Title :
Volterra signal modelling using Lagrange programming neural networks
fDate :
31 Aug-2 Sep 1998
Abstract :
Adaptive Volterra signal modelling is a multidimensional problem that involves high computational complexity. In recent years, neural networks offer a possible solution towards the implementation in real time, of adaptive Volterra signal modelling schemes. However, neural networks with a moderately large number of neurons poses a problem in VLSI realization. In this paper, the Lagrange programming neural network (LPNN) is shown to provide a methodology for modelling Volterra signals adaptively and thus providing a suitable framework for VLSI realization. In this paper, the least mean square solution and the solution to an approximation of the Chebychev norm are presented
Keywords :
Chebyshev approximation; VLSI; Volterra series; computational complexity; least mean squares methods; modelling; neural chips; neural nets; real-time systems; signal processing; Chebychev norm approximation; LPNN; Lagrange programming neural networks; VLSI realization; adaptive Volterra signal modelling; computational complexity; least mean square solution; multidimensional problem; real-time implementation; Adaptive systems; Chebyshev approximation; Computational complexity; Equations; Lagrangian functions; Least squares approximation; Multidimensional systems; Neural networks; Neurons; Very large scale integration;
Conference_Titel :
Neural Networks for Signal Processing VIII, 1998. Proceedings of the 1998 IEEE Signal Processing Society Workshop
Conference_Location :
Cambridge
Print_ISBN :
0-7803-5060-X
DOI :
10.1109/NNSP.1998.710656
