IJCSNS International Journal of Computer Science and Network Security, VOL.11 No.3, March 2011 103 Manuscript received March 5, 2011 Manuscript revised March 20, 2011 Dynamic Scheduling Reduction Algorithm for Distributed Processing Resources through Effective Factors Using Arash Ghorbannia Delavar † ,Mohsen Nejadkheirallah †† , Mohsen Khajeh naeini ††† and Reza Sookhtsaraei †††† † Computer Science Department, Payam Noor University, Tehran, Iran †† Allameh Mohaddes Noori Institute of Higher Education Noor, Iran ††† Computer Science Department, Faculty Payam Noor University, Tehran, Iran †††† Information and Communication Technology Engineering Department,Payame Noor University, 19395-4697 Tehran, I. R. of IRAN Summary By incorporating selected parent’s chromosomes with the effective parameters, we could present a new dynamic scheduling reduction algorithm for distributed processing resources through effective factors using genetics algorithm. Combining the chromosomes, a competence function is formed which has optimal efficiency compared with the previous samples, for this purpose, first a list of dependent dynamic tasks with the complexity degree of O(nlogn) will be assessed. We need to integrate the fitness function and prioritization function to reduce the time waste of the processing resources while running the tasks list. With the help of proposed algorithm of the time parameters, cost, threshold and integration of chromosomes have been applied in addition to the previous affective parameters. This is done when in several simulation of the DSGR algorithm, efficiency and reliability is increased compared to the previous modes of the technical approach and has created this threshold methods to establish the chromosome integration through a more appropriate method. Finally, through a careful evaluation of the objective function, we managed to have going and returning time in a more favorable conditions than the previous algorithms, and to use less time to choose the Selected chromosomes. Key words: Genetic Algorithm; Scheduling; Grid Systems; DSGR. 1. Introduction In the twentieth century, some case studies has been conducted on the scheduling algorithms whose concern is on the nodes aggregation as priority queue, for this purpose, we must consider the tasks associated with the scheduling algorithms in such a way that waiting queues in the processing resources could be reduced. To achieve the objectives of the distributed systems, we should evaluate the classification in two phases: in terms of software, the first phase is presented by the exact use of the previous algorithms of the new optimal algorithm which takes place with more reliability than the previous modes, the proposed algorithm must be in the similar environmental conditions in order for the obtained output to be optimum compared to the previous mode. But in the second phase, in terms of hardware, the network communication module and the speed of the processor which, considering the above conditions, can reduce the implementation time of the independent processing processes compared with the similar processes, and reduce the time allocated to the plan implementation process using the proposed algorithm and be able to optimize plan implementation faster by the above output algorithm [2]. In this paper, a new algorithm is offered for the optimal scheduling of the tasks on different sources based on the genetic algorithm divided based on the gender of the genes provided and managed to improve the makespan time and increase the grade environment efficiency by mentioning the indices and parameters effective on the previous algorithms on the fitness function and controlling the number of the created generations By the genetic algorithm. In this algorithm, roulette wheel schema is used to create the first generation of the chromosomes and Gridsim tool is used to perform simulations. 2. Literature review We first reviewed the related tasks on dynamic tasks scheduling in the computing grid. Then, we introduce our unique approach in solving the reconstruction problem and discovering in scheduling the grid tasks. As many researchers have different opinions in the dynamic scheduling tasks. The most common real function of the tasks scheduling problem is the makespan function. However, in the grid, makespan of a non-optimal scheduler might be much longer than the optimal makespan, because the computing power of the networks changes over time [10]. In the past, task scheduling was mainly proposed for supercomputers and real time and parallel computers in [7], [8], [9], [10].