Efficient Resource Provisioning in Fog Computing Using Agent-Based Contract Net Protocol: A Smart Healthcare Case Study

Abstract

In the era of pervasive computing, fog computing has emerged as a paradigm that extends cloud services to the edge of the network, enabling real-time data processing and reducing latency for time-sensitive applications, particularly in the Internet of Things (IoT). However, fog computing faces significant challenges in resource allocation and provisioning due to its limited resources. Efficient resource provisioning in fog nodes is critical, as a well-designed algorithm can significantly reduce response time and energy consumption for application requests. In this paper, we present a novel agent-based resource provisioning algorithm tailored for fog computing environments, utilizing the Contract Net Protocol (CNP) to dynamically allocate tasks to virtual machines (VMs) based on resource availability and task requirements. A smart healthcare case study shows the proposed algorithm reduces network usage by ${2 2. 9 5 \%}$ over SJF and $28.67 \%$ over FCFS, reduces energy consumption by $6 \%$ for both SJF and FCFS, and decreases loop delay by $19 \%$ over SJF and $38.48 \%$ over FCFS.