will bring complicated electromagnetic interference to them. This paper will establish a single-cell system model in which cellular users and D2D users coexist, and apply the method of power allocation to solve the problem of interference in the communication system. Then, we propose power allocation algorithm based on Q learning. Finally, the performance of the power allocation algorithm based on Q learning is analyzed and evaluated through the results of simulation experiments to verify the superiority of the algorithm over the performance of traditional power allocation algorithm.