Optimization of resource allocation using Monte Carlo

Abstract

In this paper, we study the problem of optimizing the error generated in the allocation of processing elements to tasks, based on real-time scheduling. The makespan of the tasks to be scheduled is the same, in the computational grid. Tasks to be assigned to processing elements are randomly generated and are mapped to processors. Scheduling is done, considering real-time environment and mapping is achieved via biased and unbiased mapping of tasks to processors. The heuristics are used with the simulated computational grid and the results are compared extensively against each other.