Research On Video Monitoring Data Management Technology Of College Entrance Examination Based On Cloud Environment

In recent years, with the rapid development of computer networks and transmissiontechnology, video surveillance technology also will be rapid development. Due tosocio-economic development led growing industry demand for video surveillance applications.Application requirements showed the following significant features: Applications more and morewidely; Increasing scales; Clarity increasingly demanding; Monitoring time is getting longer andlonger.Since entering the21st century; the development trend of monitoring technology is digital,networked, intelligent and mobile. The digitial means information flow transition from analog todigital state, which change the structure and the operation of system. Networked has achievedshare between network systems resources, tasks and the load. Intelligence mainly used tointelligent control, the analysis process of event, and intelligent management. Mobile is theinevitable trend of development along with the popularity of wireless networks and mobiledevices. The several development directions of video surveillance above need higher storagerequirements. Compared to the data storage, the storage capacity of the video monitoring is massstorage; stability of the storage requirement is to ensure safety and integrity of data in the state ofcontinuous write.Nowadays the Internet has become an essential part of the social structure. Network iswidely used in aspects of society. As one of the networks basic part, computer developed rapidlyas well. CPU speeds continue to improve: The launch of the2011CPU has100billion transistors,1billion instructions per second executable; PC will reach processing power of the originalhigh-performance server computer; high-performance computers using advanced data storagetechnology and achieve frequency of over1GHz clock. Greatly improve of all espectsperformance of computer makes it possible to realize the distributed storage of massive videodata by network. Cloud storage and distributed storage collaborated lots of idle network storagecapacity to achieve collaborative storage; application of distributed storage can effectively usethe high performance network resources to solve the storage problem of video data.On one hand, video surveillance storage brings problems to the storage, on the other hand,put forward higher requirements to the real-time transmission. In order to realize fast access tothe video data in different geographic positions of users, take the following methods, video datacompression coding, distributed storage, multiple replicas, streaming media transmissionprotocol and other measures to ensure the quality of service.The work of thesis mainly in the following aspects:(1) Research on status of monitoring, distributed system, cloud storage, replica allocationproblem and video streaming application protocol at home and abroad.(2) In-depth research focus on the distributed storage, analysis of the classification anddevelopment of distributed storage and video processing technology. Then proposed the distributed storage management system based on the hierarchical, designed the system functions,detailed document add and delete algorithm of the system files.(3) A detailed study of the existing replica allocation and replication management strategy,and proposed a replica tree strategy based on partial cluster optimization. Realized centralizedand structured management of large number of unstructured replicas in network. And theexperimental results showed when the node failure the structure is still stability, reducing thereplica redundancy.(4) Comparing the characteristics of RTP/RTSP/RSVP, proposes a multi-pathRSVP,changes RSVP from use routers to maintain state to ues multipath reservation guaranteetransmission; improve poor scalability problems when data transmission and the routermaintenance of path state more overhead. Simulation results show that the proposed multipathreserved strategy reduced the router maintenance overhead, and to some extent, improved thetime delay problems caused by network congestion or node crashes.(5) On the basis of research on some relative technologies, according to the specific needsof the college entrance examination video surveillance, designed the distributed managementsystem of college entrance examination video monitoring data based on cloud storage, anddesigned the functions of each hierarchy of the architecture.The innovation of this paper has the following main points:(1) Put forward a new strategy of replica allocation named replica-tree.After make thecalculation and analysis of network access; obtain the replica allocation node set. Thenestablished a replica tree through partial cluster optimization. The experimental verified strategycan some extent maintained the system’s stability when node fails.(2) Proposed improved program of RSVP named MS-RSVP (multi-state RSVP): Accordingto the poor scalability problems and the problems which complete negotiation and reservationrepeatly. Between end to end set up states that can provide service of specific quality, reduce therouter’s overhead in the path state maintenance. The experimental verified the improved RSVPcan shortening the delay in waiting time when the network is unstable.(3) Designed the system-SNVM (System of Network Video Management). Realizedremote monitoring processing, distributed storage, backup of the system. For the multilevelmonitoring is easier to appear single-point-failure, put forward a method of the interconnectionand multi-point backup, improved the robustness of the system.