4. Optimal UAV Caching and Trajectory Design in the AGVN

In this chapter, we investigate the UAV-assisted mobile edge caching to assist terrestrial vehicular networks in delivering high-bandwidth content files. To maximize the overall network throughput, we formulate a joint caching and trajectory optimization (JCTO) problem to jointly optimize content placement, content delivery, and UAV trajectory. Considering the intercoupled decisions and the limited UAV energy, the formulated JCTO problem is intractable directly and timely. Therefore, we propose a deep supervised learning (DSL) scheme to enable intelligent edge for real-time decision-making in the highly dynamic vehicular networks. Specifically, we first propose a clustering-based two-layered (CBTL) algorithm to solve the JCTO problem offline. With a given content placement strategy, we devise a time-based graph decomposition method to jointly optimize the content delivery and trajectory design, with which we then leverage the particle swarm optimization (PSO) algorithm to further optimize the content placement. We then design a convolutional neural network (CNN)-based DSL architecture to make fast decisions online. The network density and content request distribution with spatial–temporal dimensions are labeled as channeled images and input to the CNN-based model, and the results achieved by the CBTL algorithm are labeled as model outputs. With the CNN-based model, a function mapping the input network information to output decisions can be intelligently learnt to make timely decisions. Extensive trace-driven experiments are carried out to demonstrate the efficiency of CBTL in solving the JCTO problem and the superior learning performance with the CNN-based model.