In this paper, a clustered parallel training algorithm is designed for personalized federated learning (Per-FL), called CPPer-FL. CPPer-FL improves the communication and training efficiency of Per-FL from two perspectives, namely, less burden for the central server and lower interaction idling delay. CPPer-FL adopts a client-edge-center learning architecture, which offloads the central server's model aggregation and communication burden to distributed edge servers. Also, CPPer-FL redesigns the cascading model synchronization and updating procedure in conventional Per-FL and changes it to a parallel manner, thus improving the interaction efficiency in the training process. Further, for the proposed hierarchical architecture, two approaches are proposed to cater to Per-FL: similarity-based clustering for client-edge association and personalized model aggregation for parallel model updating, such that clients' personal features can be preserved in the training process. The convergence of CPPer-FL has been formally analyzed and proved. Evaluation results validate the communication efficiency, model convergence, and model accuracy improvement.

Viewport prediction and prefetch have an important influence on VR video streaming performance. This work proposes a novel federated learning-based viewport prediction model training algorithm, ComPer-FedAvg. The proposed algorithm leverages a VR video’s common viewing pattern and users’ personal viewing patterns to train the prediction model in a distributed and privacy-preserving manner. Further, considering the VR video viewport prediction accuracy, a stochastic game is formulated to solve the VR streaming network’s communication resource allocation problem, where limited communication resource blocks are auctioned to users to achieve the optimal overall VR viewing experience. For each user, the auction is decomposed into two disjoint subproblems, namely, the optimal number of data rate requesting and true value claiming (bidding). The optimal true value claiming has been analytically proved to be equal to the VR viewing reward with given data rate. Due to the lack of global information when users request data rate, we reformulate users’ data rate requesting problem as a POMDP problem. A novel deep reinforcement learning algorithm is adopted to solve the problem. Evaluation and simulation results show the proposed viewport prediction and VR streaming schemes outperform conventional solutions in terms of prediction accuracy and VR viewing experience.