GJRA: A Global Joint Resource Allocation Scheme for UAV service of PEC in IIoTs
The Industrial Internet of Things (IIoT) is an emerging paradigm to make industrial operations more efficient and intelligent by deploying a massive number of wireless devices to industry scenes. However, due to the limited computing capability and batteries, the Industrial Internet of Things Devices (IIoTDs) can't perform the computation-intensive or delay-sensitive tasks well and provide long-term services in practical. To tackle these challenges, we present an effective global joint resource allocation scheme for Unmanned Aerial Vehicle (UAV) service of Pervasive Edge Computing (PEC) in Industrial Internet of Things (IIoTs) and studied a collaborative UAV server-IIoTDs scheme in this paper. In our proposed scheme, the IIoTDs can keep high-efficiency performance even their battery ran out, by deploying the UAV as a PEC server and a mobile power source to provide task offloading and energy harvesting opportunities for IIoTDs. In consideration of the offloading position selection, devices resource allocation and system performance, we aim to minimize the overall service latency of all IIoTDs consisting of task computation latency and offloading latency, by joint optimizing the task offloading decisions, charging resources allocation, connection management, and UAV computation resources allocation. However, the formulated optimization problem is a mixed-integer nonlinear programming (MINLP) problem which is challenging to solve in general. In order to address the problem, we decompose it into multiple convex sub-problems based on block-coordinate descent (BCD) method to obtain the optimal solution. Performance evaluation demonstrates that our scheme outperforms the existing schemes in terms of the overall service latency of IIoTDs.