Intelligent context-aware fog node discovery

Abstract

Fog computing has been proposed as a mechanism to address certain issues in cloud computing such as latency, storage, network bandwidth, etc. Fog computing brings the processing, storage, and networking to the edge of the network near the edge devices, which we called fog consumers. This decreases latency, network bandwidth, and response time. Discovering the most relevant fog node, the nearest one to the fog consumers, is a critical challenge that is yet to be addressed by the research. In this study, we present the Intelligent and Distributed Fog node Discovery mechanism (IDFD) which is an intelligent approach to enable fog consumers to discover appropriate fog nodes in a context-aware manner. The proposed approach is based on the distributed fog registries between fog consumers and fog nodes that can facilitate the discovery process of fog nodes. In this study, the KNN, K-d tree, and brute force algorithms are used to discover fog nodes based on the context-aware criteria of fog nodes and fog consumers. The proposed framework is simulated using OMNET++, and the performance of the proposed algorithms is compared based on performance metrics and execution time. The accuracy and execution time are the major points of consideration in the selection of an optimal fog search algorithm. The experiment results show that the KNN and K-d tree algorithms achieve the same accuracy results of 95%. However, the K-d tree method takes less time to find the nearest fog nodes than KNN and brute force. Thus, the K-d tree is selected as the fog search algorithm in the IDFD to discover the nearest fog nodes very efficiently and quickly.