Distributed System Design And Algorithm Of Image Processing On Satellite Cluster

With the development of information technology,people have entered into the era of big data,it has become a consensus to build a cluster of distributed computing and storage for huge amounts of data,as well as to space satellite fields.lt is very meaningful to set up a distributed architecture of satellite cluster.At the national level,in the face of a sprawling complex surrounding situation,such as the United States and South Korea deployed Missile Defense system,if we achieve satellite network for distributed computing and storage,we can significantly improve the ability of the satellite,and extend the function of satellite,in addition it will play an important role on the defense investigation,situational awareness,map detection and communication demand.On the public level,people can access satellite networks at any time through mobile devices,thus they can enjoy more accurate real-time positioning,logistics tracking and remote sensing information services provided by satellite system.In recent years,the rise of Internet big data application has made distributed technology widely applied in engineering practice,which mainly includes distributed storage,computation,database and monitoring.Distributed storage technology provides the basis for data storage,ensuring redundancy and consistency of data;Distributed computing improves computing speed and provides the possibility of mining mass data.Distributed database can store multiple types of unstructured data,providing SQL interface for large data transaction processing;The distributed resource monitoring system monitors the cluster indicators at all times to ensure the safe and efficient operation of the cluster.Be different from the ground environment,satellite is limited with some constraints such as volume,power consumption and space radiation,key technology such as architecture design,operating systems and resource management should be grasped to establish distributed environment on satellite cluster.To solve above problems,this paper first studys the distributed computing architecture and storage architecture on satellite cluster,secondly studys the implementation of distributed resource monitoring on satellite cluster,In the end,transfers the distributed technology to the satellite platform,sets up the satellite group demonstration system,and applies distributed technology to the image processing algorithm in the satellite to prove its feasibility and superiority.This project is a part of the 13th five-year skybase networked computing platform project,builds a satellite network computing platform demonstration system,studys and verifys the key technologies of distributed computing,storage and monitoring through building deployment of satellite image processing.Through the project development,we can offer work base for the 13th five-year pre-research project,provide technical basis for related product planning and development,provide software technology support for the integration of heaven and earth in the future electronic system.This paper is based on the single image processing algorithm,the distributed computing technology is adopted in the matrix processing to shorten image processing time and improve imaging speed through distributing operation of fourier transform on matrix into muti-nodes.