Bi-directional trust management system in fog computing using logistic regression

Abstract

Fog computing is a decentralised computing infrastructure that brings data, storage, computation, and communication resources closer to end users by extending typical cloud computing services to the network edge. A fog node can serve another fog node based on their processing power allowing fog-to-fog interaction. Fog nodes, being independent must be trusted for delegation because they collect sensitive data and share with other discrete fog nodes, where standard cryptographic solutions are ineffective against the internal attacks tossed by rogue fog node. This paper proposes a Bi-directional trust management system for secure transactions and fog-to-fog collaboration to address this problem in a fog environment, which allows a service requester to assess a service provider’s trustworthiness and the service provider to assess the service requester's level of trust before beginning a connection. This trust management system works based on the recommendation system, which is estimated using logistic regression by fog service provider and subjective logic by fog service requester for the establishment of secured connection between them. Using quality of security parameters, the proposed work yields the result of decision making between the fog service requester and fog service provider.