Partially blind source separation of continuous chaotic signals from linear mixture

The authors consider the problem of partially blind source separation of continuous chaotic signals from a linear mixture, in which the generating systems of chaotic signals are assumed to be available and the mixture matrix is unknown. Determination of the mixture matrix is firstly formulated as a problem of the synchronisationbased parameter estimation. Then an efficient parameter estimation method by exploiting the generating dynamics of chaotic signals is introduced. The estimated mixture matrix is used to design a signal separator to blindly separate the mixed chaotic signals. Three examples are given to illustrate the applicability of the proposed approach for the mixed chaotic signals generated by different and/or same dynamical systems and contaminated in measurement noise. In comparison with statistics-based methods, the new approach can solve the magnitude scaling indeterminacy and shows the separability of the mixed signals in strong noise background.     

Blind extraction of chaotic signal from an instantaneous linear mixture

In this paper, we address the problem of blind extraction of a chaotic signal from an instantaneous linear mixture. Determination of the desired extracting vector is formulated as an optimization problem, in which the knowledge of the dynamics of the chaotic signal to be extracted is exploited. A noisy chaotic signal is first extracted and a noise reduction technique is then used to refine the extracted signal. Numerical simulation results demonstrate that the proposed approach can extract a specific chaotic signal from a linear mixture consisting of some chaotic signals and/or random signals.     

Wideband Blind Source Separation Algorithm Based on Beamforming

As the problem of array mixing model of wideband signals cannot be solved by conventional blind source separation algorithms, an improved algorithm based on beamforming is proposed in this paper. First, the received signals are transformed into time–frequency domain, and the delays of source signals are estimated. Then, the received signals are compensated with the estimated delay in frequency domain. Finally, the desired signal is acquired by using Frost wideband beamforming algorithm. Due to adopting the new methods of single source point extraction and delay estimation, the complexity of the proposed algorithm is reduced. Pre-steering delay is used in frequency domain to eliminate the compensation error when the delay is not an integer multiple of the sampling interval, which improves the separation performance significantly. The simulation results show that the proposed algorithm can adequately solve the problem of delay mismatch and achieve wideband blind source separation effectively. The existing algorithms are mostly fail for frequency hopping signals when there are numerous overlapping time–frequency points. In this case, the proposed algorithm still has good separation performance.

     

基于粒子滤波的混沌信号盲分离

该文首先建立了混沌信号盲分离问题的状态空间模型,进而通过引入Rao-Blackwellisaion策略和加扰动的方法,提出了一种基于粒子滤波的在线盲分离算法。在此基础上,为了降低分离后信号中的残留噪声,提出了新颖的延迟估计方法,很好地解决了传统的延迟加权(delay-weight)方法中信号源之间的相互干扰问题。     

基于增殖系数的混沌信号提取算法

本文提出一种基于混沌信号特性的信号盲提取算法,由于不同的混沌信号在相空间里面对应着不同的吸引子二阶增长率,利用这个特点定义了增殖系数（Proliferation Exponent,PE）并将其作为混沌信号提取的目标函数.首先分析基于增殖系数的梯度搜索方法在解决盲提取问题时存在不足,并将混沌信号的盲提取问题转化为带约束的优化问题,提出利用改进的粒子群优化算法解决信号盲提取的优化问题,通过惯性系数动态调整和最优位置的扰动,提高算法的寻优性能.实验结果表明基于增殖系数的信号提取算法能有效地提取混沌信号,提取的信号在时域和相空间与源信号接近,同时算法也表现出对噪声污染的鲁棒性.     

一种引入自适应动量项的变步长混沌信号盲分离算法

该文针对混沌信号盲分离问题,提出一种改进盲分离算法。该算法利用信号分离评价指标来构造函数实现步长和动量因子的自适应调整,然后将构造函数代入盲分离算法中并引入自适应动量项。区别于大多数算法不对混合矩阵进行估计的问题,该算法用变步长函数迭代估计出混合矩阵,从而得到全局矩阵和估计评价指标,以此迭代更新步长和动量因子,最终得出分离矩阵。仿真表明,该算法依据估计评价指标构造函数调整步长和动量因子方法是有效的,在平稳和非平稳环境下对混合混沌信号分离时都能达到收敛速度快且稳态误差小的效果;在混入色噪声时,比传统算法抗噪性能好,表明该文算法在混沌信号盲分离处理中有一定应用价值。     

