What is it about?

The paper investigates how the performance of a matched filter varies when the noise changes from stationary to non-stationary. Matched filter is known to maximise the SNR of the received signal when the noise in the model is assumed to be white and non-stationary. In underwater acoustic sensing applications, the non-constant gain to compensate for transmission loss often restores the received signal but changes the noise in the received signal, so that the noise is no longer stationary. A matched filter would no longer be an optimal receiver filter for such a scenario.However, we prove that a filter designed to maximise the SNR for a signal signal received in such noise is conditionally optimal for very long signals and at short ranges only; the matched filter is still the best choice for all signal lengths and all ranges.

Featured Image

Why is it important?

The research strengthens the fact the matched filter, a generic receiver filter, known to maximise the SNR of the received signal, is the optimal choice, regardless of the type of the noise. This is, at least true for underwater sensing applications, where a Time Variable Gain is used to compensate for non-constant transmission loss.


This research should invite some interest into viability and suitability of a matched filter for many other scenarios, beyond the scope of this paper. We believe it would be interesting to know how a matched filter would perform against another filter, that has been designed to optimise (maximise) SNR in a changed scenario.

Victoria University of Wellington

Read the Original

This page is a summary of: The Sensitivity of the Matched Filter Signal to Noise Ratio and Effects of Time Variable Gain for Sonar Systems, IET Radar Sonar & Navigation, December 2019, the Institution of Engineering and Technology (the IET), DOI: 10.1049/iet-rsn.2019.0334.
You can read the full text:




The following have contributed to this page