Multi-agent cloud computing

Cloud computing is a collection of parallel, distributed, and web-accessible resources (subject to service-level agreements) that should be dynamically composed and virtualized based on consumer requirements. Service-oriented computing promotes the loosely coupling of autonomous and independent services to compose solutions from massively distributed systems. Service providers, brokers, and consumers, are self-interested, autonomous parties that should interact and coordinate among themselves to make an efficient use of Cloud resources. The distributed nature and dynamism of service-oriented Cloud systems as well as the self-interestedness of Cloud participants emphasize the need for agent-based solutions. Agents are autonomous problem solvers that can interact with other agents through negotiation and cooperation to achieve their design objectives by using distributed artificial intelligence techniques.

My research is focused on the design, implementation, and exploitation of agent-based problem solving techniques to achieve complex objectives in continuously evolving, distributed, and service-oriented Cloud systems. This research is concerned with developing both theoretical frameworks and practical algorithms to allow agents to achieve both independent and global design objectives in Clouds in an autonomous and intelligent manner.