采用经验模态分解的语音与音频通用编码方法

为有效解决现有单一模型编码器无法在中低速率对语音和音频信号进行高质量通用编码的问题,本文借助语音与音频信号的谐波特性,建立了一种对语音和音频信号统一编码的方法。首先,本文利用经验模态分解(Empirical Mode Decomposition, EMD)提取输入信号的谐波成分;其次,利用感知匹配追踪算法,并结合正弦参数建模对谐波成分进行参数提取与量化;第三,对于量化谐波后的残差进行抖动格型矢量量化,以提升重建音频的主观听觉质量,并最终实现一套包含24kbps和32kbps码率的宽带语音与音频通用编码器;最后,对所提算法进行了客观PESQ/PEAQ和主观A/B测试,并与ITU-T G.722.1和G.722.2编码器进行了比较,实验结果表明,所提编码器对语音和音频信号的编码质量均优于参考编码器。      

Harmonic gain and noise in a frequency-doubling gyro-amplifier

Nonlinear gain in a 34-GHz three-stage frequency-doubling gyro-traveling wave tube (gyro-TWT) has been experimentally studied. The device consists of a thermionic electron gun, TE/sub 01//spl rarr/TE/sub 02/ fundamental gyro-TWT input section, second harmonic TE/sub 03/ intermediate buncher section, and a second harmonic TE/sub 02//spl rarr/TE/sub 04/ complex output circuit. Nonlinear bunching in the electron orbital phase generates harmonics of the input signal in the beam current, which excite the subsequent circuits at the second harmonic frequency. Since the gain is nonlinear, noise or applied sideband signals intermodulate with the carrier generating high-order products in the output. Therefore, it has been suggested that the noise figure of these devices may be unreasonably high. In this study, the complex harmonic transfer characteristics were experimentally measured and compared with calculations based on the assumption that the gyro-amplifier gain can be described, in the narrowband sense, as a classical frequency-doubling circuit. The results show that narrowband intermodulation gain is 6 dB higher than the carrier as predicted in the small signal limit, but as the device reaches saturation the nonlinear products become suppressed with respect to the carrier. Tests on the broadband gain characteristics show that output noise consists of second harmonic shot noise spontaneously excited in the output circuits along with the products of the intermodulation between external noise and the carrier. Good agreement between the experimental results and the calculations is demonstrated.     

Harmonic Millimeter-Wave Generation and Frequency Up-Conversion Using a Passively Mode-Locked Multisection DFB Laser Under External Optical Injection

We demonstrate that a passively mode-locked multisection distributed-feedback (PML-MS DFB) laser can perform both harmonic millimeter-wave generation and harmonic frequency up-conversion under the external optical injection. Optical multiple modes generated by a PML-MS DFB laser produce harmonic millimeter-waves by mode-beating in a photodiode (PD), and the phase quality and stability are enhanced by the injection of external optical signal modulated at f_LO. In addition, if the external optical signal is modulated simultaneously by f_LO and f_IF, the PML-MS DFB laser performs all-optical frequency up-conversion producing sidebands at f_LOplusmnf_IF when detected by a PD. Using this method, we demonstrate generation of stable harmonic millimeter-waves at 30.42 and 60.84 GHz with f_LO of 15.21 GHz, and up-conversion of 10-Mb/s quadrature-phase-shift keying data at 150 MHz f_IF into the 60-GHz band. These functions of a PML-MS DFB laser can be useful for radio-over-fiber applications in which a compact optical source is needed for processing high-frequency radio signals in optical domain     

Linearisation of RF PA by even-order nonlinear baseband signal processed in digital domain

This paper represents, for the first time, the linearisation method of RF power amplifiers, which is based on using the second- and fourth-order nonlinear signals shaped by digital signal processing in a baseband domain. The linearisation signals are generated by the second- and fourth-order nonlinear modifications of the baseband signal in the proposed manners. The composite second- and fourth-order linearisation signals modulate carrier second harmonics, and the modulated signals are then driven to the gate and drain of the amplifier transistor. The linearisation effects of the proposed method are evaluated on a single stage power amplifier for the simulated QAM and OFDM digitally modulated signals at diverse input power levels up to the 1 dB compression point. Moreover, the FPGA implementation of a system for the generation and processing of the fundamental signal, and the second-order linearisation signals is presented in the paper, and the proposed technique is verified for the generated QAM signal. In addition, the linearisation of the two-way asymmetrical Doherty amplifier is performed in the experiment by using software defined radio platforms for synthesising the fundamental signal and the second-order linearisation signals.     

