The Research And Implementation Of Bank Real-time Task Scheduling Based On Genetic Algorithm Under Cloud Computing Environment

Cloud computing will connect a large amount of computing resources,storage resources and service resources through the network to form a resource pool.And then it uniformly schedule and manage the resources according to the needs of users.How to schedule resource timely and efficiently in the resource poll to meet the various requests of users on the quality of cloud services,has been a very important part of the cloud computing study.The current real-time task scheduling algorithms only focus on satisfying the real-time demands of user tasks and they are not flexible enough,they can't well adapt for real-time change in heterogeneous systems.In this thesis,with the help of the characteristics of global optimization of genetic algorithm searching,from the point of the users'real-time demand and the overall throughput of system,we design the fitness function based on real-time and overall throughput.Therefore we can apply the scheduling based on genetic algorithm to real-time task scheduling environment.It can not only adapt to the real-time change of cloud computing environment,but also guarantee the overallefficiency of the task.Aiming at the existing problems of slow convergence speed of genetic algorithm,we propose the optimization strategy based on resources perception.According to the size of the task load,resource usage and the type of task(I.E.CPU-intensive or I/O-intensive),we guide the process of convergence of genetic algorithm.Thus we can accelerate the convergence process.Then with the comparison with traditional real-time task scheduling algorithms under the condition of different parameters,we verify that the improved genetic algorithm has a better efficiency in the task completion and resource utilization.Real-time forecasting system predicts trend of threshold value and the use of resources through machine learning method,and has great reference value to discovering the system fault and bottleneck.The proposed scheduling algorithm can better adapt to this kind of system which has a large amount of data and computational cost and needs better real-time property.Finally,the improved genetic algorithm and traditional task scheduling algorithms are both used in the bank real-time forecasting system for task scheduling.Through the contrast of the task completion,the system overall running time and system energy consumption,it can bee seen that the proposed algorithm has a higher efficiency in actual application.