Cost Optimization for Allocation of Tasks to Processors in a Distributed Processing System

AbstractDistributed Processing System (DPS) consists of one or more applications spread over several computer systems to solve some computational problems. Task allocation refers to the allocation of tasks to processors before commencement of their execution. The challenge of allocation is predicting the execution behavior of a system and partitioning tasks to processors in a way that there should be maximizes resource utilization by minimizes processing cost. Regarding static task allocation also wide spectrum of techniques, including branch and bound, integer programming, searching, graph theory, randomization, genetic algorithms, and evolutionary methods have been given. The objective of this paper is to evaluate the optimality in terms of processing cost taken to allocate tasks in a DPS. Static allocation on a DPS to minimize the processing cost is a well known problem in parallel & distributed processing applications. In the present research paper we have taken ‘n’ processors and ‘m’ tasks to allocate them on the processors in DPS, where the number of tasks is always greater then number of processors.