Adaptive Trust Calculation in Fog Computing

Abstract

Fog is well suited for situations where a huge number of decentralized devices must communicate, provide live analysis of data, and perform storage jobs because of its inherent decentralized nature and capacity to process data in transit, i.e., ability to draw conclusions in real-time. Fog computing offers the dependability that time-sensitive smart healthcare systems require because of its ability to operate near the end user and independence from centralized architecture. Because healthcare data is so vital, there is a need for stronger security and privacy solutions for fog computing, where trust is crucial The goal of this research is to provide a context-based adaptive trust solution for the smart healthcare environment using Bayesian technique and similarity measures against bad mouthing and ballot stuffing since context dependent trust solution for fogs is still an open research topic. To assess our findings, the proposed trust model has been simulated in Contiki and Cooja. In contrast to static weighting, adaptive weights assigned to direct and indirect trust using entropy values assure the least amount of trust bias, and calculations of context similarity remove recommender nodes with malevolent intent utilizing server, coworker, and service similarity. Due to its minimal trust computation overhead and linear complexity O(n), this model is effective.