Effect of focused ultrasound stimulation at different ultrasonic power levels on the local field potential power spectrum

Local field potential(LFP) signals of the rat hippocampus were recorded under noninvasive focused ultrasound stimulation(FUS) with different ultrasonic powers. The LFP mean absolute power was calculated with the Welch algorithm at the delta, theta, alpha, beta, and gamma frequency bands. The experimental results demonstrate that the LFP mean absolute power at different frequency bands increases as the ultrasound power increases.     

瞬时频率方差检测器（VIFD）及其性能评价

匹配滤波器及相关器是在理想信道中白噪声背景下检测确知信号的最佳检测器,但由于水声信道的复杂性,这类理论上的最佳检测器在实际水声信道的应用中,性能产生严重蜕化。利用瞬时频率序列及其统计特征的估计结果,根据水声信道以及主动声呐工作环境的特点,提出了一种针对CW脉冲的回波信号的性能稳健的新型检测器一瞬时频率方差(VIFD)检测器。根据理论分析、计算机仿真分析和实验验证,该检测器在非平稳的干扰背景和信道中具有很好的稳健性。对信号的起伏与衰落及多卜勒失配也有很好的适配性能。     

次级通道模型误差下滤波X型最小均方差算法收敛性分析

给出了滤波X型LMS(FXLMS)算法收敛的一个充分条件,指出如果次级通道传递函数与其估计值在所有频带上满足正实条件,则FXLMS算法对任意参考信号收敛。若上述正实条件仅在某些频带上满足,则FXLMS算法的收敛将依赖于参考信号功率谱密度的分布。收敛步长取决于某特定相关矩阵特征值的分布。将上述结论应用于时延型LMS(DLMS)算法,得出在时延估计存在误差时,DLMS算法收敛于若干离散的频带,而频带宽度完全取决于“对延估计误差频率”(时延估计误差倒数的1/4)。     

Sinusoidal modeling for nonstationary voiced speech based on a local vector transform

A voiced speech signal can be expressed as a sum of sinusoidal components of which instantaneous frequency and amplitude continuously vary with time. Determining these parameters from the input, the time-varying characteristics are crucial error sources for the algorithms, which assume their stationarity within a local analysis segment. To overcome this problem, a new method is proposed, local vector transform (LVT), which can determine instantaneous frequency and amplitude for nonstationary sinusoids. The method does not assume the local stationarity. The effectiveness of LVT was examined in parameter determination for synthesized and naturally uttered speech signals. The instantaneous frequency for the first harmonic component was determined with an accuracy almost equal to that of the time-corrected instantaneous frequency method and higher accuracy than that of spectral peak-picking, autocorrelation, and cepstrum. The instantaneous amplitude was also determined accurately by LVT while considerable errors were left in the other algorithms. The signal reconstructed from the determined parameters by LVT agreed well with the corresponding component of voiced speech. These results suggest that the method is effective for analyzing time-varying voiced speech signals.     

Wavelet modulation: An alternative modulation with low energy consumption

This paper presents wavelet modulation, based on the discrete wavelet transform, as an alternative modulation with low energy consumption. The transmitted signal has low envelope variations, which induces a good efficiency for the power amplifier. Wavelet modulation is analyzed and compared for different wavelet families with orthogonal frequency division multiplexing (OFDM) in terms of peak-to-average power ratio (PAPR), power spectral density (PSD) properties, and the impact of the power amplifier on the spectral regrowth. The performance in terms of bit error rate and complexity of implementation are also evaluated, and several trade-offs are characterized.     

Effects of flanking band proximity, number, and modulation pattern on comodulation masking release