谐波锁模光电振荡器（特邀）

提出了一种主动锁模光电振荡器（OEO）方案,可以实现高阶谐波锁模,从而产生具有高重复频率的微波脉冲信号。在所提方案中,通过在OEO腔内的电光强度调制器直流偏置端口引入一个正弦驱动信号,当该正弦信号的频率为OEO环腔自由光谱范围的整数倍时,实现基频（）或谐波（）锁模,输出重复频率为的微波脉冲信号。实验中分别实现了10阶、50阶和100阶谐波锁模,输出微波脉冲信号的重复频率分别为360 kHz、1.8 MHz和3.6 MHz。该方案为脉冲多普勒雷达等系统应用提供了一种全新的、具备低相噪潜力的微波脉冲信号产生的技术途径。     

基于1 bit时间调制技术的功率控制方法

针对传统数控衰减器结构复杂、控制信号路数多的问题,提出了一种基于1 bit时间调制技术的功率控制方法. 通过对1 bit移相器0°和180°两种状态占空比的调节,该方法可以有效实现对输入信号基波功率的动态控制,利用带通滤波器抑制多余谐波后其可具备数控衰减器的功能. 还进一步讨论了调制信号时间精度对不同衰减量和不同调制频率的约束. 仿真和实验结果验证了所提功率控制方法的有效性. 相比于传统数字控制衰减器级联网络和多路控制信号的结构,本文方法只需一个调制模块和两路控制信号即可实现多种衰减状态可调的功能,因此具备结构简单的优势.     

常规调制信号与扩频信号的调制识别算法

由于直扩信号（Direct Sequence Spread Spectrum, DSSS）通常淹没在噪声中,为了有效地识别直扩信号、跳频信号（Frequency Hopping Spread Spectrum, FHSS）和常规调制信号,提出基于瞬时特征和高阶累积量的识别算法。首先推导证明了FM、MFSK、MPSK、DSSS、FHSS信号的归一化四阶累积量切片 是相同的,并推导得到AM和MQAM的 通用公式。然后,利用归一化四阶累积量切片 特征参数检测噪声中通信信号,利用占用带宽特征参数将信号分为扩频信号和常规调制通信信号两类;最后利用瞬时特征参数和高阶累积量特征参数,分别识别扩频信号和5种常规调制通信信号。仿真结果表明,当信噪比高于1dB时,该算法对上述7种信号的正确识别率可达到100%。      

Reflective SOA-Based Bidirectional WDM-PON Sharing Optical Source for Up/Downlink Data and Broadcasting Transmission

We proposed and experimentally demonstrated the reflective semiconductor optical amplifier (RSOA)-based wavelength-division-multiplexed passive optical network (WDM-PON) scheme where not only up/downlink data services but also broadcasting service could be provided using a single optical source. In the proposed scheme, the digital signal and subcarrier multiplexing (SCM) signal for downstream were simultaneously modulated by a single distributed feedback laser diode and RSOA in optical network unit remodulated downstream source as an upstream. Without performance deterioration of digital signals both in up- and downstream, the proposed WDM-PON scheme can stably offer the SCM signal for broadcasting service. In the experiment, 1-Gb/s digital signals both for up- and downstream and 20-Mb/s SCM signal at 2.2 GHz for broadcasting were demonstrated in 10-km bidirectional link     

Numerical method for determining the quantization error PDF of single-bit /spl Sigma//spl Delta/ Modulators

Existing models for the quantizer of /spl Sigma//spl Delta/ modulators make assumptions on the probability density function (pdf) of the quantization error, or some other convenient signal of the modulator. In this paper, a method for the determination of this pdf for single-bit /spl Sigma//spl Delta/ modulators is presented. First, a numerical method is proposed in order to solve the simplified equation for the quantization error pdf for first-order systems considering noiseless and noisy dc input signals. Then, it is shown how most practical high-order (>2)/spl Sigma//spl Delta/ modulators, resulting from well-established design methods, can be modeled as first-order systems plus an additive noise source at the input. Hence, their quantization error pdf is analyzed using the proposed method. Simulation results are shown to be in considerable agreement with those of the proposed method.     

