Design And Implementation Of Cloud Computing Scheduling Method Based On Heuristic Algorithm

Cloud computing is a fresh model of service based on the Internet.It attracts many researchers and enterprises' attention because of its scalability,high reliability,low cost and on-demand services,and is a hot issue today.Cloud computing scheduling is an important part of the cloud computing research.Traditional task scheduling algorithms include FCFS algorithm,polling algorithm,Max-min algorithm,Min-min algorithm and so on,these algorithms are simple and easy to use,but easy to waste resources and lead to low efficiency of task processing.As the cloud computing platform has a large number of hardware,resources are heterogeneous and complex,resources and service type required by users are also diverse,it makes the scheduling strategy in cloud computing platform is more complicated than the traditional scheduling,and needs to consider resources' adjustment and coordination in multiple levels,also need to be able to quickly respond to user request.Therefore,it is very important to design an efficient scheduling strategy to solve the above related problems.The scheduling objectives of task scheduling in cloud computing platform include:shorter task completion time,load balance,execution cost and so on.First of all,aiming at the completion time of the task,to achieve the shortest task processing time,two kinds of cloud task scheduling methods are proposed,which are improved ant colony optimization algorithm,and improved artificial bee colony algorithm.The ant colony optimization algorithm,is improved in two aspects of pheromone evaporation coefficient and pheromone update method;For another algorithm,the select probability of honey model and bee search strategy are improved and optimized respectively.Then the multi-objective scheduling strategy of improving the genetic algorithm is proposed.The strategy can be handled according to the user's preference to achieve different processing effects.Finally,all the algorithm are tested on CloudSim.The result show that the improved ant colony optimization algorithm can effectively reduce the total processing time of tasks.And so do the improved artificial bee colony algorithm.And in multi-objective task scheduling experiment,the improved genetic algorithm also shows better performance than standard genetic algorithm.