Rousseau, David and Anand, GV and Chapeau, Francois B (2006) Noise-enhanced nonlinear detector to improve signal detection in non-Gaussian noise. In: Signal Processing, 86 (11). pp. 3456-3465.
Restricted to Registered users only
Download (297Kb) | Request a copy
We compare the performance of two detection schemes in charge of detecting the presence of a signal buried in an additive noise. One of these is the correlation receiver (linear detector), which is optimal when the noise is Gaussian. The other detector is obtained by applying the same correlation receiver to the output of a nonlinear preprocessor formed by a summing parallel array of two-state quantizers. We show that the performance of the collective detection realized by the array can benefit from an injection of independent noises purposely added on each individual quantizer. We show that this nonlinear detector can achieve better performance compared to the linear detector for various situations of non-Gaussian noise. This occurs for both Bayesian and Neyman–Pearson detection strategies with periodic and aperiodic signals.
|Item Type:||Journal Article|
|Additional Information:||Copyright of this article belongs to Elsevier.|
|Keywords:||Detection;Quantizer;Nonlinear arrays;Suprathreshold stochastic resonance|
|Department/Centre:||Division of Electrical Sciences > Electrical Communication Engineering|
|Date Deposited:||26 May 2008|
|Last Modified:||19 Sep 2010 04:45|
Actions (login required)