【摘要】：以紫外光作為通信載體的紫外光語(yǔ)音通信系統(tǒng),以其對(duì)噪聲極強(qiáng)抗干擾能力、保密性能好、非視距通信(Non Line of Sight,NLOS)和全天候通信等優(yōu)點(diǎn)逐漸成為軍事通信領(lǐng)域的研究焦點(diǎn)。但是,光通信領(lǐng)域又面臨著信道資源有限和系統(tǒng)誤碼率高的問題,如何解決光通信中的問題,研究者主要從改進(jìn)調(diào)制方式和進(jìn)行信源編碼兩個(gè)方向進(jìn)行攻關(guān),其中以信源編碼方式較為有效,研究高語(yǔ)音質(zhì)量和低編碼速率的編碼方式成為光通信領(lǐng)域信源編碼的重點(diǎn)。論文研究的目的是在參數(shù)編碼的理論基礎(chǔ)上,對(duì)特征參數(shù)進(jìn)行提取,并在現(xiàn)場(chǎng)可編程邏輯陣列(FieldProgrammable Gate Array,FPGA)上實(shí)現(xiàn),從而達(dá)到低速率語(yǔ)音編碼的目的,編碼后的數(shù)據(jù)加載到輸出光波長(zhǎng)為266 nm的紫外激光器上并調(diào)制輸出,紫外光在大氣中傳播由光電倍增管接收并轉(zhuǎn)換輸出電流信號(hào),電流信號(hào)被信號(hào)處理電路進(jìn)行電流電壓轉(zhuǎn)換和放大處理,在FPGA開發(fā)板上對(duì)放大后的電壓信號(hào)進(jìn)行處理,完成解碼工作,最終還原語(yǔ)音信號(hào),從而實(shí)現(xiàn)高質(zhì)量低速率紫外光語(yǔ)音通信。首先闡述了大氣吸收效應(yīng)和散射效應(yīng),進(jìn)一步以瑞利散射和米氏散射為理論基礎(chǔ),建立了紫外光通信的單散射橢球模型。另外提出了參數(shù)編碼中LPC-10編碼算法,該算法能夠在8 kHz的音頻采集頻率下,以180個(gè)樣點(diǎn)為一幀,從而實(shí)現(xiàn)語(yǔ)音的編碼速率為2.4 kb/s。在對(duì)語(yǔ)音進(jìn)行線性預(yù)測(cè)分析時(shí),重點(diǎn)研究了提取濾波器參數(shù)的理論推導(dǎo),為參數(shù)編碼奠定了理論模型。在對(duì)參數(shù)編碼研究的同時(shí),本文也給出了波形編碼中較為典型的A律語(yǔ)音壓縮編碼的基本原理方法。針對(duì)系統(tǒng)的發(fā)射光源和光電探測(cè)器工作原理進(jìn)行了簡(jiǎn)要介紹,同時(shí)對(duì)不同類型的光源和光電探測(cè)器進(jìn)行了對(duì)比分析,最終確定發(fā)射光源選擇紫外激光,光電探測(cè)器選擇光電倍增管(Photomultiplier Tube,PMT)。使用FPGA開發(fā)板對(duì)數(shù)據(jù)流進(jìn)行處理,根據(jù)FPGA的設(shè)計(jì)流程和編程原則,設(shè)計(jì)了串并轉(zhuǎn)換、并串轉(zhuǎn)換、RAM調(diào)用、參數(shù)計(jì)算、編碼輸出等底層模塊,并對(duì)每個(gè)底層模塊的功能進(jìn)行相應(yīng)的時(shí)序仿真和驗(yàn)證。搭建了兩個(gè)通信系統(tǒng),一個(gè)是驗(yàn)證A律編碼的紅光激光通信,另一個(gè)就是以LPC-10參數(shù)編碼為核心的紫外光語(yǔ)音通信。并對(duì)以上兩個(gè)語(yǔ)音通信系統(tǒng)在實(shí)驗(yàn)室進(jìn)行了多次測(cè)試。測(cè)試結(jié)果表明,設(shè)計(jì)的A律編碼的紅光激光通信的語(yǔ)音編碼速率為64 kb/s,客觀平均意見得分(Mean Opinion Score,MOS)≥3.0,設(shè)計(jì)的LPC-10編碼的紫外光語(yǔ)音通信的語(yǔ)音編碼速率為2.4 kb/s,實(shí)現(xiàn)了中低速率語(yǔ)音編碼通信。
[Abstract]:Ultraviolet (UV) voice communication system, with its strong anti-jamming ability to noise, good security performance, non-line-of-sight communication (NLOS) and all-weather communication, has gradually become the focus of research in the field of military communications. However, the field of optical communication is faced with the problems of limited channel resources and high bit error rate (BER). Among them, the source coding is more effective, and the research of high speech quality and low coding rate has become the focus of source coding in the field of optical communication. The purpose of this paper is to extract feature parameters based on the theory of parameter coding and implement them on Field Programmable Gate Array (FPGA), so as to achieve the purpose of low rate speech coding. The encoded data is loaded into the output wavelength of 266 nm UV laser and modulated. The ultraviolet light propagates in the atmosphere and is received by the photomultiplier tube and converted into the output current signal. The current signal is converted and amplified by the signal processing circuit, the amplified voltage signal is processed on the FPGA development board, the decoding work is completed, and the speech signal is finally restored. In order to achieve high quality low-rate UV voice communication. Firstly, the atmospheric absorption effect and scattering effect are described. Based on Rayleigh scattering and Michlet scattering theory, a single scattering ellipsoid model for ultraviolet communication is established. In addition, the LPC-10 coding algorithm in parameter coding is proposed. The algorithm can take 180 samples as a frame at 8 kHz audio acquisition frequency, thus the coding rate of speech is 2.4 kb / s. In the linear prediction analysis of speech, the theoretical derivation of extracting filter parameters is studied, which establishes a theoretical model for parameter coding. In addition to the research on parameter coding, this paper also gives the basic principle and method of A law speech compression coding, which is typical in waveform coding. This paper briefly introduces the working principle of the system's emitting light source and photodetector. At the same time, the different light sources and photodetectors are compared and analyzed. Finally, it is determined that the emitting light source should choose the ultraviolet laser. The photomultiplier PMT is selected by the photodetector. Using FPGA development board to process the data flow, according to the design flow and programming principle of FPGA, the bottom modules, such as serial-parallel conversion, parallel conversion RAM call, parameter calculation, encoding output and so on, are designed. And the function of each underlying module is simulated and verified. Two communication systems are built, one is the red laser communication which verifies A-law coding, the other is the ultraviolet voice communication based on LPC-10 parameter coding. The above two voice communication systems have been tested in the laboratory many times. The test results show that The speech coding rate of A law coded red light laser communication is 64 kb / s, the mean opinion score (MOS) is 鈮，

本文編號(hào)：2138857