Abstract: Cloud computing is a model for enabling a convenient, on demand network access to a common shared pool of configurable computing resources that can be immediately provisioned and released with least effort . The technologies used in cloud have provided ample opportunities for scalability, cost efficiency, reliability and high resource utilization. Many applications deployed on clouds are real time in nature .Performance of real time cloud applications depends on the generated results and time at which result becomes available .Scheduling has essential role in mapping real time tasks to machines such that deadlines and response time requirements are satisfied. Agent based scheduling technology is flexible to meet different requirements while scheduling. The agent based systems facilitate the interaction between processes by cooperating, coordinating and negotiating with each other. Based on this approach scheduling mechanisms are implemented and corresponding dynamic scheduling algorithms for real time tasks are executed in cloud. In this paper an approach for real time scheduling is implemented for virtualized clouds. This concept is based on agent based scheduling with deadline and cost constraints. Based on these constraints a bidirectional announcement bidding mechanism and probability strategy are developed that will give an improved scheduling of tasks. Investigated the problem of agent-based scheduling for independent real-time tasks in virtualized cloud environments and proposed mechanism for improved scheduling based on deadlines and cost constraints.

Keywords: Real time scheduling, cloud, virtual machines, bidding, cost effective ratio, market approach.