Deep Compression Based Intelligent Power Allocation Algorithm for Internet of Vehicles.

The high-speed movement of vehicle brings about the rapid change of communication environment. This leads to a sharp increase in the exchange amount of vehicle environment status information (ESI), which results in the communication quality dissatisfaction for Internet of Vehicles (IoV) user. To achieve efficient resource allocation, how to reduce the ESI exchange imposes crucial challenge for the rate improvement of IoV. In this paper, we propose the ESI deep compression based intelligent power allocation (ECIA) algorithm for IoV. First, a deep neural network (DNN) is used to compress the complex ESI to reduce the amount of information exchange in IoV. Then, a decision network composed of the total power network and power allocation network is designed to maximize the sum rate of Vehicle-to-Vehicle (V2V) links. Finally, simulation results show that the proposed ECIA algorithm can effectively increase the sum rate of V2V links and guarantee the communication quality of IoV user.