Parametric Homomorphic Deconvolution (Prony, ARMA)

Parametric homomorphic deconvolution using Prony (ARMA) modeling to recover time-delay structure from a single-channel signal
6 Downloads
Updated 12 Aug 2025

View License

Purpose.
Recover the propagation (time-delay) response from a single-channel analog-summed microphone signal using homomorphic deconvolution followed by Prony (ARMA) modeling. Returns the parametric HD magnitude plus numerator/denominator coefficients and poles—handy for compact feature extraction and AoA regression.
[z22, vp2, Num2, Den2] = cep_prony6(Aord, data0, N, W, MM, novlp, normonoff, plotonoff)
Inputs
  • Aord — Prony model order (AR and MA orders both set to Aord).
  • data0 — Input vector; length must equal N + novlp*(MM-1).
  • N — Frame length (power of two recommended).
  • W — Quefrency gate length (samples).
  • MM — Ensemble length (# of frames to average).
  • novlpHop size (overlap = N - novlp).
  • normonoff1 to normalize z22 to unity max.
  • plotonoff1 to show HD magnitude and pole/zero plot.
Outputs
  • z22[N/2+1 × 1] parametric HD magnitude (DC…Nyquist).
  • vp2 — Prony poles (roots(Den2)).
  • Num2, Den2 — Prony MA/AR coefficients.

Cite As

Keonwook Kim (2026). Parametric Homomorphic Deconvolution (Prony, ARMA) (https://www.mathworks.com/matlabcentral/fileexchange/181788-parametric-homomorphic-deconvolution-prony-arma), MATLAB Central File Exchange. Retrieved .

Park, Yeonseok, et al. “Parametric Estimations Based on Homomorphic Deconvolution for Time of Flight in Sound Source Localization System.” Sensors, vol. 20, no. 3, Feb. 2020, p. 925, https://doi.org/10.3390/s20030925.

View more styles
MATLAB Release Compatibility
Created with R2021a
Compatible with R2021a to R2025a
Platform Compatibility
Windows macOS Linux
Version Published Release Notes
1.0.0