Research On Adaptive Scheduling Of Remote Sensing Task Based On Tabu Search

As more and more remote sensing satellites are launched in China,the amount of remote sensing information transmitted by satellite is growing geometrically.The remote sensing products are obtained by processing the original remote sensing images,but the production tasks of remote sensing products are longer,the tasks are relatively heavy and the demand is relatively urgent.Therefore,the production of remote sensing products are mostly used in the cluster system in order to improve the processing capacity.How to properly and efficiently allocate a large number of tasks in a cluster so that each type of production task can obtain a better scheduling result and ensure that the system load is maintained in a relatively low state and increase the cluster throughput has become one of the hotspots in the field of computer cluster.However,the task scheduling strategy in the existing remote sensing cluster is mostly based on the task scheduling method of traditional operating system or improved on the basis of its,the scheduling algorithm itself has some limitations.Therefore,it is of great theoretical and practical significance to carry out the research on the task scheduling method of remote sensing cluster to enhance the service capability of remote sensing cluster.The main research work and innovation of this paper are as follows:(1)Aiming at the different resource requirements of different remote sensing task in remote sensing cluster,a resource demand calculation method is proposed which suitable for remote sensing task.In remote sensing clusters,the number of users is variable,but the type of production order is relatively fixed.The task resource demand calculation method proposed in this paper according to the different production order type and remote sensing algorithm dependency which provides the basis for the subsequent task scheduling.(2)Aiming at the problem that the task scheduling effect is not ideal in the remote sensing cluster system,an adaptive remote sensing task scheduling algorithm based on tabu search is proposed.The scheduling algorithm proposed in this paper relies on the calculated demand of remote sensing task resources.Then,the tabu search algorithm is used in remote sensing task scheduling,and the approximate global optimal solution of task scheduling is obtained.Finally,an adaptive remote sensing task scheduling algorithm is designed in combination with a single task scheduling algorithm.In the large-scale cluster environment,this method can effectively improve the cluster task scheduling efficiency and cluster throughput,and can achieve cluster load balancing.(3)Relying on the real project production environment to build experimental platform,and then compare with the current commonly used in the field of remote sensing cluster scheduling algorithm.The experimental data show that the proposed remote sensing task scheduling algorithm has better load balancing effect,it can reduce the cluster imbalance effectively,improve the utilization rate of cluster resources and increase the cluster throughput.