A Mobile Devices Assisted Adaptive Content Delivery Strategy in Urban Public Transportation System

The Internet of Vehicles (IoV) is one of the classic scenarios for the Internet of Mobile Things (IoMTs). While the entertainment demand (videos, photos, etc.) of public vehicle passengers has explosively grown in recent years, it results in huge traffic pressure on the IoV. Edge computing is regarded as a potential solution, and an edge server can act as a powerful end node to provide computing or storage service to mobile passengers. However, the fixed edge server (road side unit, base station, etc.) is hard to establish a stable communication connection with passengers for high mobility. To tackle this problem, we propose a mobile devices assisted adaptive content delivery strategy, where Deep Reinforcement Learning is applied to generate the processing sequence of the request considering the delivery deadline, content size, and payment. Compared with the model-based delivery strategy, the proposed scheme can adjust the criteria according to the request processing queue. By the way, a D2D transmission model is designed to further improve the transmission rate under the scheduling of the proposed method. The simulation results show that the proposed mechanism can deal with more passenger requests and decrease the content delivery delay than the First-Come-First-Serve mechanism.