论文部分内容阅读
由于检测有色高斯噪声的秩-1高斯信号可以采用似然比测试统计形式,根据这一启发,在检测弱正弦信号之前我们把早期的低-秩信号估值应用到估计和消除强正弦干扰波形问题上。我们认为,在很短的时间间隔或空间内要得到数据采样是很困难的,而且正弦信号和正弦干扰的频率间隔比窗口范围的倒数的间隔更密。本文介绍的一种方法可以用到非正弦和/或随机信号和干扰两种情况上。最重要的假设是:当干扰采样以矩阵的形式排列时,并且在仅干扰矩阵由于高概率而近似为低-秩矩阵时,这个矩阵将近似地有个低-秩。
Since the rank-1 Gaussian signal for the detection of colored Gaussian noise can be used as the likelihood ratio test statistic, based on this, we apply early low-rank signal estimates to the estimation and cancellation of strong sinusoidal interference waveforms prior to the detection of weak sinusoids Question In our opinion, it is difficult to obtain data samples within a short time interval or space, and the frequency interval between the sinusoidal signal and the sinusoidal interference is more dense than the reciprocal of the window range. This article describes a method that can be used for both non-sinusoidal and / or random signals and disturbances. The most important assumption is that when the interference samples are arranged in a matrix, and only the interference matrix approximates a low-rank matrix due to the high probability, this matrix will have approximately a low-rank.