Volterra system identification and Kramer's sampling theorem

  • Authors:

  • R.J. Martin

  • Affiliations:

  • Inst. for Telecommun. Res., Adelaide, SA, Australia

  • Venue:
  • IEEE Transactions on Signal Processing
  • Year:
  • 1999

Quantified Score

Hi-index 35.69

Visualization

Abstract

Nonlinear systems increase signal bandwidth, but Volterra systems may be identified uniquely even when the output is sampled at the Nyquist rate of the input. We present a more transparent and general proof. Using the theory of sampling expansions for general integral transforms, we permit the sampling to be nonuniform