Priority-Based Task Offloading of Cooperative Edge Computing for Security Monitoring IoT System

Mobile edge computing (MEC) is expected as a promising technology toincrease the computational capability of security monitoringinternet of things (IoT) system which integrates various artificialintelligence (AI)-based applications such as intelligent videoanalysis, dynamic target tracking, etc., However, due to the complexwireless environment and dynamic computing demands, MEC may sufferinaccurate task offloading issue and result in performancedegradation of security monitoring IoT system. In this paper, wefirstly design a cooperative MEC framework for IoT devices andunmanned monitoring vehicles (UMVs) collaboratively to process taskswith different priority levels. Then, we propose a priority-basedtask offloading method to maximize the system utility under energyconstraints of IoT devices and UMVs. Given that the policy of taskoffloading depends on the system state, we formulate the taskoffloading problem as a Markov decision process (MDP) and propose anadaptive long-term optimization algorithm to solve it. Suchalgorithm combines deep deterministic policy gradient (DDPG) and areward estimation method in case the reward function may be lost dueto communication errors. Finally, numerical results validate theeffectiveness and superiority of our proposed scheme compared totraditional algorithms.