Abstract : Combining mobile edge computing (MEC), themulti-drone aided space-air-ground integrated Internet of things(SAG-IoT) networks can provide ground IoT devices (GIDs) highqualitywireless access and computing services. However, the diversetasks, moving drones, and limited network resources revealgreat challenges for the task offloading and resource allocationscheme exploitation. Especially, given the restricted computationresources, how to make full use of available applications deployedon MEC servers (MECSs) to compute various types of tasks, iseven an important issue. To the best of our knowledge, it is anentirely new problem since most existing works in this line assumethat all types of applications can be deployed on one MECS so asto process various offloaded tasks. Toward this end, we presentthis paper to investigate inter-server computation offloading, resourceallocation, and drone deployment to minimize the overallcomputation overhead of all GIDs. An iteratively optimizationalgorithm is proposed which alternately utilizes heuristic greedyand successive convex approximation methods. Simulation resultsverify that, for different GID numbers, optimization schemes, andcomputing models, our devised schemes can not only significantlyreduce the overall computation overhead but also achieve optimaldecisions of computation offloading, spectrum allocation, anddrone deployment.
Index terms : Bandwidth allocation, inter-server computation offloading, multi-drone, space-air-ground integrated IoT network.