基于Turbo均衡和信道估计的单通道盲信号恢复算法

定时同步是单通道盲信号接收端处理的难点,提出了一种无需定时同步基于Turbo均衡的单通道盲信号恢复算法。该算法将定时同步偏差等效为符号间干扰（ISI,inter-symbol interference）信道,通过信道估计和Turbo均衡相互反馈软信息来改善源信号信息恢复性能。重点研究了初始盲均衡算法、信道估计算法、混合信号的MMSE均衡算法以及三者间的软信息交互。算法复杂度低、计算量小,适用于高阶调制信号。仿真结果表明,对BPSK、QPSK和8PSK信号,该算法都能得到较好的性能,且对等功率和不等功率信号同样适用。     

A Discrete‐Amplitude Pulse Width Modulation for a High‐Efficiency Linear Power Amplifier

A new discrete‐amplitude pulse width modulation (DAPWM) scheme for a high‐efficiency linear power amplifier is proposed. A radio frequency (RF) input signal is divided into an envelope and a phase modulated carrier. The low‐frequency envelope is modulated so that it can be represented by a pulse whose area is proportional to its amplitude. The modulated pulse has at least two different pulse amplitude levels in order that the duty ratios of the pulse are kept large for small input. Then, an RF pulse train is generated by mixing the modulated envelope with the phase modulated carrier. The RF pulse train is amplified by a switching‐mode power amplifier, and the original RF input signal is restored by a band pass filter. Because duty ratios of the RF pulse train are kept large in spite of a small input envelope, the DAPWM technique can reduce loss from harmonic components. Furthermore, it reduces filtering efforts required to suppress harmonic components. Simulations show that the overall efficiency of the pulsed power amplifier with DAPWM is about 60.3% for a mobile WiMax signal. This is approximately a 73% increase compared to a pulsed power amplifier with PWM.     

基于判别式受限玻尔兹曼机的数字调制识别

为了提高大动态信噪比下数字调制识别性能,提出一种基于高阶累积量和判别式受限玻尔兹曼机的联合调制识别方法。该方法提取数字信号的高阶累积量作为信号特征,综合利用判别式受限玻尔兹曼机的生成能力和分类能力,分析了含有高斯噪声、时变相位偏移或瑞利衰落环境下的数字信号识别率。实验结果表明,与传统识别方法相比,所提方法的识别性能有明显改善。此外,利用该模型的生成能力对输入特征进行重构,可有效提高低信噪比下的信号识别率。     

Band-pass nonlinearities

This paper presents a simple method for analyzing the effect of a nonlinearity sandwiched between band-pass filters, applying it to the case of weak signals when both filters pass the same band of frequencies, and obtaining completely general expressions for the output signal and noise when the output filter passes a harmonic of the input band.     

负SNR数字调相信号的干涉仪测向技术

传统意义上的干涉仪测向数学模型与处理流程一般要求目标信号具有正信噪比(SNR),在对负SNR调制信号实施测向时通常会产生较大误差甚至失效。在分析传统方法失效原因的基础上,利用数字调相信号的特点,通过非线性变换对负SNR信号进行高次载波恢复,然后针对恢复后的高次载波进行相位差提取,从而获得被测信号的来波方向。在此基础上对该方法所能达到的精度进行了理论分析,并将其推广至凡是具有可恢复载波分量的测向应用情况。仿真结果验证了该方法的有效性,这对于干涉仪在电子侦察和电磁频谱监测中更加广泛的应用提供了新的参考。     

