Search

Parameter estimation is important in signal modeling and signal filtering. This paper uses the filtering technique to study the parameter estimation problems of a class of bilinear systems, for which the input-output representation is derived by eliminating the state variables in the systems. According to the obtained identification model, a two-stage least squares based iterative algorithm and a filtering based least squares iterative algorithm are proposed for estimating the parameters of bilinear systems with colored noises by using the hierarchical identification principle and the data filtering technique. The least squares based iterative algorithm is given for comparison. The simulation results show that the proposed algorithms have good performance in estimating the parameters of bilinear systems. [ABSTRACT FROM AUTHOR]

Augmented adaptive filters provide superior performance over their conventional counterparts when dealing with noncircular complex signals. However, the performance of such filters may change considerably when they are implemented in finite-precision arithmetic due to round-off errors. In this paper, we study the performance of recently introduced augmented complex least mean-square (ACLMS) algorithm when it is implemented in finite-precision arithmetic. To this aim, we first derive a model for the finite-precision ACLMS updating equations. Then, using the established energy conservation argument, we derive a closed-form expression, in terms of the excess mean-square error (EMSE) metric which explains how the quantized ACLMS (QACLMS) algorithm performs in the steady-state. We further derive the required conditions for mean stability of the QACLMS algorithm. The derived expression, supported by simulations, reveals that unlike the infinite-precision case, the EMSE curve for QACLMS is not monotonically increasing function of the step-size parameter. We use this observation to optimize the step-size learning parameter. Simulation results illustrate the theoretical findings. [ABSTRACT FROM AUTHOR]

This paper introduces a novel approach to derive norm-constrained adaptive algorithms for sparse system identification. In contrast to other similar approaches found in the literature, the proposed approach is focused primarily on keeping the a posteriori error equal to zero (which is a characteristic of the normalized least-mean-square algorithm) while seeking to satisfy a norm constraint. To this end, the proposed algorithms look directly for a vector belonging to the intersection of a zero-error hyperplane and a hyperplane resulting from a relaxed norm constraint. This somewhat simpler strategy leads to effective sparsity-promoting adaptive algorithms that exhibit low computational complexity and use parameters that are easy to adjust. In this context, a general framework that allows obtaining adaptive algorithms using different norm functions is devised. From this framework, two norm-constrained algorithms based on the ℓ 1 and ℓ 0 norms are obtained. Moreover, enhanced versions of these algorithms are developed aiming to make them independent of user-defined norm-bound parameters. Numerical simulation results corroborate the effectiveness of the proposed framework as well as the very good performance of the obtained algorithms. [ABSTRACT FROM AUTHOR]

Abstract: The zero attracting normalized least mean square (ZA-NLMS) algorithm achieves lower steady-state error than the normalized least mean square (NLMS) algorithm for sparse system identification. Most of the available analytical results on several versions of the zero attracting least mean square algorithms assume white Gaussian inputs. This paper presents the individual weight error variance (IWV) analysis of the ZA-NLMS algorithm without Gaussian inputs assumption. The IWV analysis is based on exact individual weight error relation and used to derive the transient and steady-state behavior of the ZA-NLMS algorithm without restricting the input to being Gaussian or white, whereas some assumptions are introduced to overcome weight nonlinearity in evaluating certain expectations involved. Extensive simulations are used to verify the analysis results presented. [Copyright &y& Elsevier]

Abstract: According to the iterative identification technique and the hierarchical identification principle, this paper presents a two-stage gradient based and a least squares based iterative parameter estimation algorithms (i.e., the hierarchical gradient based iterative algorithm and the hierarchical least squares based iterative algorithm) for controlled autoregressive autoregressive moving average systems. The proposed two-stage least squares based iterative algorithm requires less computation compared with the least squares based iterative algorithm. The simulation results indicate that the two-stage least squares based iterative algorithm converges faster than the two-stage gradient based iterative algorithm. [Copyright &y& Elsevier]

