In this paper, we propose an extension of the well-known FLANN filter using trigonometric expansions to include suitable cross-terms, i.e., products input samples with different time shifts. It is shown that in some applications nonlinear active noise control, generalized can offer performance as good high-order Volterra filters a reduced complexity.

SmartRoadSense is a crowdsensing project aimed at monitoring the conditions of road surface. Using sensors smartphone, monitors vertical accelerations inside vehicle traveling and extracts roughness index conveying information about conditions. The smartphone GPS data are periodically sent to central server where they processed, associated with specific road, aggregated measured by other smartphones. This paper studies how related speed. It shown that dependence can be locally approximated...

The paper introduces two improvements in the feedforward active noise control system with online secondary path modeling developed by Akhtar, Abe, and Kawamata: (1) optimal variable step-size parameters are derived for adaptation algorithms of filter (2) a self-tuning power scheduling auxiliary is introduced. proposed chosen so that every operating condition specific ratio between powers at error microphone residual achieved. It shown same conditions equipped improve convergence speed...

In this paper, a bounded-input bounded-output (BIBO) stability condition for the recursive functional link artificial neural network (FLANN) filter, based on trigonometric expansions, is derived. This filter considered as member of class causal shift-invariant nonlinear filters whose output depends linearly coefficients. As all filters, its should be granted or, at least, tested. The relevant conclusion we derive from that FLANN not affected by instabilities whenever linear part stable. fact...

Room response equalization aims at improving the sound reproduction in rooms by applying advanced digital signal processing techniques to design an equalizer on basis of one or more measurements room response. This topic has been intensively studied last 40 years, resulting a number effective facing different aspects problem. review paper giving overview existing methods following their historical evolution, and discussing pros cons each approach with relation characteristics, as well...

A multichannel controller based on Volterra filters is described. filtered-X affine projection algorithm derived in detail for homogeneous quadratic filters. The proposed can be also extended to higher-order kernels and includes linear controllers as a particular case.

This paper provides an analysis of transient and steady-state behavior different filtered-x affine projection algorithms. Algorithms suitable for single-channel multichannel active noise controllers are treated within a unified framework. Very mild assumptions posed on the control system model, which is only required to have linear dependence output from filter coefficients. Therefore, applies not finite impulse response models but also nonlinear Volterra filters, i.e., polynomial other...

In this paper, a novel sub-class of linear-in-the-parameters (LIP) nonlinear filters, formed by the so-called even mirror Fourier (EMFN) is presented. These filters are universal approximators for causal, time invariant, finite-memory, continuous systems as well-known Volterra filters. However, in contrast to their basis functions mutually orthogonal white uniform input signals. Therefore, adaptive applications, gradient descent algorithms with fast convergence speed and efficient system...

The multiple-variance method is a cross-correlation that exploits input signals with different powers for the identification of nonlinear system by means Volterra series.It overcomes problem locality solution traditional methods, based on mean square error minimization or cross-correlation, well approximate only inputs have approximately same power signal.The permits to improve performance models systems high dynamic, like audio amplifiers.This used, first time, identify three tube...

We discuss the optimal regularization parameter of Filtered- Affine Projection (FX-AP) algorithm suitable for feedforward active noise control. While original FX-AP always provides a biased estimate minimum-mean-square solution, we show that regularized is capable to eliminate bias asymptotic solution and thus can optimize both convergence speed residual MSE algorithm. derive some expressions parameter, heuristic estimations in practical conditions.

This paper introduces a novel sub-class of linear-in-the-parameters nonlinear filters, the Legendre filters. Their basis functions are polynomials, specifically, products polynomial expansions input signal samples. filters share many properties recently introduced classes Fourier and even mirror which based on trigonometric functions. In fact, universal approximators for causal, time invariant, finite-memory, continuous, systems their mutually orthogonal white uniform signals. adaptive...

In this paper, a novel class of recursive nonlinear filters is presented as an extension finite-memory even mirror Fourier filters. These are characterized by two relevant properties: i) they able to arbitrarily well approximate any discrete-time, time-invariant, causal, infinite-memory, continuous, system and ii) always stable according the bounded-input-bounded-output criterion. Even though models can represent many systems with fewer coefficients than their counterparts, it still possible...

This paper deals with methods for multiple position room response equalization. Differently from a well-known technique working in the time domain and based on fuzzy <i xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">c</i> -means clustering, proposed approach most of operations are performed frequency and, particular, clustering is applied to magnitude responses at different positions. It shown that allows us obtain equalization performances least...