Node selection and utility maximization for mobile edge computing–driven IoT

Abstract

AbstractOptimal user association and resource utilization entails a challenging problem of high latency in mobile edge computing (MEC)–driven Internet of Things (IoT) applications. Increased power consumption is another aspect that requires attention, specifically for the systems that involve huge number of users (IoT nodes) and computing devices (cloudlets/MEC nodes/fog nodes). Though MEC/fog networks are designed for low latency and low transmit power connections, yet the substantial increase in IoT devices signify the need to redesign it for future demands. In this article, we formulate an optimization problem related to quality of service–driven node selection and utility maximization subject to power and workload constraints for applications that require very low latency. Outer approximation algorithm is a proven technique in the field of optimization theory and is scalable with number of nodes. A distributed self‐converging algorithm based on the outer approximation algorithm is presented in this work, which proved to be efficient for the problem formulated. Extensive simulations are done to validate the numerical results. This work concludes by comparing results of outer approximation method with matching theory to validate its effectiveness.