E.N. Efimov - Post-graduate Student, Department 405, Moscow Aviation Institute (MAI). E-mail: omegatype@gmail.com D.V. Filimonova - Student, Department 405, Moscow Aviation Institute (MAI). E-mail: daria-fili@rambler.ru T.Y. Shevgunov - Ph. D. (Eng.), Associate Professor, Department 405, Moscow Aviation Institute (MAI). E-mail: shevgunov@mai-trt.ru

This paper introduces an approach for implementing maximum likelihood parameter estimators using feedforward artificial neural network of multilayer perceptron architecture. A theoretical foundation of the proposed approach is presented in the assumption that the model of observation is known as well as the values of its vector of parameters. For a practical example the implementation of direction of arrival estimator for the active ring antenna array is shown. In order to estimate a performance and accuracy of the proposed approach, the results of numerical calculation are presented, compared to the algorithm based on optimal numerical solution and referenced to Cramer-Rao lower bound. The results also indicate that there is no significant dependency of the accuracy of estimation on actual parameter value. Moreover, the calculations take significantly less time, although some of it is spent on the initial training of the neural network.

 

1. Galushkin A.I. Nejjronnye seti. Osnovy teorii. M.: Gorjachaja Linija-Telekom. 2012. 496 s.
2. KHajjkin S. Nejjronnye seti: polnyjj kurs. Izd. 2-e., ispr. / Per. s angl. M.: OOO «I.D. Viljams». 2006. 1104 c.
3. Baum E. Supervised Learning of Probability Distributions by Neural Networks. AmericanInstituteofPhysics. 1988. P. 52−61.
4. Setiono R. A neural network construction algorithm which maximizes the likelihood function // Connection Science. 1995.V. 7. № 2. P. 147−166.
5. R. Battiti First- and Second-Order Methods for Learning: Between Steepest Descent and Newton-s Method // Neural Computation. 1992. V. 4. № 2. P. 141−166.
6. C. Cervellera, D. Maccio, M. Muselli Deterministic learning for maximum-likelihood estimation through neural networks // IEEE Transactions on Neural Networks. 2008. V. 19.№ 8. P. 1456−1467.
7. D.W. Marquardt An algorithm for least-squares estimation of non-linear parameters // J. Soc. Ind. Appl. Math. 1963.V. 11. P. 431−441.
8. Efimov E.N., SHevgunov T.JA. Formirovanie ocenki napravlenija prikhoda signala s ispolzovaniem iskusstvennykh nejjronnykh setejj // 70-ja Mezhdunar. konf. «Radioehlektronnye ustrojjstva i sistemy dlja infokommunikacionnykh tekhnologijj» (REDS-2015). M.: MTUSI. 2015. S. 98−101.
9. Efimov E.N., SHevgunov T.JA. Postroenie nejjronnykh setejj prjamogo rasprostranenija s ispolzovaniem adaptivnykh ehlementov // ZHurnal radioehlektroniki. M.: IREH RAN. 2012.№ 8.
10. Dubrovin A.V. Potencialnaja tochnost pelengovanija kompleksami s antennymi reshetkami, imejushhimi konfiguraciju v vide nabora proizvolnogo chisla kolec // Radiotekhnika i ehlektronika. 2006. T. 51. № 13. S. 268−270.