Joint Security-Reliability Analysis and Optimization for Time Sensitive Mobile Edge Computing Networks in Smart Grid

This paper considers both the physical layer (PLS) and reliability in time sensitive MEC network for smart grids. Specially, sensors are deployed to monitor the electrical devices and offload the collected confidential task (communication phase) to the MEC server for calculating and decision-making (computation phase). In the meantime, an eavesdropper attempts to intercept the offloaded task during the communication phase. The whole process is designed to fulfill the demand of low-latency, ultra-reliability and high-security. Therefore, the joint reliability-security performance for the communication phase is characterized with finite blocklength (FBL) codes, and the delay violation performance for the computation phase is modeled based on the queue tail distribution. Then, the end-to-end security-reliability (leakage-failure) performance is characterized. To minimize the end-to-end leakage-failure, an optimal framework for time allocation is presented. In particular, we prove the convexity of proposed framework with respect to the time length of the communication phase and the computation phase. Numerical results validate the convexity proven in our analytical and demonstrate the impact of various setups on end-to-end leakage-failure performance.