Adaptive neural fuzzy inference system-based scheduler for cyber–physical system

Task scheduling is one of the challenging research problems in distributed computing, especially in a complex scenario like a cyber–physical system. The cyber–physical system consists of a physical system and cyber system which are operated on the different domain of response time. The performance of the scheduling algorithm under the cyber–physical system depends on both cyber and physical factors. But both the factors are unpredictable one in reality which makes the scheduling a challenging one. This paper proposes a fuzzy logic controller based on an efficient scheduler which tackles the above problem. The proposed dynamic scheduler involves three scheduling algorithms which will be selected by the fuzzy controller based on the dynamic behavior of a cyber–physical system. A neural network is incorporated into the fuzzy controller to provide the learning capability, and an adaptive neural fuzzy inference system (ANFIS) is designed. The simulation results demonstrate the superiority of the proposed mechanism in comparison with the existing one.