Method and apparatus for noise cancellation

• US 9,131,915 B2
• Filed: 07/06/2012
• Issued: 09/15/2015
• Est. Priority Date: 07/06/2011
• Status: Expired due to Fees

• Current Assignee: University of New Brunswick
• Original Assignee: University of New Brunswick
• Inventors: Amiri Farahani, Mohsen, Guerra, Eduardo Castillo, Colpitts, Bruce G., Brown, Anthony W.
• Primary Examiner(s): Kim, Kevin
• Application Number:
  US13/543,658
• Publication Number:
  US 20130011144A1
• Time in Patent Office: 1,166 Days
• Field of Search: 375/232, 375/348, 375/346, 375/349
• US Class Current: 1/1
• CPC Class Codes: A61B 5/318 Heart-related electrical modalities, e.g. electrocardiography [ECG]
• CPC Class Codes: A61B 5/725 using specific filters therefor, e.g. Kalman or adaptive filters A61B5/316 take precedence
• CPC Class Codes: A61B 6/5258 involving detection or reduction of artifacts or noise
• CPC Class Codes: G01D 3/032 affecting incoming signal, e.g. by averaging; gating undesired signals
• CPC Class Codes: G01D 5/35364 using inelastic backscattering to detect the measured quantity, e.g. using Brillouin or Raman backscattering

Abstract

Method and apparatus of denoising recurrent signals comprising: providing first and second noisy signals each comprising a deterministic signal and (different realizations of) white noise; selecting one of the noisy signals as an input signal and the other noisy signal as a desired signal; providing an estimate of the deterministic signal in the input signal; determining the error between the estimated deterministic signal and the desired signal; adjusting the estimate of the deterministic signal using the error and producing an adjusted estimate of the deterministic signal.

Claims

1. A method of denoising recurrent signals comprising:

providing first and second noisy signals each comprising a deterministic signal and a white noise;

selecting one of the noisy signals as an input signal and the other noisy signal as a desired signal;

providing an estimate of the deterministic signal in the input signal;

determining the error between the estimated deterministic signal and the desired signal;

adjusting the estimate of the deterministic signal using the error; and

producing an adjusted estimate of the deterministic signal.

2. The method of claim 1 wherein the white noise is a different realization of white noise.

3. The method of claim 1 wherein the white noise in the first signal is different from the white noise in the second signal.

4. The method of claim 1 wherein the step of adjusting of the deterministic signal is carried out by minimizing the error using a given criterion.

5. The method of claim 4 wherein the step of minimizing the error is carried out using an error minimization method selected from the group consisting of least mean squares, normalized least mean squares, recursive least squares, affine projection, fast affine projection, fast Newton transversal and Kalman.

6. The method of claim 1 wherein the noisy signals are recurrent signals.

7. The method of claim 6 wherein the recurrent signal is from an optical sensor.

8. The method of claim 7 wherein the optical sensor is an SBS sensor.

9. The method of claim 7 wherein the optical sensor is a BOTDA sensor.