Research On Signal Recognition Technology Of Perimetersecurity System

The security protection system is called SPS(Security Protection System)and is the precondition for the stable development of social security.It is an intrusion prevention system composed of security protection devices.Optical fiber perimeter security system is a very important application of fiber optic technology in the non-communication field.The distributed optical fiber perimeter security system is an advanced security system for national security that uses modern optical technology and information processing technology as the core components of optical fiber vibration sensors.It has the advantages of high sensitivity,low energy consumption,etc.Because the external signals are complex and changeable,the difficulty of sensor detection is increased.Therefore,the identification and analysis of the intrusion behavior of vibration signals caused by intrusion behavior will be the fiber perimeter.One of the key technologies of the security system,the effect of the processing will directly affect the system's ability to detect intrusion signals.This article will focus on the perimeter security system of fiber optics,and use intrusion behavior as the object.The target of vibration signal feature extraction and signal recognition is the specific research target.A series of research work is carried out as follows:(1)Combining the development of perimeter security technologies at home and abroad,the current perimeter security system is improved,and a perimeter security system based on optical fiber vibration sensing is proposed.At the same time,the signal is collected using the characteristics of the Internet of Things sensing layer.(2)In order to be able to thoroughly discuss the accuracy of perimeter intrusion detection,based on the characteristics of the fiber vibration signal,based on the traditional denoising method of fiber vibration signal,a new threshold denoising method is proposed.The results show that: The new denoising method has a significant improvement in denoised SNR and RMS error data.(3)Based on the denoising of fiber vibration signals,three signal feature extraction algorithms are proposed.For each kind of signal feature extraction algorithm.(4)We proposed a fuzzy function slice as a feature extraction method of optical fiber vibration signal.At the same time,we use ReliefF feature selection method to optimize the selected slice,and then obtain more sparse feature subsets and simplify the following classification and identification work.(5)Explained the related theories of SVM classification,constructed a multi-class SVM by adopting one-to-one method,combined with the signal feature extraction mentioned in the article,and verified through fiber-optic vibration signal recognition experiments.The results show that: The SVM combined optical fiber vibration signal recognition algorithm can effectively mine the phase modulation information it contains.Through the research of this topic,we will effectively solve the problems such as the high false alarm rate in alarm systems in perimeter security and the inability to accurately determine the location,effectively improving the level of public security defense in China.