<![CDATA[The life signals are built on harmonic mode for their low frequency, quasi-periodicity, low SNR, and the easy submerged in strong clutter noise. The method for detecting life signal based on adaptive filter and high order statistics is presented, in which neither the Gaussian supposition of the observed signal, nor a prior information about the waveform and arrival time of the observed signal is necessary. The principle of method is to separate the spectrum of input signal into many narrow frequency bands, whose Sub-band signal is followed by a short-time estimation of higher-order statistics so as to suppress Gaussian noises. Simulated results show that the method can effectively detect life signals from noise with good convergence speed and stability, and greatly improve the signal quality with respect to LMS method. DOI: http://dx.doi.org/10.11591/telkomnika.v10i6.1597 ]]>
Copyrights © 2012