Comodulation masking release for a 700-Hz pure-tone signal was investigated as a function of the number and spectral positions of 20-Hz-wide comodulated flanking bands. In the first experiment, all stimuli were presented diotically. CMR was examined as a function of the number of flanking bands present, in conditions where the bands were arranged symmetrically around the signal frequency, were below the signal frequency, or were above the signal frequency. The number of flanking bands ranged from one to eight, and the magnitude of the diotic CMR ranged from approximately 5-16 dB. The results indicated: (1) bands closer to the signal resulted in larger masking release, and (2) more bands gave rise to larger CMR (but with diminishing returns above two flanking bands). Two additional sets of diotic conditions were examined and compared to the condition where all eight comodulated flanking bands were present: In one set of conditions, two of the eight flanking bands were removed; in the other set of conditions, two of the eight flanking bands were replaced with bands (termed "deviant" bands) that were not comodulated with respect to the other bands. There was very little effect of reducing eight bands to six, even when the removed bands were relatively near the signal frequency; however, CMR was substantially reduced when deviant bands were introduced, particularly when the deviant bands were placed relatively near the signal frequency. These reductions in CMR were slightly greater when each of the deviant bands had a unique modulation pattern (bideviant bands) than when the two deviant bands themselves shared the same modulation pattern (codeviant bands). In the second experiment, dichotic conditions were examined where the number and spectral positions of the flanking bands in the nonsignal ear were varied (the signal ear received only a 20-Hz-wide noise band centered on the signal frequency). The magnitude of the dichotic CMR ranged from approximately 2-10 dB, depending on condition. Effects of proximity and the number of flanking bands were similar to the effects obtained in diotic conditions. For both the diotic and the dichotic data, the effects of proximity were more consistent with an interpretation based upon across-channel processing than upon a within-channel interaction. The results obtained using deviant bands indicate that it is difficult for the auditory system to disregard the modulation pattern of flanking bands that differ from the modulation pattern of the on-signal band, particularly if such bands are proximal to the signal frequency.     

Microwave electromagnetic pulse from asymmetrical nonstationary gamma source

The coherent microwave radiation emitted in air by a low-intensity asymmetrical nonstationary gamma source is examined. A dependence of the spectral power of the microwave radiation on the frequency and characteristics of the gamma source is obtained.     

Development and comparison of spectral methods for passive acoustic anomaly detection in nuclear power plants

We have developed spectral signal processing methods for passive acoustic anomaly detection in nuclear power plants. Furthermore, we compared the developed and existing methods by applying them to stationary sounds recorded in a controlled environment. Our new methods show significant improvement, in particular concerning robustness against false alarms. The results also demonstrate that clear detection of a given sound at a given signal-to-noise ratio is highly dependent on the distribution of characteristic frequency content in the spectrum in relation to the background noise and the spectral uncertainty. Since the frequency monitoring principle used here is quite rigid, we stress the need for research on more flexible methods, also taking into account differences between experiments and real reactor systems.     

Broadband sound generation by confined turbulent jets

Sound generation by confined stationary jets is of interest to the study of voice and speech production, among other applications. The generation of sound by low Mach number, confined, stationary circular jets was investigated. Experiments were performed using a quiet flow supply, muffler-terminated rigid uniform tubes, and acrylic orifice plates. A spectral decomposition method based on a linear source-filter model was used to decompose radiated nondimensional sound pressure spectra measured for various gas mixtures and mean flow velocities into the product of (1) a source spectral distribution function; (2) a function accounting for near field effects and radiation efficiency; and (3) an acoustic frequency response function. The acoustic frequency response function agreed, as expected, with the transfer function between the radiated acoustic pressure at one fixed location and the strength of an equivalent velocity source located at the orifice. The radiation efficiency function indicated a radiation efficiency of the order (kD)2 over the planar wave frequency range and (kD)4 at higher frequencies, where k is the wavenumber and D is the tube cross sectional dimension. This is consistent with theoretical predictions for the planar wave radiation efficiency of quadrupole sources in uniform rigid anechoic tubes. The effects of the Reynolds number, Re, on the source spectral distribution function were found to be insignificant over the range 20002.5. The influence of a reflective open tube termination on the source function spectral distribution was found to be insignificant, confirming the absence of a feedback mechanism.     

基于Hilbert变换的水下多源声信号频率的相干探测

为实现水下中低频声信号的探测识别,通过研究水下多声源相干探测信号的特征,理论上给出了相干探测信号频谱混叠情况下的特征表达式,并提出了一种基于Hilbert变换的信号解调处理方法,实现了水下多声源相干探测信号频谱混叠情况下各声源发声频率的解调.该方法将探测信号经过滤波平滑处理之后进行Hilbert变换,得到信号的解析形式,然后对解析信号模值的平方进行二次滤波平滑等处理,分离混叠在一起的频带,将得到的信号进行频谱分析,根据频移值计算得到水下各个声源的发声频率.在光学暗室下搭建激光相干探测系统,对2~6kHz的水下声信号进行实验,实验结果表明,该方法可以有效分离探测信号中混叠在一起的信号频带,并准确提取各水下声信号的发声频率,频率提取重复性不大于2.5Hz.     

