Low-Pass NGD Voice Signal Sensoring With Passive Circuit
Fayu Wan, Zhongzhu Yuan, Blaise Ravelo, Junxiang Ge, Wenceslas Rahajandraibe
Abstract
Nowadays, the speech signal processing is still challenging with the multimedia ubiquity. Classical signal processing techniques as filtering were developed for the speech enhancement to overcome unintentional effects as the propagation channel. However, the delay effect remains an open challenge for audio signal processing engineers. A voice/song signal (VSS) prediction is investigated theoretically and experimentally in this paper. This VSS engineering allows to predict an arbitrary waveform signal by using a lowpass negative group delay (NGD) function. This outstanding function presents the VSS advance properties when its parameters are matched with the operated signals. This NGD engineering analytical approach is described with the audio signal processing. The low-pass NGD analysis and property are deeply investigated with two NGD RC-and RL-passive cells. NGD transient study is performed with parametric analytical signals presenting raised cosine waveform. The introduced VSS NGD sensoring is verified with experimental real-time tests of microphone VSS generator. VSS time-advance of approximately 50 μs was observed with a good agreement between simulations and measurements.