Cellular Network Power Allocation Algorithm Based on Deep Reinforcement Learning and Artificial Intelligence

In the shortest path planning problem, the old algorithm usually has many defects, such as the robot's cognition being contrary to reality, the lack of practical operation feasibility, or the limitation of problem processing. Nowadays, with deep learning, artificial intelligence algorithms tend to be mature; it has become a mainstream trend to adopt end-to-end learning system instead of traditional old algorithms. In recent years, with the rise of the Internet of things emerging technology industry and the explosive surge of network data traffic, the drawback is the increasingly severe shortage of wireless spectrum resources. In order to effectively reduce the cochannel interference of D2D communication technology in the system and enhance the useable range of the cellular network, it is necessary to distribute the useful and efficient cellular resources of the system. In this article, we will study the D2D users and the selection scheme of D2D users' transmission power control mode and allocate the spectrum resources in the uplink of the cellular users in the communication network. In order to reduce the cochannel interference in a cellular network and improve the spectrum utilization of the system, the research direction of this article is to solve the problem of user communication resource allocation in a single-cell hybrid cellular network.