Research On Model And Algorithm Of Fuzzy Cloud Computing Resource Scheduling

As one of the key problems in cloud computing system,efficient scheduling approach of resources in complex and changeable environment remains a challenge.Scholars have established various models for the problem of cloud computing resource scheduling,and intelligent optimization algorithms have been widely used to solve the models.However,most of their studies do not take into account problems in the uncertain environment,and the performance of current algorithms can still be optimized.In this paper,fuzzy methods were used to establish more realistic models,and on the basis of existing algorithms,two better scheduling algorithms were proposed.In this paper,triangular fuzzy number was used to represent the task execution time,and a fuzzy model with time-cost constraints was established.The decision variable is corresponding relation matrix between tasks and virtual machines,and the objective function is to minimize the total execution time,and total cost.On the basis of chaotic ant colony algorithm,infinite folding mapping and elite strategy were introduced to propose an adaptive elite chaotic ant colony algorithm(ECACO).The experimental results show that the existence of uncertain factors reduces the efficiency of cloud resource scheduling,so it must be considered.And in the comparison of the algorithms,ECACO outperforms the others in both small-scale and large-scale scheduling.In order to express the uncertain values more accurately,Z-number was introduced in the later research to represent the execution time of tasks,a cloud computing resource scheduling model based on Z-number was established,and the ranking method of Z-number was given.On the basis of ACO,three strategies were introduced to propose a hybrid ant colony algorithm(HACO).The strategies are heuristic rules based on minimum idle strategy,state transition rules and local search.Simulation results show that compared with other four algorithms,HACO have better performance in solving ability,convergence rate and system load balance.In order to analyze and validate the models and algorithms proposed in this paper,an open source framework was used to develop a cloud computing resource scheduling simulation algorithm platform.The platform is universal and scalable,integrates different cloud computing resource scheduling models and algorithms,and it can be used to compare the performance between different models and different algorithms conveniently,so as to provide experimental basis for models and algorithms performance analysis.