Naturally sampled triangle carrier PWM bandwidth limit and output spectrum

This paper first presents a general theory predicting harmonic components in naturally sampled pulse-width modulation (PWM) output signals, so that any input signal with a Fourier series representation may be handled. This theory provides a basis for the mathematical analysis of PWM systems, such as converters transmitting a main signal and a broadband of feedback or feedforward signals, or Class D amplifiers. Ultimately, the results of this theory are applied to two concrete problems, and conditions similar to the Nyquist theorem or Carson's rule for FM modulation are derived for the cases of recovery of an input signal consisting of a main signal and a bandwidth of small feedback or feedforward signals and a bandlimited signal consisting of several harmonics of comparable magnitude. A rule of thumb, and conservative estimate of the bandwidth in both cases is one third of the carrier frequency (/spl omega//sub upper/<(1/3)/spl omega//sub c/). This research has been based on the premise that the results obtained will provide valuable insight into the general behavior of PWM systems, and provide a supporting theory for concrete systems utilizing signals of this kind, such as PWM converters whose purpose is to ensure a dominant sinusoidal waveform along with a broadband frequency channel for small feedback signals.     

V-BLAST OFDM信号调制识别算法

提出了基于高阶累积量的V-BLAST OFDM信号识别算法,用以区分频率选择性衰落信道条件下V-BLAST结构的多载波信号(0FDM)和单载波信号(MFSK、MPSK、MQAM).该算法不需要预先知道信号的载波频率、波特率以及信道的先验信息,直接在中频对信号进行识别处理.仿真结果表明该算法具有良好的识别性能.     

卫星通信中数字调相信号调制方式识别方法研究

通信信号的调制样式识别在非协作通信中具有重要的研究意义。针对卫星通信中调制方式不断向高阶发展的情况,提出了一种针对常用数字调相信号调制方式自动识别的算法。本文首先给出了卫星通信中数字调相信号的函数模型,介绍了不同调制样式的特点,给出了不同调相信号的星座图。在调制样式识别的研究中,六阶及更高阶的累积量很少用到,本文利用调制信号的八阶累积量和四阶累积量特征的关系,提取出信号的特征呢高参数,对QPSK、8PSK和{16APSK、32APSK}信号进行了区分,证明四阶以上累积量也可以为信号特征提供有用信息。然后通过对高阶APSK信号进行统计分析,采用一种快速收敛且有效的遗传算法,进行实数编码、交叉等操作,计算出采样信号信号平方幅度比,提取出APSK信号的统计特征参数,从而对{16APSK、32APSK}信号进行区分。这种算法对信号的相位偏差具有不变性,同时可抑制加性高斯噪声,特征参数具有很强的鲁棒性;与其它识别算法比较,它具有稳健、应用广泛、实时性强的特性。通过计算机仿真表明,在给定的数据长度和中等信噪比条件下,可得到很高的识别率(>96%)。      

光子晶体中二阶非线性过程产生三次谐波的研究

运用非线性耦合波理论,分析了在非线性光子晶体中,通过二阶过程产生三次谐波的准相位匹配条件。数值模拟结果表明,在入射光中引入倍频光信号,能够极大地改变三次谐波的转换效率。当入射光中倍频光与基频光强度比以及两者复振幅的初相角满足一定的条件时,可在宽的入射光强范围内获得很高的三次谐波转换效率。     

Automatic chromatic dispersion compensation using alternating chirp signal for installation of high-speed transmission systems

Proposes a simple automatic chromatic dispersion compensation system that can detect the zero-dispersion point of an installed fiber span. This technique realizes high sensitivity without requiring the system to be extensively modified. The alternating chirp signal, which is used as the zero-dispersion-detection signal, has alternating optical frequency shift. It can be generated simply by changing the input signals of multiplexer circuits and shifting the bias voltage of the LiNbO/sub 3/ modulator. This technique is shown to detect the zero-dispersion point in the range of -600 to +600 ps/nm by using the signal modulated at 10 GHz; automatic dispersion compensation is successfully demonstrated in a 20-Gb/s, 400-km transmission experiment in our laboratory. Compensation is also demonstrated in 281-km-installed dispersion-shifted fiber using a 20-GHz alternating chirp signal.