Research On Distributed Task Scheduling Method And Parallel Optimization Method

Distributed computing has been widely used in network transmission,data processing and other fields.With the popularity of distributed computing,it saves a lot of time and space when processing information data,and also makes the computing efficiency of data processing much higher than before.Among them,reasonable task scheduling and efficient data processing methods are the key to improve the computing efficiency.The current design methods of task scheduling in many fields mainly include task scheduling algorithm based on distributed framework,traditional task scheduling algorithm based on single objective optimization and intelligent thinking method based on multi-obj ective.Parallel processing method is mainly based on distributed computing framework,through cluster management,to reduce the cost of hardware.Although most of the task scheduling methods and data processing methods have achieved good results,but in the aspect of task scheduling optimization and video data serial order processing caused by low computational efficiency,it still needs to be improved.Based on the idea of parallelization,this paper uses improved ant colony optimization algorithm and speculative parallel video decoding scheduling method respectively to solve the problem of task scheduling optimization and low computational efficiency caused by serial sequence processing of video data,and verifies them through comparative experiments.The research,contents of this paper are as follows:(1)Aiming at the problem of task scheduling optimization in distributed computing technology and low computing efficiency caused by serial processing of video data,this paper summarizes the research status at home and abroad,analyzes the advantages and disadvantages of different technologies and methods,and puts forward corresponding solutions to solve the two problems.(2)An ant colony optimization algorithm with adaptive parameter control is proposed.On the basis of the original algorithm,an adaptive parameter control method is used to improve,and the parallel virtual cluster simulation tool Cloudsim is used to simulate the improved ant colony optimization algorithm.Experiments are carried out under the condition of the same computing scale,by comparing the experimental results of other intelligent thought algorithms,the effectiveness and scalability of the improved ant colony optimization algorithm in solving task scheduling problems are verified.(3)The proposed scheduling algorithm is verified based on video decoding scheduling.In the computer cluster environment,the original video coding data arranged in serial order are divided and tested by this method,and the parallel decoding processing and result merging are carried out.Under the condition of the same computing scale,by comparing the experimental results of decoding in serial order,it is verified that the speculative parallel scheduling method is effective,reliable and scalable in solving the problem of low efficiency caused by serial data processing.