Timeliness Analysis of Service-Driven Collaborative Mobile Edge Computing in UAV Swarm

Abstract

Unmanned aerial vehicles (UAVs) can be conveniently deployed for reconnaissance, firefighting, disaster rescue, and so on. However, the utmost challenge is how to support a wide variety of services efficiently in face of communication and computation constraints. Therefore, a service-driven collaborative mobile edge computing model is proposed to support the computation- intensive and latency-critical services for UAV swarm. In terms of different service requirements, UAVs are divided into two categories, including the cloud service support (CSS) UAVs, and the local service support (LSS) UAVs. For CSS UAVs, the enhanced collaborative computing model is designed and the closed-form solution of the decision threshold is also achieved by using queueing theory. Furthermore, in terms of LSS UAVs, the system status update algorithm is proposed based on the age of information in collaboration with CSS UAVs and the control center. In contrast to the typical UAV network with MEC servers, simulation results verify that our proposed model and algorithm can significantly improve the timeliness of information transmission.