Action of nonstationary linear systems on the time-dependent spectrum of ultrashort light pulses

The spectral and temporal properties of laser light pulses are adequately characterized by the time-dependent spectrum. Response functions are investigated in the frequency as well as in the temporal domain which are suitable for describing the interaction of these pulses with nonstationary (and stationary) systems. The results are applied to physically important systems: the Fabry-Perot étalon, transmission through glass, absorption of molecules, reflection by semiconductors and the optical gate. The conditions are investigated under which the response functions can be measured by applying a square-law detector. An uncertainty relation is formulated for the time-dependent spectrum.     

Cyclic feature suppression for physical layer security

Cyclic prefix (CP) deploying techniques such as orthogonal frequency division multiplexing (OFDM) and single carrier frequency domain equalization (SC-FDE) offer considerable advantages in terms of equalizing time dispersive effect of wireless channel at the expense of a reasonable spectral redundancy. However, CP introduces cyclic features to the signal which can also be exploited for signal interception, blind parameter estimation and synchronization, and therefore, compromises the security of the signal against eavesdropping attacks. In order to provide a covert communication against such attacks, in this paper, we present two novel techniques that suppress the cyclic features of the CP utilizing signals while maintaining their advantages in equalization without reducing spectral efficiency. The first technique is built on a CP selection strategy while the second one is based on randomizing the symbol time. We also performed peak-to-average power ratio mitigation and out-of-band leakage suppression along with the cyclic feature concealing in the second technique at the expense of a reasonable complexity and signaling. Subsequent to the presentation of the proposed techniques, their performances are discussed and compared for OFDM and SC-FDE in terms of complexity and bit-error-rate along with cyclic feature suppression.¹     

基于时变功率谱的激光器调谐瞬时线宽测量方法研究

半导体激光器的线宽通常采用激光外差测量技术,通过差拍信号的功率谱密度函数来确定,受傅里叶变换方法的限制,得到的均是在一定时间段内的静态平均线宽。为了获得半导体激光器在电流调谐过程中的瞬时线宽特性,提出了利用时变功率谱获知调谐瞬时线宽的相干和非相干测量方法,并分别进行了理论分析和实验验证。首先对半导体激光器输出光信号及差拍信号进行了时间-频率域下的数学描述,确定了时变功率谱与调谐瞬时线宽的关系;其次,针对差拍信号的趋向性特征,提出了趋势局部均值分解方法,并研究了利用分解出的乘积函数建立差拍信号及激光器输出光信号的时变功率谱的方法;最后利用非相干和相干测量法分别获得了分布反馈式半导体激光器在50~51及50~100 mA锯齿波电流调谐过程中的瞬时线宽。     

A Linear Frequency Principle Model to Understand the Absence of Overfitting in Neural Networks

Why heavily parameterized neural networks(NNs) do not overfit the data is an important long standing open question. We propose a phenomenological model of the NN training to explain this non-overfitting puzzle. Our linear frequency principle(LFP) model accounts for a key dynamical feature of NNs: they learn low frequencies first, irrespective of microscopic details. Theory based on our LFP model shows that low frequency dominance of target functions is the key condition for the non-overfitting of NNs and is verified by experiments. Furthermore,through an ideal two-layer NN, we unravel how detailed microscopic NN training dynamics statistically gives rise to an LFP model with quantitative prediction power.     

Proposals for the directive characterisation of electroacoustic emitters

In this paper, the concept of directivity is generalised to acoustic sources radiating transients or signals evolutive with time. (These are the most common cases in the power or reinforcement electroacoustic systems.) The generalisation proposed is based upon the anamorphism relating the signal levels emitted into the free space in different directions. The relationship between the signals observed in two arbitrary directions is essentially independent of time. Therefore, the anamorphical relationship offers the possibility of obtaining the directivity patterns simply by using as a test signal that signal commonly emitted by the system (i.e. speech in a reinforcement system for a conference room). This principle and method can be applied without major restrictions to any other system, or piece or part of machinery emitting acoustic energy in discontinuous form.Concerning electroacoustic sources, it appears advantageous to replace the usual test signal consisting of pure tones by the signal proper to the system (music, speech, etc.) filtered into the standardised frequency bands. The complete signal (not filtered) can also give significant results. As a simplifying and reasonable compromise regarding the directivity for speech and music, bands of white noise are proposed as test signals.     

Continuous wavelet transform and exactly solvable model of nonstationary signals

An analytic expression for a continuous wavelet transform (CWT) is obtained for a nonstationary signal with spectral properties varying with time. As a model of the nonstationary signal, a superposition of elementary nonstationary signals is considered, each of which is the product of the Gaussian-shape envelope and an oscillating function. To obtain the CWT, the mother Morlet wavelet is used. The result is compared with the window Fourier transform (Gabor transform).The advantages of the CWT are illustrated by concrete examples. The applications of the exactly solvable model of nonstationary signals for simulating transient processes in physics are considered.     

