Optimal Real-Time Synchronized Scheduling for Collaborative Content Delivery

Motivated by applications such as cloud gaming and collaborative mixed reality, we consider the real-time synchronized scheduling for collaborative content delivery from a server to a group of users, ensuring that all users can view the content simultaneously for each delivery. We introduce latency compensation to control the server's content delivery time for each user, aiming to maximize synchronization performance (i.e., minimizing the gap between maximum and minimum latencies) while keeping the overall latency as low as possible. Our objective function includes a non-smooth component. To address this, we employ the stochastic proximal point (SPP) that sequentially updates the delay compensation, requiring only the previous latency samples instead of latency distribution. We then explore the structure of the subproblems and utilize the alternating direction method of multipliers (ADMM) to decompose the computation, resulting in a low-computational update with the complexity of at most O (N\log N) per iteration, where N is the number of users. Furthermore, we capture the convergence rate of our proposed algorithm. Finally, we demonstrate the superiority of our proposed algorithm against two baselines through simulations, including those based on real network latency trace datasets.