基于偏度和时序结构的声音信号盲抽取算法研究

信号盲抽取是盲信号处理领域的热点研究方向,它仅抽取感兴趣的信号,能有效减小运算量,解决盲分离中信号顺序不确定性的难题,因而在生物医学信号分析（如EEG、MEG、fMRI等）、语音和图像识别领域得到广泛应用。针对传统的基于时序结构的盲抽取算法存在较弱的抗噪性和对时延估计误差比较敏感的不足,论文提出了将偏度和时序结构相结合的信号盲抽取算法。该算法首先利用偏度的非对称性来度量分离信号的非高斯性,以减弱噪声,同时减小了传统的利用峭度度量非高斯性方法的运算量;其次利用基音周期作为声音信号的最佳时延估计,以实现对感兴趣信号的盲抽取,将两者结合后使得算法对时延估计误差不敏感,且对噪声更具鲁棒性。仿真实验部分选取了标准TIMIT语料库中一男、两女分别单独朗读同一语句的语音信号,盲抽取的实验结果表明:本文算法与文献3中算法相比具有较好的分离效果且抽取速度快,与文献4中算法相比分离效果相当但大大地提高了抽取速度,从而验证了本文算法的有效性。      

基于独立分量分析的PCMA信号盲分离算法

针对非协作通信中成对载波多址(Paired Carrier Multiple Acess,PCMA)信号的盲分离问题,提出了一种基于独立分量分析(Independent component analysis,ICA)的单通道盲分离算法。首先对接收到的单路PCMA信号进行参数估计得到其残余载波频率,再对其处理得到两路基带混合信号,最后利用ICA算法分离出源基带信号。该算法在未知两个卫星地面站发送信号的情况下,从接收到的PCMA信号中恢复出两路源基带信号。仿真实验表明,本文算法在信噪比为-10dB时仍具有良好的分离效果,两路基带信号的波形相似系数可分别达到0.94与0.86以上。     

具有延迟的欠定语音盲分离

探讨具有延迟的欠定语音盲分离问题,针对语音信号的部分W-分离正交性,提出一种新的基于混叠矩阵估计的语音盲分离方法.首先对观测信号进行短时傅里叶变换,选定一个信号作为参考信号,构造幅度衰减向量和时间延迟向量,提取单源主导区间的观测信号作为观测样本;然后对观测样本进行聚类,通过估计参考信号的混叠系数来估计混叠矩阵;最后结合...     

动态变化混叠模型下盲源分离中的源数估计

在进行欠定盲分离时,特别是对于源信号数目及混合矩阵动态变化的情况,常规的欠定盲分离及源数估计方法不能对源信号数目的变化时刻做出判断,因此很难实现动态变化的源信号数目实时和准确的估计。针对这个问题,提出了一种动态变化混叠模型下欠定盲源分离中的源数估计方法。首先,建立动态变化混叠情形下盲源分离的数学模型及动态标识矩阵。其次,基于构建的动态标识矩阵统计和判断动态源信号数目的变化情况。最后,通过分段时间内多维观测矢量采样点聚类区间局部峰值统计,实现动态变化混叠模型下盲源分离中的源信号数目的有效估计。仿真结果表明,该方法能有效实现动态变化混叠模型下欠定盲源分离中的源数估计,并且信号估计效果良好。     

一种有效的语音信号盲分离方法

频域盲语音信号分离存在着排序模糊问题,提出一种基于相邻频点幅度相关和DOA估计相结合的解排序模糊方法,并且通过对一系列预处理（白化）、独立分量分析和后处理算法的优化和有机组合,很好地实现了卷积混合语音信号的盲分离。用真实录制的语音信号进行了仿真实验,恢复出来的源信号的信干比较分离之前提高了约13dB,证明了算法的有效性。     

基于遗传算法的有序盲信号提取

本文针对盲信号分离中,如何根据信号特征进行有序提取的问题进行了探讨,提出了一种基于遗传算法的有序盲信号提取算法.该方法能够确保源信号按照四阶累计量的绝对值降序提取,解决了目前一些基于梯度的提取算法容易陷入局部极值而不能保证有序提取的问题;另外,在信号提取的消源过程中,我们还提出了一种基于Schmidt正交化的消源去相关算法,该方法不仅简化了Cichocki-Thawonmas-Amari(1997)消源算法的复杂计算,同时还对消源后的混叠信号进行了白化.仿真结果表明,该算法能够保证实现盲信号的有序提取.     

