A Novel Genetic and Intelligent Scheme for Service Trading in IoT Fog Networks

Abstract

The evolution of the current centric cloud to distributed clouds such as fog presents a suitable path to counteract the intolerable processing delays for time-critical applications. It is anticipated that more fog nodes (FN) will be connected to the IoT paradigm to improve the quality of service and meet the requirements of emerging IoT applications. Typically, the owner manages these FN nodes opening up promising doors towards new business opportunities. Thus, this paper considers fog computing driven network that consists of a set of FNs, distributed on the network edge to serve cloud clients.  Cloud service provider (CSP), in turn, can offer new services, define a profile for each service, and set generate revenue.  However, new schemes should be developed to make this dynamic business model economically feasible. In this context, we propose a new intelligent scheme for service trading, in which a new genetic algorithm is developed for selecting a set of optimal clients that maximize CSP’s profit using game theory for setting the service price. Game theory captures the conflict between cloud clients and CSP, where clients and CSP try to maximize their respective utilities. While CSP attempts to maximize profit, each client tries to negotiate for a lower service price.  Simulation results stress that the CSP can maximize profit by utilizing computational resources efficiently and selecting service requests with the highest possible bid.