Design And Experimental Study Of Performance Improvement Scheme For Ultraviolet Voice Communication System

With the rapid development of human communication needs,the spectrum resource is increasingly scarce.Wireless optical communication has received widespread attention due to the advantages of high modulation rate,broad bandwidth,and no spectral license required.However,large background noise and difficulty in non-line-of-sight transmission limit the application potential of wireless optical communication.The ultraviolet communication(UVC)system working in the "solar-blind" spectral band(200-280nm)has the advantages of low background noise,anti-interference,and the ability of non-line-of-sight communication,which is an important direction for the future development of wireless optical communication technology.However,there are many challenges in the research of UVC,such as the research on UVC channel model is imperfect,the simulation research on UVC system is insufficient,the design and experimental research of full-duplex UVC system,misjudgment easily caused by traditional judgment methods in communication so it needs to improve the performance of UVC by new judgment technology.This paper studies the above problems,and the main work and achievements are as follows:1.We analyze the interaction between atmospheric particles and UV light is analyzed firstly.Combined with the existing ultraviolet transmission model at atmospheric channel,the complete system of UVC is modeled by optical simulation software Opti-system.The effects of different parameters on the performance of UV communication system are simulated and analyzed.2.A full-duplex ultraviolet voice communication system is designed and implemented.The key points of system design are described from the selection of key components,the design of key hardware circuit and software architecture of microcontroller unit(MCU).The realized UV voice communication system uses MCU to realize the voice signal processing,UV LED as the emission light source,UV photomultiplier as the detector,with FPGA core board to complete the receiving and receiving processing of the voice signal.The system can work in full-duplex mode and realize point-to-point line-of-sight communication with a distance of 180 meters at the rate of 115.2kbps,which provides a system experiment platform for further research and testing of UVC system.3.In order to improve the integrity of UVC system,a decision algorithm for UVC system based on Support Vector Machines(SVM)is designed to solve the problem that traditional decision technology is prone to misjudgment in long-distance communication.The proposed algorithm is verified by the realized UVC system.The off-line experimental results show that the SVM-based decision technology can recover the received data more accurately than the traditional decision technology,can meet the communication needs of longer transmission distance,and can effectively improve the overall performance of the UV communication system.