Adaptive Bitrate Video Caching in UAV-Assisted MEC Networks Based on Distributionally Robust Optimization

To alleviate the pressure on the ground base station (BS) from intensive video requests, unmanned aerial vehicle (UAV)-assisted mobile edge computing (MEC) has become a promising and flexible solution. The UAV carries a MEC server to provide caching and transcoding services for adaptive bitrate video streaming, which can reduce duplicate transmissions of the BS and the content acquisition latency of users, while improving the flexibility of video delivery. However, considering the uncertainty of user requests and content popularity distribution, improving the robustness of video caching is a challenge to promote practical applications. Thus, by integrating caching and transcoding on the UAV, as well as backhaul retrieving, we study the bitrate-aware video caching and processing with uncertain popularity distribution. Then, the problem of joint cache placement and video delivery scheduling under the worst-case distribution is formulated to minimize the total expected system latency with energy consumption constrained. Specifically, we use
$\zeta$
-structure probability metrics to characterize the uncertainty and construct confidence sets of arrival distribution. Furthermore, a distributionally robust latency optimization algorithm based on convex optimization theory is designed to obtain a robust solution. Finally, we conduct extensive simulations using real-world datasets to evaluate the effectiveness and robustness of the proposed scheme.