作为非平稳信号的重要特征,瞬时频率(instantaneous frequency,IF)和瞬时调频率(instantaneous frequency rate,IFR)的准确估计具有重要意义。现有方法在处理存在时频交叠的多分量非平稳信号时易发生关联错误等问题。短时调频傅里叶变...作为非平稳信号的重要特征,瞬时频率(instantaneous frequency,IF)和瞬时调频率(instantaneous frequency rate,IFR)的准确估计具有重要意义。现有方法在处理存在时频交叠的多分量非平稳信号时易发生关联错误等问题。短时调频傅里叶变换通过将信号在时间频率调频率三维空间中进行表征,使不同分量发生交叠的可能性大幅降低,且基于频率调频率的变化规律可实现分量的时序关联。据此,提出一种基于检测跟踪算法的多分量IF-IFR估计方法。首先,针对传统检测算法在噪声环境下精度不足问题,提出了基于改进YOLOX网络的检测方法,实现了信号瞬时频率调频率的估计和瞬时形状特征的提取。然后,提出基于卡尔曼滤波的瞬时估计值和形状特征时序关联方法,以形成稳定连续的多分量IF和IFR估计。通过仿真及实测实验对所提算法进行了验证,在设置的仿真场景中,-5 dB信噪比条件下最优估计误差小于0.8 Hz,证明了所提方法的有效性。展开更多
A generalized approach for narrowband interference （NBI） suppression in direct sequence spread spectrum （DSSS） communication systems using adaptive infinite impulse response （IIR） filter is presented. The excisi...A generalized approach for narrowband interference （NBI） suppression in direct sequence spread spectrum （DSSS） communication systems using adaptive infinite impulse response （IIR） filter is presented. The excision filter coefficients depend on both the jammer power and its instantaneous frequency. The dependency of the filter construction on the jammer power is significant as it allows optimal tradeoff between interference removal and signal distortion by maximizing the receiver signal to noise ratio improvement（SNRI）. Instead of traditional adaptive line enhancer （ALE） estimator, a preferable NBI estimator-Fourier interpolation estimator （FIE） is proposed. Closed-form expressions of the SNR improvement and theoretical bit error rate （BER） based on the assumption that the output of the correlator is Gaussian distributed are both derived. Performance results obtained by numerical simulation are also presented and compared with theoretical results.展开更多
基金Sponsored by the Beijing Municipal Natural Science Foundation(4052024)
文摘A generalized approach for narrowband interference （NBI） suppression in direct sequence spread spectrum （DSSS） communication systems using adaptive infinite impulse response （IIR） filter is presented. The excision filter coefficients depend on both the jammer power and its instantaneous frequency. The dependency of the filter construction on the jammer power is significant as it allows optimal tradeoff between interference removal and signal distortion by maximizing the receiver signal to noise ratio improvement（SNRI）. Instead of traditional adaptive line enhancer （ALE） estimator, a preferable NBI estimator-Fourier interpolation estimator （FIE） is proposed. Closed-form expressions of the SNR improvement and theoretical bit error rate （BER） based on the assumption that the output of the correlator is Gaussian distributed are both derived. Performance results obtained by numerical simulation are also presented and compared with theoretical results.
