Advertisement

Neural Network Detectors for Composite Hypothesis Tests

  • D. de la Mata-Moya
  • P. Jarabo-Amores
  • R. Vicen-Bueno
  • M. Rosa-Zurera
  • F. López-Ferreras
Part of the Lecture Notes in Computer Science book series (LNCS, volume 4224)

Abstract

Neural networks (NNs) are proposed for approximating the Average Likelihood Ratio (ALR). The detection of gaussian targets with gaussian autocorrelation function and unknown one-lag correlation coefficient, ρ s , in Additive White Gaussian Noise (AWGN) is considered. After proving the low robustness of the likelihood ratio (LR) detector with respect to ρ s , the ALR detector assuming a uniform distribution of this parameter in [0,1] has been studied. Due to the complexity of the involved integral, two NN based solutions are proposed. Firstly, single Multi-Layer Perceptrons (MLPs) are trained with target patterns with ρ s varying in [0,1]. This scheme outperforms the LR detector designed for a fixed value of ρ s . MLP with 17 hidden neurons is proposed as a solution. Then, two MLPs trained with target patterns with ρ s varying in [0,0.5] and [0.5,1], respectively, are combined. This scheme outperforms the single MLP and allows to determine a solution of compromise between complexity and approximation error. A detector composed of MLPs with 17 and 8 hidden units each one is proposed.

Keywords

Additive White Gaussian Noise Hide Neuron Hide Unit Target Pattern Importance Sampling Technique 
These keywords were added by machine and not by the authors. This process is experimental and the keywords may be updated as the learning algorithm improves.

Preview

Unable to display preview. Download preview PDF.

Unable to display preview. Download preview PDF.

References

  1. Van Trees, H.L.: Detection, estimation, and modulation theory, vol. 1. Wiley, Chichester (1968)MATHGoogle Scholar
  2. Aref, M.R., Nayebi, M.M.: Likelihood-ratio detection. In: IEEE Int. Symp. on Information Theory, Trondheim, Norway, p. 260 (1994)Google Scholar
  3. Ruck, D.W., Rogers, S.K., Kabrisky, M., Oxley, M.E., Suter, B.W.: The multilayer perceptron as an aproximation to a Bayes optimal discriminant function. IEEE Transactions on Neural Networks 1(4), 296–298 (1990)CrossRefGoogle Scholar
  4. Wan, E.A.: Neural network classification: a bayesian interpretation. IEEE Transactions on Neural Networks 1(4), 303–305 (1990)CrossRefGoogle Scholar
  5. Richard, M.D., Lippmann, R.P.: Neural network classifiers estimate Bayessian a posteriori probabilities. Neural Computation 3, 461–483 (1991)CrossRefGoogle Scholar
  6. Jarabo-Amores, P., Rosa-Zurera, M., Gil-Pita, R., López-Ferreras, F.: Suficient Condition for an Adaptive System to Approximate the Neyman-Pearson Detector. In: Proc. IEEE Workshop on Statistical Signal Processing, Bordeaux, France (2005)Google Scholar
  7. Gandhi, P.P., Ramamurti, V.: Neural networks for signal detection in non-gaussian noise. IEEE Transactions on Signal Processing 45(11), 2846–2851 (1997)CrossRefGoogle Scholar
  8. Andina, D., Sanz-Gonzlez, J.L.: Comparison of a neural network detector vs. Neyman-Pearson optimal detector. In: Proc. ICASSP 1996, Atlanta, GA, pp. 3573–3576 (1996)Google Scholar
  9. Munro, D.J., Ersoy, O.K., Bell, M.R., Sadowsky, J.S.: Neural network learning of low-probability events. IEEE Transactions on Aerospace and Electronic Systems 32(3), 898–910 (1996)CrossRefGoogle Scholar
  10. Mata-Moya, D., Jarabo-Amores, P., Rosa-Zurera, M., López-Ferreras, F., Vicen- Bueno, R.: Approximating the Neyman-Pearson detector for Swerling I Targets with Low Complexity Neural Networks. In: Duch, W., Kacprzyk, J., Oja, E., Zadrożny, S. (eds.) ICANN 2005. LNCS, vol. 3697, pp. 917–922. Springer, Heidelberg (2005)Google Scholar
  11. Jarabo-Amores, P., Gil-Pita, R., Rosa-Zurera, M., López-Ferreras, F.: MLP and RBFN for detecting white gaussian signals in white gaussian interference. In: IWANN 2003. LNCS, vol. 2687, pp. 790–797. Springer, Heidelberg (2003)Google Scholar
  12. El-Jaroudi, A.A., Makhoul, J.: A new error criterion for posterior probability estimation with neural nets. In: Proc. of the Int. Conf. on Neural Networks IJCNN, U.S.A., June 1990, pp. 185–192 (1990)Google Scholar
  13. Nguyen, D., Widrow, B.: Improving the learning speed of 2-layer neural networks by choosing initial values of the adaptive weights. In: Proc. of the Int. Joint Conf. on Neural Networks, pp. 21–26 (1990)Google Scholar
  14. Sanz-Gonzalez, J.L., Andina, D.: Performance analysis of neural network detectors by importance sampling techniques. In: Neural Proc. Letters, vol. 9, pp. 257–269 (1999)Google Scholar

Copyright information

© Springer-Verlag Berlin Heidelberg 2006

Authors and Affiliations

  • D. de la Mata-Moya
    • 1
  • P. Jarabo-Amores
    • 1
  • R. Vicen-Bueno
    • 1
  • M. Rosa-Zurera
    • 1
  • F. López-Ferreras
    • 1
  1. 1.Departamento de Teoría de la Señal y Comunicaciones, Escuela Politécnica SuperiorUniversidad de AlcaláAlcalá de Henares, MadridSpain

Personalised recommendations