Walsh Transform Pulse Detection in High Noise Environments
Abstract
The Walsh Hadamard Transformation is shown through computer generated graphics to be very useful for detecting rectangularly shaped information carrying pulses that exist in high noise environments. This transformation is used to process the time domain input signal and determine if one or more pulses are present in the noise. A Fast Walsh Transform (FWT) algorithm is used to generate the sequency domain coefficients. It is shown that by low pass filtering of the Walsh Spectrum coefficients, the pulse can be detected since the higher order coefficient contribute only to the noise. A combination of low pass filtering and amplitude thresholding of the Walsh spectrum is shown to provide an optimal signal detection capability. (Author)
Document Details
- Document Type
- Technical Report
- Publication Date
- Mar 01, 1977
- Accession Number
- ADA038651
Entities
People
- Ronald M. Okupski
Organizations
- Rome Laboratory