Aerospace Instrument-Making Annotation << Back DEVELOPMENT OF ALGORITHMS OF THE AIRCRAFT NOISE SUPPRESSION SYSTEM BY THE METHOD OF DIRECT BLIND SOURCE SEPARATION A. Dramchini The paper investigates a way to improve the voice communication quality in airplanes in order to improve the flight safety. The problem of separating a useful signal, distorted by additive noise, from two mixed signals, each of which is the sum of the useful signal and noise, is considered. An algorithm for noise elimination and speech signal extraction is proposed to improve the quality of voice communication between the pilot and the controller and between the pilot and passengers. The algorithm uses Wiener filtering, least squares method and direct structure of blind signal separation. Keywords: communication, signal, noise, speech, mix of signals, LMS algorithm, Wiener filter, blind signal separation, light aircraft, flight safety. DOI: 10.25791/aviakosmos.6.2022.1285 Pp. 40-50.