差频可调谐太赫兹技术的新进展

太赫兹技术在最近30年来得到快速发展, 并在医学、生物、农业、材料、安检、通信、天文等领域得到广泛应用. 从太赫兹源的频谱特性可以分为窄带(单频)太赫兹源和宽带太赫兹源. 从频谱技术方面来说, 相干的宽带和窄带太赫兹谱是一种互补性关系, 具有各自的技术特点和应用范围. 宽带太赫兹谱可以用于快速获取较宽频谱范围的分子振转谱, 实现混合特征谱的快速检测或成像. 窄带太赫兹源具有很好的光谱灵敏度和分辨率, 适用于太赫兹抽运-探测、分子振转能级谱精细结构分辨 以及太赫兹远程探测和成像. 因此研制具有可调谐的高峰值功率的窄带太赫兹源是适用于探测和识别分子振转能级指纹谱的应用需求, 而差频技术是获得高功率和宽调谐窄带太赫兹源最重要的技术之一. 为了突出该技术的最新进展, 本综述引证论文仅仅限于近5 年来基于差频技术产生太赫兹波的研究进展, 分为光学激光差频源和量子级联激光器差频源两大部分. 对于光学激光差频源, 分别对目前文献报道的各种双波长差频源和太赫兹产生用的非线性晶体进行分类介绍, 并给出所采用的技术和实验结果; 对于量子级联激光器差频源, 分别介绍了量子级联激光器中的差频产生技术和波长调谐技术的最新进展. 量子级联激光器差频太赫兹源是目前实现量子级联激光器在太赫兹波段室温运转的惟一技术, 是实现小型化、窄带宽调谐和室温运转太赫兹源的新发展领域, 值得关注.     

Kinetics of ions in a neutral gas upon abrupt application of an electric field. I. CEM model

A new method for calculating matrix elements of the collision integral is used for solving problems of the mobility of ions against the background of atoms and for constructing the distribution functions for ions upon an abrupt application of an electric field. It is shown how the stationary distribution function can be constructed using the nonstationary moments method in the case when the stationary moments method is completely inapplicable. The solution to the nonstationary problem for the CEM model corresponding to resonant charge exchange with a constant collision frequency, which is constructed analytically, is used for analyzing the limits of applicability of the nonstationary moments method.     

Fermat principle for a nonstationary medium

One possible formulation of a variational principle of the Fermat type for systems with time-dependent parameters is suggested. In a stationary case, it reduces to the Mopertui-Lagrange least-action principle. A class of Hamiltonians (dispersion relations) is indicated, for which the variational principle reduces to the Fermat principle in a general nonstationary case. Hamiltonians that are homogeneous functions of momenta are in this category. For the important case of nondispersive waves (corresponding to Hamiltonians being homogeneous function of momenta order 1) the Fermat principle fully determines the geometry of the rays. Equations relating the variation of signal frequency with the rate of change of propagation time are established.     

基于毫米级单晶石墨烯的倍频器性能研究

石墨烯作为一种拥有高电子迁移率和高饱和速度的二维材料,在射频电子学领域具有很大的应用潜力,引起了人们广泛的研究兴趣.近些年随着化学气相沉积制备石墨烯技术的发展,高质量大尺寸的单晶石墨烯生长技术也愈加成熟.本文基于化学气相沉积生长的毫米级单晶石墨烯,在高介电常数介质上制备出高性能的石墨烯倍频器,并且对其倍频特性做了系统的研究.研究结果表明:在输入信号频率为1 GHz时,倍频增益可以达到-23.4 dB,频谱纯度可以达到94%.研究了不同漏极偏压以及输入信号功率下倍频增益的变化特性,随着漏极偏压以及输入信号功率的增加,倍频增益增加.对具有不同跨导和电子空穴电导对称性的器件的倍频增益和频谱纯度随输入信号频率f_(in)的变化关系进行了研究.结果表明,跨导对于倍频增益影响显著,在f_(in)=1 GHz时器件的频谱纯度差别不大,均大于90%,但是随着f_(in)增加至4 GHz,电子空穴电导对称性较差的器件频谱纯度下降至42%,电子空穴电导对称性较好的器件仍能保持85%的频谱纯度.这是电子空穴电导对称性和电子空穴响应速度共同作用的结果.本文的研究结果对于高性能石墨烯倍频器设计具有一定的指导意义.