LMS Algorithm Research Articles

Design and Simulation of Adaptive Filters in Real-Time Environment

The results of the study showed that adaptive filters were designed and simulated to enhance signal processing capabilities in real-time environments. The study focused on the LMS and NLMS algorithms for echo and noise cancellation in audio signals, exploring the impact of different mu values (0.01, 0.05, 0.1, and 0.5) on algorithm performance. A comprehensive literature review underscored the importance of effective adaptive filtering techniques in real-time audio processing and highlighted the trade-off between convergence speed and stability. The results indicated that a mu value between 0.05 and 0.1 optimally balances convergence speed and stability for both LMS and NLMS algorithms. The NLMS algorithm consistently outperformed LMS due to its superior adaptation to signal power variations and better stability at higher mu values. Recommendations include using the NLMS algorithm for most applications, careful tuning of the mu value, and further research into adaptive and hybrid methods. These findings aimed to enhance echo and noise cancellation, resulting in clearer and more intelligible audio signals in real-world applications. Keywords: Adaptive Filters, LMS Algorithm, NLMS Algorithm, Real-Time Signal Processing

The Socio-Economic Situation of a Population as a Determinant of Childhood Obesity: Answers from Big Data

Big data tools are currently a major tool for assessing population changes. There is a causal relationship between low economic levels and a higher prevalence of conditions associated with malnutrition and obesity. One of the causes of low income in different areas could be a higher unemployment rate and a confluence of people who are migrating for economic reasons. Objective: To assess in a child population the effect of the unemployment rate, average income and immigration rate as a possible effect of increasing the prevalence of malnutrition associated with childhood obesity. Material and methods: Data collected from computerized clinical history episodes, studying the variables of sex, age, weight, height, of a pediatric population (year 2022), comparing it with the average income of their residential district, unemployment rate and immigration rate. Use of big data methods for the study of variables. Using the Cole-Green LMS algorithm with penalized likelihood, implemented in the RefCurv 0.4.2 software (2020), which allows managing large amounts of data. The hyperparameters have been selected using the BIC (Bayesian Information Criterion). To calculate population deviations from the reference, the reference was taken as being above 1.5 standard deviations from the average according to age. Results: 66,975 computerised episodes of children under 16 years of age and a total of 1,205,000 variables studied. The data and comparative graphs between districts of the population studied are represented with respect to the variables analysed. There are significant differences, with an increase in the rate of overweight in those areas with lower economic income and higher unemployment and immigration rates. Big data technology allows for more efficient population studies, selecting populations most in need of health intervention, optimizing scarce health resources.

Selective collaboration in distributed FxLMS active noise control systems

This paper addresses the selection of the collaboration configuration on distributed active noise control (ANC) systems. ANC systems aim to cancel out acoustic noise within a listening area. In distributed systems, the control task is delegated among multiple acoustic control nodes that generate the control signals by filtering a noise reference signal. The coefficients of each node filter are iteratively calculated using the filtered-X LMS algorithm. The stability is achieved when the adaptive filters computed in each node converge to finite values. However, acoustic coupling among nodes could lead to instability (i.e., divergence). Collaboration among selected nodes may avoid this phenomenon, although not just any collaboration configuration guarantees network stability. On the other hand, a collaborative distributed system presents two drawbacks: the stability assessment is computationally expensive, and communication requirements increase with the number of collaborations among nodes. In this paper, we propose and discuss several methods to establish a collaboration configuration that ensures system stability. The optimal configuration, which is characterized by the minimal number of necessary collaborations between nodes, can be identified through exhaustive search. However, this approach incurs a high computational cost, particularly in networks with many nodes. To address this challenge, we introduce several heuristic methods aimed at efficiently obtaining stable configurations.

