Research On Collaboration Mechanism Of Client Aggregation Services In The Application Of Dynamic Monitoring By Remote Sensing

The big data and the development of cloud computing make it easier to dataacquisition and data storage. The exponential growth of data, led up to the innovation ofthe scientific method, which transformed a thinking model verifying causality topursuing the regularity of data characteristics. The age of big date is promoting theinformation revolution and driving another industrial revolution. The application andimplementation of the big data, which is used for the management, storage, analysis,management, services, etc, are based on cloud computing. Therefore, managementsystem by big data and cloud computing technology to establish client aggregationservice cooperative work mechanism is of positive significance to the System ofDynamic RS Monitor.The data of Dynamic RS Monitor has the characteristics of big data; quantitativefluctuation of data will cause the qualitative change of problems. For that reason, itreduplicates the difficulty of problem solving in following aspects:(1) It is difficult torealize the general processing method under the condition of wide data resourcesacquisition channels, numerous preconditioning methods and different data sources.(2)The large scale of the data storage resources and its complex management, remotesensing image data with high resolution are beyond the processing capacity of thesingle node graphics workstation.(3)The complex relationship of data resources,regular computing model and system structure are difficult to access and serve.(4) TheDynamic RS Monitor in the practical applications mainly relies on the artificialexperience, in which there are various original image processes and a huge number ofextraction patterns. The low efficiency of information exchange and data sharing in thenational and local coordination cause the difficulty to achieve statistical analysis anddata mining of integrity.(5) Whereas the data resource is classified information, dataprivacy and security need a powerful safeguard mechanism.(6) There are manybottleneck problems in the construction of digital land information platform, theimplementation of the “land map” and the visualization service mechanism of two or three dimensional integration at the data level or the technical level.(7) According to itscharacteristics, the collaborative mechanism of client aggregation service can store datadispersedly, gather information on a real-time basis and serve on demand. Based on this,it provides an application model for dynamic monitoring by remote sensing.In this paper, aiming at the problems of client collaboration and its application inthe dynamic remote sensing monitoring, relying on the related topic, orienting the datalife cycle, the author researches from the methods for data preprocessing under thebackground of big data, the data exchanging and sharing, data security and permissionsmanagement etc., to construct the complex information system through the conceptbased on data and data oriented architecture, to establish of data classification systemand access permissions which marks the data identification as recognition andpositioning data marker and to realize a prototype for data sharing and collaborationservices by the management and exchange of data in data registration center and by theestablishment of various data function unit. This paper mainly investigates thefollowing aspects:1. Research on the functional collaboration preprocessing method under thebig dataResearch on the classification method based on clustering data on big dataenvironment, including the key technology of clustering algorithm, data cleaning, datasampling, data transformation and data reduction, etc; Research on Modeling of typicalremote sensing image data processing, including the selection of different data sourceprocessing model and parameter definition; Research on data classificationmanagement based on high computing capability and high processing capacity of cloudservice, including user authorization and pretreatment of bulk data.2. Research on client collaboration mechanism based on PDPResearch on data exchange specification C2ML for client collaborative, includingdata encapsulation and release criteria, data organization and management technology;Research on aggregation mechanism on-demand of data information, including keytechnologies of cooperative concurrency control algorithm, efficient and reliable datatransmission mechanism, computer-supported collaborative technology, Web servicetechnology and simple object access protocol; Research on message service mechanismto support data collaboration, including key technologies of sharing collaborativeinformation, collaborative access control, cooperative member communication andcollaborative visualization.3. Research on data security model based on the management mechanism of life cycleResearch on authority management based on cloud services, including datasharing, privacy protection and other key technologies; Research on encryption designmethod of data of the whole life cycle, which goes through the whole process includingproduction, storage, transmission and application; Research on the managementmethods of access authorization, in accordance with the data set, the user and the typeof access; Research on nonsymmetrical security strategy of systematic storage,including the key technology of asymmetric distributed storage, high and low level datapartition management etc..4. Research on Application of client collaboration mechanism in dynamicmonitoring by remote sensing technologyResearch on application technology of client collaboration in cloud services,including multi scale representation of multidimensional and interactive visualizationman-machine technology, such as remote sensing metadata management, remotesensing image, elevation data, the geological survey data, vector data, real-timemonitoring data, remote sensing dynamic monitoring map spot etc.; Research onremote sensing dynamic monitoring map spot on large data environment identification,labeling and measurement, including remote sensing data batch processing; Researchon distributed storage strategy, data based on PDP data management and data exchangestandard.5. Research and prototype system for dynamic remote sensing monitoringapplicationsResearch on design methods of the prototype system, including the design anddevelopment of functional modules; Research on data registry and key technology,including data cooperative security model and data management; Research on cloudservice mechanism based on DOA, including the cloud storage model, cloud model,cloud governance model and loud security model.The innovative achievements and contributions of this thesis are as follows:1. Proposes a multidimensional collaborative clustering method of dataclassificationCombining K-Means Clustering (KC) and Fuzzy C-Means Clustering (FCM)algorithm, this paper proposes a multidimensional data classification combinedclustering (KC-FCM) method. KC was used to optimize the FCM algorithm toinitialize clustering center selection problem, using the FCM algorithm to optimizethe clustering center, complete fuzzy clustering, in combination with fuzzy reasoning method is simple, realize the clustering classification.Through the theoretical analysis and programming experiment results verify, KC-FCM algorithm can converge to the global optimal solution less number of iterations,combined with support vector machine (SVM) which can get better training effect,can effectively improve the big environment data classification precision andefficiency, better adapt to big data environment, reduce the complexity of thealgorithm itself, can be used for dynamic data classification of remote sensingmonitoring.2. Put forward a kind of DOA "Peer-Data Register Center-Peer"(PDP)service mode, and puts forward a client synergy mechanism based on the PDPUsing big Data thinking and cloud services architecture, from two aspects toobtain, infrastructure and Data without too much consideration of storage capacityand computing power, such as XML, GML, KML, HGML markup language as theblueprint, this paper proposes a Client synergy logo (Client Collaboration, C2ML)which combining the Data Oriented Architecture (DOA) and C2ML identification,and a Peer-Data Register Center-Peer,(PDP) Message Collaboration Service mode,used for the equality of peer to peer communication between the Client. Any nodeexchanges information through by Data Registration Center (DRC). Combining thePDP and data distributed multi-level storage strategy, we form a highly efficient"scattered data storage, data real-time gathering, on-demand aggregation service"synergy mechanism.Validated by experimental results, the working mechanism to adapt to dataresources which are multiple source/heterogeneous, mass/volume, real-time/dynamic,complex relationship, big total value but low density. Its characteristics are so easy toaccess, to update, to retrieve and scheduling that can effectively improve theefficiency of client synergy in large data environment, provide theoretical guidanceand practical basis for the dynamic remote sensing monitoring prototype systemconstruction, and provide certain theoretical significance and practical value for thedevelopment of land resources informatization and transformation of service mode.3. Proposes a rights management based on total life cycle data, and putsforward a decentralized stored data security model based on asymmetryIn the application of dynamic remote sensing monitoring, the high precisionremote sensing data and information products which require top secret, need to beensured the safety of the storage and application service in the whole life cycle.Combined with Product Data Management (PDM) and Product Lifecycle Management (PLM), this dissertation puts forward a Data Lifecycle Management(DLM) method. In combination with rights management strategy, puts forward amethod of rights management based on DLM. Combining all the methods above, thispaper proposes a asymmetric scattered data storage security mode. The user saveshigh security classification data in the terminal or on its own cloud as require forseparate management, and excutes separate management security policy from lowsecurity classification data in applications. Using its camera obscura characteristic, bymeans of bit division, each data block does not contain local complete information.At the same time, taking full account of data’s attribute of life, through theadoption of the safety of the whole life cycle of data model, the data can be verticalintegration encryption in the process of production; storage, transmission andapplication, and it can only display and interpretation on the client. According to thedata set, the user permission and access types, authorizes dynamic/static data;according to different data and the client users, fine manage the authority of release,modify and application, so as to meet the requirements of the safe and efficient.4. Under the DOA architecture, build a client collaborative dynamic remotesensing monitoring prototype system based on PDP collaborative informaitionservice modeBased on Data Oriented Architecture (DOA), big data, cloud computing,information security and distributed technology, combining with essential data issuedby government (including the second national land survey data, elevation data, DEMdata, etc.) and2012-2013national dynamic remote sensing monitoring data (includingthe remote sensing image data of different resolution, monitoring patch data, etc.),taking C2ML logo as the core, organizing and describing information service in whichdynamic remote sensing monitoring industry application requirements takes animportant part, a prototype system is designed and implemented, through the dataregistry and the PDP collaborative information service mode, implements the clientsynergy and application.The running result proves that on the basis of protecting data security,collaborative efficiency has been improved. This system provide a scientific basis andtheoretical support for implementation DOA applying to dynamic remote sensingmonitoring in large data environment; it also has a certain reference value insystematic development of land and related field; and it is a guide to avoid repetitivefunction development and data redundant construction in informatizationconstruction. Oriented data architecture as a new mode of software architecture, its workingmechanism and service mode is not fully mature, by the way the big data is at theconceptual level and explore stage, related theory, method and technology are still inthe early stage of development. In the application of dynamic remote sensingmonitoring, theoretical system and application mode of big data can be to constantlyenrich and develop. The client collaborative dynamic remote sensing monitoringsystem still has to be optimized. Therefore, in this paper, on the basis of research ofsynergy mechanism prototype system based on the client aggregation service, somefurther study of this topic should be continued. The next step of work mainly includes:(1) Learn more knowledge of digital land, continue in-depth studies of effects ofspace information technology of dynamic remote sensing monitoring, to enrich thecontent of the client synergy service, provide more support services for the dynamicmonitoring and auxiliary decision analysis;(2) Continue in-depth research on Data Oriented Architecture workingmechanism and application framework, combine the latest progress of DOA and theapplication of dynamic remote sensing monitoring;(3) Continue to study the language specification of client synergy logo C2ML,complete C2ML architecture with application requirements dynamic remote sensingmonitoring.