FogPrime: Dynamic Pricing-Based Strategic Resource Management in Fog Networks

In this paper, the problem of strategic resource management in fog networks is discussed while considering a pay-per-use model, similar to that used in cloud. Fog networks are distributed in nature, because of which resource management in these networks is an NP-hard problem. In the existing literature, the researchers focused on resource management in fog networks, while considering the network delay constraint. However, none of these works considered the effect of pricing policy while deciding on resource allocation. Hence, there is a need for pricing-based resource management in fog networks. In this work, we proposed a dynamic pricing-based resource allocation scheme, named FogPrime, for analyzing the trade-off between the service delay and the associated price. In FogPrime, we use dynamic coalition-formation game to decide the resource allocation strategy locally within a cluster. On the other hand, we use utility game to choose the fog nodes, strategically, while considering the aforementioned trade-off. Through simulation, we observed that FogPrime outperforms the existing schemes in terms of satisfaction of the involved entities - the end-user and the fog nodes. Using FogPrime, the satisfaction of the end-users and the fog nodes increases by 24.49-47.82%, respectively. Additionally, we observe that FogPrime ensures an even distribution of profit among the fog nodes and enables the end-users to pay less at most by 15.88-47.27%.

Automated summary

This paper discusses the problem of strategic resource management in fog networks under a pay-per-use model, proposing a dynamic pricing-based resource allocation scheme, FogPrime. Through simulation, it was found that FogPrime outperforms existing schemes in terms of satisfying end-users and fog nodes, leading to increased satisfaction levels and profit distribution among fog nodes, as well as reduced costs for end-users.