Multi-agent Systems for Resource Allocation in Cloud Computing
Gopal Krishna Shyam, Priyanka Bharti · 2023
In the field of cloud computing, resource sharing, and task prioritization are primary problem domains. These problems can be resolved with efficient resource allocation. An efficient resource allocation enhances the resource utilization of cloud systems. But, the hierarchical design of cloud systems (with many devices communicating and behaving in complex ways) makes designing and stability of resource allocation algorithms difficult. Software-based agents have the important characteristic of being autonomous, as they adapt to different conditions based on requirements. Multi-agent systems aid in coordination among several agents. In this article, the algorithm we present develops a unique method for an agent system-based resource allocation in a cloud environment with reinforcement learning techniques. The algorithm shows an improvement in resource utility, resource availability, and resource scalability over static resource allocation. Compared to the multi-agent systems work in the field, the algorithm shows a good improvement with regard to these performance parameters.