A temporal neural network for the noise subspace of the array signal
Guojie Dong, Ruey-Wen Liu · 2002
In certain array signal processing problems, it is necessary to find the signal or noise subspace. Several neural networks have been presented to perform the principal Component Analysis (PCA), which can be used to find the signal and noise subspace. However, under certain situations, it is more efficient to find noise subspace directly. In this paper, we present a neural network to find the noise subspace directly. The neural network has a constant learning rate, and globally converged to the solution.