Based On Ant Algorithm Optimal Grid Scheduling Strategy And Design

Grid scheduling strategy is a core component of grid computing, scheduling strategy will directly affect the performance of grid computing system. However, due to a wide range of grid resources,and it has a heterogeneous,scalable,dynamic self-adaptive characteristics, which makes the traditional scheduling method has been difficult to apply in grid environment. Therefore, Scheduling strategy on the grid and extensive research, it is valuable of both from a theoretical or practical value to the technology itself.Ant colony algorithm as one of the bionic optimization has good positive feedback, robustness and adaptability, which is very similar grid scheduling environment. Many scholars use the ant colony algorithmgrid in Grid scheduling technology,it shows that the ant colony algorithm to solve the grid scheduling problem with inherent advantages. In this paper, ant colony algorithm and the grid scheduling problem in-depth analysis and discussion, and based on the design of grid-based ant colony algorithm for scheduling task scheduling and routing model and implementation.This paper proposes a new algorithm: Adaptive Windows Rank and Screening Strategy algorithm based on the Max-Min ant colony algorithm. Window's valve is adaptively adjusted in the terms of the distribution of cities, and gradually reduces during the operation process. It effectively narrow the search space, and improve the convergence speed of the algorithm.Screening strategy pheromone method emphasizes the optimal path while it ensures the path to explore. This enables this algorithm to avoid trapping in local optimal solution.For the ant colony algorithm is easy to fall into the shortcomings of the local optimal solution, combining with the concept and characteristics of information entropy, this paper proposed a kind of Ant task scheduling model and the realization of grid which is based on Entropy, through GridSimTookit simulation experiments ,it shows that the improved algorithm has more good search capability,and can improve the grid task scheduling time effectively. For such problems as network congestion and delay defects of AntNet(Ant Colony Routing Scheduling), this paper presents a relatively simple calculation of the statistical information, and improves the network routing node, this paper also proposes two different behaviors of the Forward ants, they can change the network load ratio of adaptive regulation of their largely reduced network congestion and delay updates and accelerate the convergence of the algorithm.Simulation experiments through NS2 simulator shows that the improved algorithm has a good network throughput and lower delay.