Optimizing resource utilization in cloud and fog computing environments using Hidden Markov Model

Abstract

Abstract The massive volumes of data IoT devices generate are processed on the cloud. Fog computing, a hybrid cloud, and IoT solution, is used to distribute workloads and allocate resources efficiently, but it requires much research to implement. The study investigates load balancing in fog and cloud computing environments, which is crucial for managing IoT data. Cloud computing centralizes data processing and storage in remote storage facilities, whereas fog computing decentralizes functions to intermediary nodes closer to data generation and consumption. This close proximity facilitates optimal bandwidth usage, decrease delay, and faster computing. This study uses Hidden Markov Model (HMM) to examine task behaviour’s across diverse computing nodes and consider resource requirements. This study captures the probabilistic nature of load distribution by constructing emission and transition matrices from observed task dynamics and node-specific information. This research demonstrates the effectiveness of HMM in describing and optimizing load-balancing tactics, HMM play a vital role in optimizing load balancing tactics, as they model resource utilization, providing insights into enhancing resource allocation efficiency within complex computing infrastructures. The accuracy of the proposed method is approximately 92%.