Torpedo Detection using Multi-Path Signals and Fast Orthogonal Search Techniques


  1. McGaughey, C.D.R.
  2. Theriault, J.A.
  3. Pecknold, S.
Corporate Authors
Defence R&D Canada - Atlantic, Dartmouth NS (CAN)
Detecting a high-speed torpedo by means of a passive acoustic detector is very challenging for most acoustic operators. Coupled with a very noisy environment, multiple sources in a multi-path scenario and varying environmental factors, a time-constrained assessment will prove difficult. In addition, a passive sensor cannot estimate the range of a torpedo approaching it at a constant bearing with conventional signal processing. The passive acoustic sensor will receive a direct path signal from the torpedo as well as a signal that has reflected off the surface. Due to the different angles of arrival, the direct-path and surface-reflected signals have different Doppler shifts. This paper introduces a torpedo detection algorithm, which is primarily developed in MATLAB. The Torpedo Detection Algorithm (TDA), employs the fast orthogonal search (FOS) algorithm for high-resolution spectral analysis to detect the closely spaced direct-path and surface-reflection signals. When a direct-path and surface-reflection are found, an automatic alert of a torpedo detection is initiated. In simulation, a torpedo is detected 20 times out of 20 as it travels from 5000 to 750 metres from the receiver. Simple trigonometric expressions are used to estimate the torpedo’s range given the two frequencies estimated by FOS and apriori information about the torpedo speed and depth. The predicted range for a simulation in which a torpedo approaches from 5000 to 750 metres is shown.

Il y a un résumé en français ici.

Torpedo;signal processing;detection
Report Number
DRDC-ATLANTIC-SL-2004-270 — Scientific Literature
Date of publication
01 Dec 2005
Number of Pages
Electronic Document(PDF)

Permanent link

Document 1 of 1

Date modified: