The Intrusion Signal Recognition Of Perimeter Security For Distributed The Fiber-optic Senor

With the rapid socio-economic development and the rapid increase in information capacity, optical fiber communication technology to the optical fiber for the transmission of information mainly medium and the theoretical basis of the principle of guided wave fiber optical fiber sensing technology in humans over the past few decadeslife and in many areas has been widely applied. Based distributed fiber sensing perimeter security systems is a new generation of security systems, as opposed to the traditional perimeter security system, it has long distance monitoring passive sensor, high sensitivity, short response time, low energy consumption and adapt to the complex prominent advantages such as the topography, the research focus of the current international and domestic, but how to improve the probability of detection and reduce the false alarm rate of the system is a difficult problem.The core of research work described in distinguish between the various cable vibration incentive target based on signal detection theory and wavelet analysis theory of digital signal processing, in accordance with the theoretical analysis, testing programs, test validation order, distributed fiber sensing perimeter security systemssignal detection in-depth analysis and discussion.The paper completed the following work:(1)The model for strain effect in FPSS with dual Mach-Zehnder interferometer is presented. By researching on the effect of strain, the model for locating technique of the FPSS based on dual Mach-Zehnder interferometer is established. Due to analysis based on the established model, the conclusion is obtained as that the system’s sensing arm and reference arm could be concluded in the same cable, as a result, the cost of FPSS can be reduced.(2) The original experimental result is analyzed, by which the characteristics of the perturbation signal and the noise are presented. Due to the characteristics of the actual perturbation signal and low detecting possibility of energy detector in the condition of low SNR, the signal detecting method, which is realized by threshold measuring energy amplitude of wavelet d7frequency signal is presented.(3) The time-frequency characteristic vector is constructed based on the multi-scale wavelet analysis. The method of distinguishing different disturbances based on the characteristic vector is presented. And this proposed called "scale-variance" can be used to clearly distinguish the different disturbances. In order to improve the system SNR, wavelet threshold denoising filter design.(4) According to the design specifications of the topic and the theoretical approach proposed in this paper, and field experiments. The experimental results verify the effectiveness of the proposed model and method.