Recursive inverse adaptive filtering algorithm

Abstract

In this paper, a new FIR adaptive filtering algorithm is introduced. This algorithm is based on the Quasi-Newton (QN) optimization algorithm. The approach uses a variable step-size in the coefficient update equation that leads to an improved performance. The simulation results show that the algorithm has very similar performance to the Robust Recursive Least Squares Algorithm (RRLS) while performing better than the Transform Domain LMS with Variable Step-Size (TDVSS) in stationary environments. The algorithm is tested in Additive White Gaussian Noise (AWGN) and Correlated Noise environments.

Robust Recursive Inverse Adaptive Algorithm in Impulsive Noise

“…where w(k) is the tap-weight vector of length N , I is an N × N identity matrix, R(k) is an N × N estimate of the tap-input vector autocorrelation matrix, p(k) is the estimate of the cross-correlation vector between the desired output signal and the tap-input vector of length N , and μ(k) is the variable step-size given by [2] …”

“…However, the RLS algorithm suffers from its high computational complexity, stability problems when the forgetting factor is relatively low, and tracking capability when the forgetting factor is relatively high. The RI algorithm [1,2], was recently proposed to overcome these problems. It is also known that the RLS and RI algorithms both provide a poor performance in impulsive-noise environments when the SNR is low.…”

“…In this paper, we propose a new RI algorithm that is robust to impulsive noise and provides better MSE performance than the recently proposed Robust RLS algorithm [4], with a considerable reduction in the computational complexity [1,2] as shown in Table 1. The proposed method employs the technique introduced in [4].…”

Joint Estimation of States and Parameters for an Input Nonlinear State-Space System with Colored Noise Using the Filtering Technique

“…The filtering technique has been applied to many fields, such as adaptive control [25,32], fault detection [1,48] and system identification [4,11,16]. Wang proposed the least squares-based recursive and iterative algorithms for output error systems by combining the auxiliary model idea with the data filtering [39].…”

Data Filtering-Based Multi-innovation Stochastic Gradient Algorithm for Nonlinear Output Error Autoregressive Systems

“…The data filtering technique has been used in the system identification [1,24,30]. Recently, by using the data filtering technique, Wang and Tang [32] studied several gradient-based iterative estimation algorithms for a class of nonlinear systems; Chen et al [4] derived a data filtering-based least squares iterative identification algorithm for output error autoregressive systems; they use the noise transfer function to filter inputoutput data, resulting in a noise model and a filtered model, and then to derive new methods for identifying the parameters of these two models, respectively.…”