Research On Heterogeneous Cloud Task Scheduling Strategy Based On Improved Fireworks Algorithm

With the development of cloud computing,the scale and quantity of cloud platform data center are also growing rapidly.High energy consumption has become one of the research focuses in the field of cloud computing.Task scheduling is one of the important technical components of the cloud platform,reasonable scheduling strategy can effectively optimize the energy consumption of cloud platform.However,the previous task scheduling algorithms only focus on minimizing task execution time to optimize energy consumption.For heterogeneous clouds composed of physical machines with different computing performance,these algorithms ignore the heterogeneity of the cloud platform,which may result in the consumption of more energy consumption of the heterogeneous cloud platform.In task scheduling algorithm,swarm intelligence algorithm shows excellent performance in solving scheduling problems,fireworks algorithm is a new emerging swarm intelligence algorithm.Its good explosion mechanism and mutation strategy make the algorithm search more thoroughly in the scope space and have better optimization performance,it has obvious advantages in solving scheduling and combination problems,and it is very suitable for task scheduling in cloud platforms.so it is the preferred algorithm in this thesis.According to the above analysis,this thesis considers the impact of physical machine type and task execution time on the energy consumption of cloud platform,and proposes a new task scheduling strategy combined with fireworks algorithm to minimize the energy consumption.The main work done is as follows:(1)The model construction of heterogeneous cloud task scheduling and the analysis of the factors influencing the energy consumption of cloud platform.Firstly,the energy consumption composition of cloud platform is analyzed,the task scheduling energy consumption model of heterogeneous cloud platform is constructed,and the relevant variables in the model are described.The model can quantify the energy consumption of heterogeneous cloud platform,and then the task scheduling optimization problem is established with the goal of minimizing energy consumption;Finally,experiments are designed to analyze the impact of physical machine type and total task execution time on the energy consumption of heterogeneous cloud platform.(2)The improvement of fireworks algorithm.This thesis analyzes the working principle and performance of fireworks algorithm.In view of the problem of waste of computing resources and poor convergence,the population of fireworks algorithm is classified according to the fitness value,and a new calculation method of explosion amplitude is designed,and the mutation and selection strategy are optimized,the benchmark function is used to test the improved algorithm.The results show that the improved algorithm has better convergence and optimization performance than the original algorithm.(3)The research of heterogeneous cloud task scheduling method based on improved fireworks algorithm.According to the discretization of scheduling problem,the corresponding coding and decoding scheme is designed to initialize the fireworks individuals and the matching of task scheduling results,and the adaptability of each operation process of the algorithm is improved to solve the optimization problem.Finally,the comparison experiment is set up on GpuCloudSim,and the task scheduling algorithm proposed in this thesis is compared with the initial fireworks algorithm,round-robin algorithm and genetic algorithm.The experimental results show that the proposed algorithm has better performance in energy consumption optimization of cloud platform than other algorithms,and can effectively reduce the energy consumption of heterogeneous cloud platform.