The Research And Design Of Security Monitoring System Based On Embedded And Cloud Storage

With the development of society and technology, more and more security monitoring system gradually into people’s vision and social. Security has become the most things that be talked about. Security monitoring systems, whose purpose is to use all kinds of modern technology, real-time continuous monitoring and controlling regional monitoring. As the most common anti-theft alarm, gas leak warning, fire alarm systems, their functions are too single and have a serious flaw-can not remote real-time automatic alarm, the guards on duty have to timely discovery of unusual circumstances. To address these shortcomings, as well as to meet the higher requirements of the monitoring system, to achieve the monitoring terminal and upper-level monitoring points have the abilities of two-way SMS communication, I develop an ARM-based remote wireless security security monitoring system. This system combines the V4L2video interface technology and the socket network technology, to design an embedded Linux-based system monitoring terminal, which transmits data and image via thesocket communication with the host PC. Security monitoring terminal SMS alarm to the monitoring station via GSM communication and embedded technology.However, for large, complex security monitoring system, users are concerned about not only the video picture information, often also concerned about some of the other data, these data were user-friendly real-time and in a different location within the database query. Appeared to be inadequate for centralized storage technology, centralized storage for server configuration is also very high, so prices tend to be very expensive and the face of the massive data storage and query operations. If a server fails, will result in the collapse of the entire system. Hadoop cloud storage technology can solve the above problem. Hadoop Cloud storage, distributed database and distributed file system logic is a unified whole, but physically data are stored in different nodes in a different location, so you can use a cheap PC, to store data instead of expensive servers. Cloud storage node network, whose each node has the characteristics of site autonomy, and each node has the ability to process data independently. The biggest advantages of distributed storage is when a node fails in cloud storage networks, the same copy in another node works so as not to cause the failure of the entire system because of a node failure. In addition, it can be operated by selecting the most recent copy of the data of the user, and reduce the communication cost to improve query latency.In this paper, in order to achieve the above-mentioned requirements, do the following work:(First)design based on the Samsung S3C2440chip embedded monitoring terminal. The terminal can be real-time acquisition of video data received through the serial communication, the lower sensor sending to the various data, by analysis of these data exceeds a warning value for alarm.(Second)The client is a Linux-based system, and sent to a variety of data received by monitor terminal through the TCP/IP protocol, receiving the monitoring terminal sends video data through the UDP protocol.(Third)Design a hadoop cloud storage-based distributed database and distributed file system to store a variety of data received by the client program, and optimize cluster load balancing algorithm, which greatly improved the system performance,while using the Map-reduce programming model to implement the parallel computing system is capable of multi-node parallel computing, final I analysis advantage of cloud storage compareing to the general memory.