A Multi-agent Matchmaker Based on Hidden Markov Model for Decentralized Grid Scheduling

Abstract

Design of scheduling algorithms for a heterogeneous computing system interconnected with an arbitrary communication network is one of the actual concerns in distributed systems research, due to the heterogeneous nature of resource, computing capacity, memory, devices, internet and institutional policies. Multi-agent systems have been a nice option of solution because of its distributed nature, autonomy, solidarity and proactivity. Though, in works related, is ignored the lack of control that the scheduler has the resources. We show the process of designing of a scheduler based on multi-agent systems. We used for the development, approach is proposed by Telecom Italia Lab, because the system was developed using the Java Agent Development Framework (JADE) because it meets all the standard of Intelligent Physical Agents Foundation. For the matchmaker process, we propose to use hidden Markov models because it facilitates the prediction of availability of the grid's resources and can evolve over time.