Profit Maximization for Resource Providers Using Dynamic Programming in Edge Computing

Abstract

Edge Computing provides mobile and Internet of Things (IoT) application users with a new distributed computing paradigm that allows them to offload their tasks at edge servers. This way, the application users can reduce latency and energy consumption. To realize edge computing services in practice, it is necessary to design a mechanism that motivates edge service providers by promising higher utility in return (higher payment) for providing edge computing facilities. So, in this paper, we formulate an optimization problem called application assignment for servers’ profit maximization (SPM-AA) that aims to maximize servers’ profit in the system. First, we show that it is NP-hard to find the optimal solution for SPM-AA. Then using a dynamic programming approach, we propose a distributed algorithm using dynamic programming in a synchronized edge computing environment for maximizing servers’ profit called SPM-AAA. Finally, we conduct simulations and compared existing schemes to demonstrate the performance improvement of our algorithms in terms of servers’ profit, the number of applications assigned, server utilization, and resource request service ratio.