Research On Models And Algorithms For Multi-installment Divisible-load Scheduling In Visual Sensor Networks

With the development of Internet of Things and smart mobile devices,visual sensor network has been widely used.It can be used to obtain abundant multimedia data.How to deal with a large amount of multimedia data has brought great challenges to both academia and industry.The divisible-load scheduling problem in visual sensor networks is a key issue of processing large data for visual sensor networks.The scheduling strategy in the visual sensor network directly affects the processing time of the multimedia data and the transmission time of the data in the entire network.Therefore,it is very important to seek an optimal scheduling strategy to minimize the completion time of the visual tasks.In this dissertation,the task scheduling problem in visual sensor networks is deeply studied.New scheduling models are established and new scheduling algorithms are designed.The main innovative work is as follows:1.For the divisible-load scheduling problem of visual sensor networks,existing studies only consider the establishment of a single-installment scheduling model for this problem,but when the workload in the problem is very large,single-installment scheduling models may take more time to complete all tasks.Thus it is necessary to set up multi-installment divisible-load scheduling models.This paper first establishes a new multi-installment divisible-load scheduling model which can improve the real-time processing time of multimedia signals without considering machine(processor)order(i.e.,the processor order is given in advance)in visual sensor networks,and then designs a new algorithm for solving this model: NMISA–VSN.Finally,simulation experiments were performed,and the proposed algorithm was compared with the latest three existing algorithms.The experimental results show that the proposed model and algorithm are effective and can get shorter completion time of the visual tasks.2.When taking into account the processor order in visual sensor networks,and also considering the number of processors,the number of installments and the task partitioning strategy,a new multi-installment divisible-load scheduling model is proposed first,and then an improved genetic algorithm is designed to solve the model.Finally,the simulation experiments are conducted and the proposed algorithm is compared with three existing algorithms.The experimental results indicate that the proposed algorithm and model perform best.It can divide the workload more reasonable by using the optimal processor order and get the minimum completion time of the visual tasks.