Gradient Adaptive Step Size Research Articles

Signed-gradient adaptive step size LMS algorithm for biomedical applications.

Gradient adaptive step size adaptive filters have been widely used to adapt different biomedical application environments and obtain useful life signals from serious ambient noise and interferences. In order to further improve the signal-to-noise ratio (SNR) of the life signals, this paper presents a class of signed-gradient adaptive step size least mean square (LMS) adaptive filters. The proposed algorithms introduce a sign function to replace the gradient of squared error in the step size updating process of the gradient adaptive step size LMS adaptive filters. The performance of both gradient and signed-gradient algorithms with dual adaptive filters is compared by extracting heartbeat signals from ambient noise in stethoscopes. Simulation results demonstrate that though the signed-gradient adaptive step size LMS algorithm converges at a slower rate at the early stage of iteration, it has a smaller mean squared error (MSE) at the stage of convergence, thus achieves a higher SNR.

Stochastic Gradient-Adaptive Complex-Valued Nonlinear Neural Adaptive Filters With a Gradient-Adaptive Step Size

A class of variable step-size learning algorithms for complex-valued nonlinear adaptive finite impulse response (FIR) filters is proposed. To achieve this, first a general complex-valued nonlinear gradient-descent (CNGD) algorithm with a fully complex nonlinear activation function is derived. To improve the convergence and robustness of CNGD, we further introduce a gradient-adaptive step size to give a class of variable step-size CNGD (VSCNGD) algorithms. The analysis and simulations show the proposed class of algorithms exhibiting fast convergence and being able to track nonlinear and nonstationary complex-valued signals. To support the derivation, an analysis of stability and computational complexity of the proposed algorithms is provided. Simulations on colored, nonlinear, and real-world complex-valued signals support the analysis.

Improved stochastic gradient adaptive filter with gradient adaptive step size

The performance of a stochastic gradient adaptive filter can be significantly improved by introducing a forgetting factor. The complexity of the original algorithm can also be reduced by using only the signs of error signals and input signals in the gradient adaptive step size computation.