Intelligent Resource Allocation in Cloud Using JADE Agents

Abstract

Cloud computing environments require intelligent and adaptive resource allocation mechanisms to maintain stable performance under continuously changing workloads. This work presents an agent-based resource allocation model developed using the JADE framework to simulate cloud behaviour and automate task distribution. The system consists of the Manager, Resource Client and Ping/Monitoring agents, each performing independent roles while coordinating through FIPA-ACL communication. Resource Agents periodically update their CPU load values, enabling the Manager Agent to identify the least- loaded machine and assign incoming tasks dynamically. The model reduces overload situations, minimizes idle resource time, and ensures balanced utilization across virtual nodes. An HTTP-based task submission interface and a web dashboard are integrated to allow external input and real-time monitoring of agent activity. Experimental results show improved load distribution, faster response to workload changes, and stable system performance under varying demand levels. The study demonstrates that multiagent coordination provides a practical and efficient approach for intelligent cloud resource management using JADE.