Abstract: Parameter estimation methods have wide applications in signal processing, communication and system identification. This paper derives an iterative least squares algorithm to estimate the parameters of output error systems and uses the partitioned matrix inversion lemma to implement the proposed algorithm in order to enhance computational efficiencies. The simulation results show that the proposed algorithm works well. [Copyright &y& Elsevier]

Abstract: The zero-attracting LMS (ZA-LMS) algorithm is one of the recently published sparse LMS algorithms. It usesan l 1-norm penalty in the standard LMS cost function. In this paper, we perform convergence analysis of the ZA-LMS algorithm based on white input signals. The stability condition is examined and the steady-state mean square deviation (MSD) is derived in terms of the system sparsity, system response length, and filter parameters (step size and zero-attractor controller). In addition, we propose a criterion for parameter selection such that the ZA-LMS algorithm outperforms the standard LMS algorithm. The results are demonstrated through computer simulations. [Copyright &y& Elsevier]

Abstract: This paper proposes a new variable-step-size control for the normalized least-mean-square (NLMS) algorithm. A step-size vector is used, with a different value for each adaptive weight. With prior knowledge of the channel impulse response statistics, the optimal step-size vector is obtained by minimizing the mean-square deviation (MSD) between the optimal and estimated filter coefficients. In addition, filter convergence is proved and the relationship between the proposed and existing algorithms are analyzed. The proposed method achieves better steady-state performance compared to existing algorithms. The effectiveness of the proposed algorithm is demonstrated through computer simulations. [Copyright &y& Elsevier]

Abstract: In actual nonlinear active noise control (NANC) systems, there often exist nonlinear distortions in such cases: the primary path may be nonlinear, the reference noise may exhibit nonlinear distortion, and the secondary path may have nonminimum-phase. To solve the problems of nonlinear distortions, two novel feedback adaptive filters based on the functional link neural network (FLNN) for NANC systems with low computational complexity are proposed in this paper, which are a feedback functional link neural network (FFLNN) and a reduced feedback functional link neural network (RFFLNN), respectively. To train the proposed nonlinear filters for NANC systems, a reduced complexity filtered-s least mean square (FSLMS) algorithm using filter bank approach is developed. The analysis of computational complexity shows that the RFFLNN adaptive filter involves less computation as compared to FFLNN and FLNN adaptive filters. Moreover, it is demonstrated through computer simulations for nonlinear noise processes that the RFFLNN adaptive filter outperforms FLNN and FFLNN in term of convergence speed and steady-state error. Furthermore, it is more effective in reducing nonlinear effects in NANC systems than other filters. [Copyright &y& Elsevier]

Abstract: In systems, such as hands-free mobile telephony, voice over internet protocol (VoIP), voice control, hearing aids, public address and teleconferencing, we are still faced with echo problem. Adaptive digital filters have been vastly used for the echo canceling by estimating the reverberation room impulse response. When a strength near-end talker (NET) signal exists, the adaptation process of the conventional algorithms misleads. In this paper, based on the correlation functions, we introduce a new algorithm named as adaptive step-size normalized correlation-based least mean square (ASNCLMS) algorithm, which is robust even in the presence of NET signal. This algorithm does not freeze the adaptation process during double-talk (DT) mode. Computer simulations, which prove the robustness of this algorithm, are also presented. [Copyright &y& Elsevier]

The linear least-squares algorithms for emitter localization such as the pseudolinear estimator are known to exhibit large estimation bias because of the correlation between the measurement matrix and the bearing noise. The paper presents a new bearings-only emitter localization algorithm with a closed-form solution that overcomes this bias by making use of instrumental variables obtained from a biased location estimate. By way of computer simulations, the new algorithm is shown to outperform the pseudolinear estimator while having a performance identical to that of the computationally more expensive maximum likelihood estimator. [Copyright &y& Elsevier]