An Incentive Framework for Resource Sensing in Fog Computing Networks
Fei Shen, Guowei Zhang, Chongchong Zhang, Yang Yang, Rong Yang · 2018
Fog computing is expected to excavate and make full use of the inherent idle communication, cache, computation, and control resources of massive devices, and to relieve the pressure of cloud computing on link congestion, delay, and energy consumption. However, how to accurately sense the resources of all fog nodes (FNs) in real time is vital to efficient resource scheduling in the fog computing networks. Frequent sensing will result in both high sensing accuracy at the fog controller (FC) and cost at the FNs. To this end, we propose a novel incentive framework to motivate the FNs to feed back their resource sensing data frequently to the FC based on Stackelberg game. The FC plays as the leader with the sensing reward prices as its strategy, and the FNs play as the followers with the sensing frequency as their strategies. The utility functions of the FC and the FNs are proposed, considering the payment for resource sensing, the accuracy of sensing and the cost of sensing. The existences of the global optimum of both utilities for the FC and the FNs are proved. Closed-form solutions for the optimal sensing frequencies of the FNs are derived. Numerous simulations are done verifying our theoretical analyses and indicating the importance of our proposed incentive framework for resource sensing in the fog computing network.