Collaborative Edge Processing by mmWave Access Providing UAVs

A wireless IoT system consisting of unmanned aerial vehicles (UAVs) is investigated. They work with a ground base station to provide mmWave access to user equipment and IoT devices. In parallel, they collaboratively share edge processing load. We design a modified Lyapunov optimization framework for assigning tasks to multi-hop connected UAVs. The proposed strategy constructs a task flow control policy that is near-optimal with respect to energy consumption and response time. The best task migration and distribution over the UAVs can be found out at a bearable complexity. Numerical results confirm the soundness of the framework regarding power cost, execution latency and overhead data.