Robust Risk-Sensitive Task Offloading for Edge-Enabled Industrial Internet of Things

Abstract

Edge-enabled Industrial Internet of Things (E-IIoT) has gained massive attention as a new type of IIoT for hosting emerging low-latency applications. However, due to device variations and complex communication environments, the edge servers and channel information are usually uncertain which poses significant challenges to the computational capacities in practical E-IIoT networks. To address these challenges, in this paper, we introduce a risk-sensitive task offloading scheme for the practical E-IIoT networks. Firstly, a distributed robust offloading optimization problem is formulated by jointly considering the high latency risks caused by network uncertainty and the average latency of the system. Then, by characterizing the latency risk as the conditional value, the formulated problem is transformed into a distributed robust mean conditional value-at-risk (CVaR) optimization problem. Further, we use a fuzzy set to capture the uncertainty of the first and second-order statistical information of the system to overcome the computational difficulties of the problem, and a distributionally robust optimization (DRO) method is employed to transform the mean-CVaR optimization problem into a computable semi-definite programming (SDP) problem. Finally, an improved DRO-based task offloading algorithm is proposed to obtain the optimal decision under uncertain E-IIoT networks. The simulation results show that the proposed offloading scheme not only improves the reliability in computation but also reduces the high latency risks. Therefore, our proposed model is more suitable for practical E-IIoT networks.