Application Of Ant Colony Optimization Algorithm In Cloud Computing Resource Allocation

In recent years, with the widely application of information technologies in varied fields,the amount of Internet data increases incredibly and the Internet has yet to enter into the BigData Time when traditional computing models cannot satisfy current dynamic needs any more.Accordingly, Cloud Computing emerges. In complex cloud computing environment, however,actual resource distribution is often complicated. It may happens that the number of resourcesis not enough to support the current user operation or the estimated dema nd for resources istoo high to make full use of resources. Therefore, how to reasonably and effectively allocateuser assignments to resources nodes becomes the key issue of Cloud Computing.Ant colony algorithm is a kind of intelligent bionic algorithm. It is based on theobservation of ant’s food hunting in the nature,which finds that ants always find shortest pathbetween their nest and food source. The main principle is ants can release pheromone andpercept its concentration to identify and determine the next step, and release a certainconcentration of pheromone according to the path length after food source is founded. Thehigher the pheromone concentration of the path, the greater chances of the path is choose. Withpositive feedback mechanism and robustness, ant colony algorithm is widely studied andimproved since put forward. As ant colony algorithm shows great advantage in solvingcombinatorial optimization problems, this paper applied the ant colony algorithm to solveresources allocation problem in Cloud Computing. Whereas, ant colony algorithm has itsinherent shortcomings in the specific application, so genetic algorithm and the rule ofMetropolis accept are introduced in this paper to improve ant colony algorithm.In this paper, the main work and innovations are as follows:(1) first of all, this paper will briefly introduce the basic knowledge of cloud computing,MapReduce model, and the resource allocation problem under Cloud Computing environment.(2)Ant Colony Algorithm (ACO) has a defect: it would be a quite long time blind search ifthe path’s pheromone concentration is same. Instead of simply improve the parameters orformula of the original algorithm, this paper brings in the basic idea of genetic algorithm (GA),and applies the rapid global search ability of genetic algorithm into the initial pheromonedistribution. Genetic Ant Colony Algorithm (GAAA) is presented by the fusion and applied toTraveling Salesman Problem (TSP), and then prove the feasibility of GAAA algorithm throughthe experimental results.(3)The pheromone update mechanism introduce Metropolis accept principle of simulatedannealing algorithm for path optimization and pheromone update. According to the rule ofMetropolis accept, some "bad" data processing is accepted and by comparing the targeted function value of the new and initial solution to determine whether to accept data processing.By this way, it can avoid to a certain extent(4) As to the resource allocation problem in cloud computing environment, this paper putsforward the objective function and applies to genetic ant colony algorithm (HGAACO) to theresource allocation problem, and list the steps and process of the algorithm. Algorithmdevelopment is under the Matlab platform. This paper compares and analyzes HGAACO,ACOand GAAA proposed by setting algorithm parameters.