Military aircrafts' classification based on their sound signature

Barbarosou, M; Paraskevas, I; Ahmed, A

Abstract

Purpose: This paper aims to present a system framework for classifying different models of military aircrafts, which is based on the sound they produce.

Design/methodology/approach: The technique is based on extracting a compact feature set, of only two features, extracted from the frequency domain of the aircrafts’ sound signals produced by their engines, namely, the spectral centroid and the signal bandwidth. These features are then introduced to an artificial neural network to classify the aircraft signals.

Findings: The current system identifies the aircraft type among four military aircrafts: Mirage 2000, F-16 Fighting Falcon, F-4 Phantom II and F-104 Starfighter. The experimental results show that the aforementioned types of aircrafts can be accurately classified up to 96.2 per cent via the proposed method.

Practical implications: The proposed system can be used as a low-cost assistive tool to the already existing radar systems to avoid cases of missed detection or false alarm. More importantly, the same method can be used for aircrafts that use stealth technology that cannot be detected using radar devices.

Originality/value: The proposed method constitutes a novel approach to classifying military aircrafts based on their sound signature. It utilizes only two spectral features extracted from the sound of the aircraft engine; these features are then introduced to a neural network classifier.

Citation

Barbarosou, M., Paraskevas, I., & Ahmed, A. (2016). Military aircrafts' classification based on their sound signature. Aircraft engineering, 88(1), 66-72. https://doi.org/10.1108/AEAT-04-2014-0040

• Journal Article Type: Article
• Acceptance Date: Nov 3, 2014
• Online Publication Date: Jan 4, 2016
• Publication Date: Jan 4, 2016
• Deposit Date: Mar 18, 2019
• Journal: Aircraft Engineering and Aerospace Technology
• Print ISSN: 0002-2667
• Publisher: Emerald
• Volume: 88
• Issue: 1
• Pages: 66-72
• DOI: https://doi.org/10.1108/AEAT-04-2014-0040
• Publisher URL: https://doi.org/10.1108/AEAT-04-2014-0040
• Related Public URLs: https://www.emeraldinsight.com/loi/aeat