Kalman滤波下的多信号单通道盲源分离

窄带物联网环境中,接收机收到的信号通常为多路混合信号,对单通道接收来说,利用常规盲源分离方法很难实现混合信号的分离和源信号提取。针对这一问题,本文提出了一种利用Kalman滤波算法进行信号估计,解决单通道盲源分离的方法。该方法利用信号间的时序结构,通过Kalman滤波算法对多信号混合中的源信号不断估计并迭代更新,最终得到分离信号。仿真实验结果表明,该方法能有效估计并分离出源信号。      

基于波束域预处理的阵列盲信号分离方法

相控阵雷达采用多个阵元进行信号的接收处理,其多通道信号处理可以采用盲信号处理的方法进行目标源信号分离.由于阵元间距导致信号在阵元间产生相位延迟,在进行盲分离的时候一般只能采用卷积混合模型,盲分离过程是较为复杂的多通道反卷积问题.文中对阵列接收信号进行波束域预处理,通过确定的空间波束相位补偿,将阵元域的多时延混合信号变换为瞬时混合信号,从而采用简单的实数分离算法即可完成信号分离,分离信号可用于后置处理.所提方法简单有效,相比常规阵列信号处理方法,可直接适用于宽带信号.仿真实验验证了其有效性.     

基于DSP的音频信号采集和盲分离系统设计

针对语音信号的盲提取问题,以提出的二阶特征窗盲分离算法和模式识别算法为基础,设计实现了以TMS320VC5509DSP为核心的音频信号采集和盲分离系统。详细介绍了系统结构及软硬件实现,并对3路音频信号的DMA同步采集进行了重点分析。该系统结构简单、扩展方便、实时性好,实验结果表明对瞬时混合信号有良好的分离效果。     

基于盲源分离的小波域多重音频水印方法

该文利用盲源分离理论,提出一种小波域的多重音频水印方法。为了解决多水印嵌入过程中经常需要考虑的嵌入顺序问题,同时增强水印方法的安全性,本文将两路水印信号与一路等长的混沌序列进行混合,得到嵌入水印信号。然后,利用线性混合方法,将嵌入水印信号与选定的小波系数进行混合,得到隐秘信号。水印提取时,利用独立分量分析算法,提取嵌入水印信号,再经过后处理过程,得到原始水印。该水印方案是一种盲水印方法,可以将多个作者信息同时嵌入到音频作品中,而不需要考虑水印的嵌入顺序。实验结果表明,该方法对常规的信号处理操作具有良好的鲁棒性,以及良好的抵抗时间轴同步攻击的能力。     

Sequential Blind Source Extraction For Quasi-Periodic Signals With Time-Varying Period

A novel second-order-statistics-based sequential blind extraction algorithm for blind extraction of quasi-periodic signals, with time-varying period, is introduced in this paper. Source extraction is performed by sequentially converging to a solution that effectively diagonalizes autocorrelation matrices at lags corresponding to the time-varying period, which thereby explicitly exploits a key statistical nonstationary characteristic of the desired source. The algorithm is shown to have fast convergence and yields significant improvement in signal-to-interference ratio as compared to when the algorithm assumes a fixed period. The algorithm is further evaluated on the problem of separation of a heart sound signal from real-world lung sound recordings. Separation results confirm the utility of the introduced approach, and listening tests are employed to further corroborate the results.     

一种卫星隐蔽通信信号盲分离算法

重构抵消算法是卫星隐蔽通信信号分离的关键技术,算法的性能主要依赖于参数估计的精度。然而在实际环境中,参数估计误差带来的算法性能损失无法被避免。此外,对信号的重构使得该算法计算复杂度较高。针对这个问题,首先分析了参数估计误差对分离性能的影响,然后提出基于盲均衡算法的协作分离算法,提升信号分离性能的同时降低了算法的计算量。仿真实验表明,新算法相较于重构抵消算法,降低了对参数估计精度的依赖,当参数估计误差大于0.05时,信号的解调误码率降低了一个数量级左右。     

应用萤火虫算法的混沌信号盲源分离

针对现有的独立成分分析法分离混合混沌信号精度不理想的问题,提出了一种新的混沌信号盲分离方法。该方法以求解最优解混矩阵为目标,利用峭度构造目标函数,将混沌信号的盲源分离转化为一个优化问题,并用萤火虫算法求解。同时,通过预白化和正交矩阵的参数化表示降低优化问题的维数,能有效提高分离精度。仿真结果表明,无论是处理混合的混沌映射信号还是混合的混沌流信号,该方法都能快速收敛,并且其分离精度在各项实验中都优于独立成分分析法等现有的盲源分离方法。     

On the Use of Spatial Time Frequency Distributions for Signal Extraction

This paper deals with the extraction of signals from their instantaneous linear mixtures using time-frequency distributions. Fundamentally, this problem is a signal synthesis from the time-frequency (t-f) plane. However with the incorporation of the spatial information provided by a multisensor array, the problem can be posed as special case of blind source separation. So far, the blind source separation has been solved using only statistical information available on the source signals. Herein, we propose to solve the aforementioned problem using time-frequency signal representations and the spatial array aperture. The proposed approach relies on the difference in the t-f signatures of the sources to be separated. It is based on the diagonalization of a combined set of spatial time-frequency distribution matrices. A numerical example is provided to illustrate the effectiveness of our method.