SkipNet: an adaptive neural network equalization algorithm for future passive optical networking

In this paper, we propose an original adaptive neural network equalizer (NNE) algorithm named SkipNet, which is suitable for rapid training on a packet-by-packet basis for burst-mode non-linear equalization in upstream PON transmission. SkipNet uses the simple LMS algorithm and avoids complex neural network training algorithms such as backpropagation and mini-batch training. We demonstrate SkipNet on captured continuous mode 100 Gbit/s PAM4 signals using an SOA preamplifier to achieve the challenging 29 dB PON optical loss budget. The adaptive SkipNet equalizer is shown to overcome combinations of severe SOA patterning effects and fiber dispersion impairments to achieve >29dB dynamic range back-to-back and >22.9dB dynamic range for up to 81.6 ps/nm accumulated dispersion. It can adapt in as little as 250 training symbols to each impairment scenario, which is equivalent to existing FFE/DFE solutions, while matching the non-linear performance of previously proposed static NNE solutions. To the best of our knowledge, SkipNet is the first ever adaptive NNE framework that can realistically be trained and adapted on a packet-by-packet basis and within strict PON packet preamble lengths.

Bayesian-Learning-Based Diffusion Least Mean Square Algorithms Over Networks.

To improve the learning performance of the conventional diffusion least mean square (DLMS) algorithms, this article proposes Bayesian-learning-based DLMS (BL-DLMS) algorithms. First, the proposed BL-DLMS algorithms are inferred from a Gaussian state-space model-based Bayesian learning perspective. By performing Bayesian inference in the given Gaussian state-space model, a variable step-size and an estimation of the uncertainty of information of interest at each node are obtained for the proposed BL-DLMS algorithms. Next, a control method at each node is designed to improve the tracking performance of the proposed BL-DLMS algorithms in the sudden change scenario. Then, a lower bound on the variable step-size of each node of the proposed BL-DLMS algorithms is derived to maintain the optimal steady-state performance in the nonstationary scenario (unknown parameter vector of interest is time-varying). Afterward, the mean stability and the transient and steady-state mean square performance of the proposed BL-DLMS algorithms are analyzed in the nonstationary scenario. In addition, two Bayesian-learning-based diffusion bias-compensated LMS algorithms are proposed to handle the noisy inputs. Finally, the superior learning performance of the proposed learning algorithms is verified by numerical simulations, and the simulated results are in good agreement with the theoretical results.

Random Fourier features based nonlinear recurrent kernel normalized LMS algorithm with multiple feedbacks

The performance of kernel adaptive filtering algorithms (KAFs) with nonlinear recurrent structures surpasses traditional KAFs, attributed to the nonlinear contribution of feedback. Nevertheless, the existing nonlinear recurrent KAFs primarily focus on a single feedback output, potentially limiting their latent filtering capabilities. In this paper, we introduce a novel alternative, named nonlinear recurrent kernel normalized least-mean-square with multiple feedbacks (NR-KNLMS-MF), which leverages the information from multiple feedback outputs. Additionally, to tackle the computational complexity challenges associated with KAFs, we integrate random Fourier features (RFF) into NR-KNLMS-MF, resulting in an efficient variant called as RFF-NR-KNLMS-MF. Furthermore, we conduct a theoretical analysis of the mean-square convergence for RFF-NR-KNLMS-MF. Simulation results on time-series predictions demonstrate the superiority of our proposed algorithms over other competing alternatives, validating their effectiveness.

Secure multiple adaptive kernel diffusion LMS algorithm for distributed estimation over sensor networks

Secure multiple adaptive kernel diffusion LMS algorithm for distributed estimation over sensor networks

A hybrid fractional order LMS algorithm for power system harmonic estimation

This article proposes a novel multi-innovation fractional order least mean square algorithm for estimating harmonic parameters in power systems with unknown amplitudes and phases. The algorithm incorporates fractional order calculus with variable initial values, which, in contrast to the fixed initial value approach, overcomes the non-locality issue, and converges to its true extreme value. In addition, by extending scalar innovation term to a vector form, the algorithm fully utilizes innovation within a fixed window to improve computational accuracy. The convex combination between integer order and novel fractional order ensures that the performance of the algorithm will not change significantly with the fractional order. Finally, the effectiveness is verified by numerical simulations and comparisons.

Robust and fast backbone tracking via phase-locked loops

Phase-locked loops are commonly used for shaker-based backbone tracking of nonlinear structures. The state of the art is to tune the control parameters by trial and error. In the present work, an approach is proposed to make backbone tracking much more robust and faster. A simple PI controller is proposed, and closed-form expressions for the gains are provided that lead to an optimal settling of the phase transient. The required input parameters are obtained from a conventional shaker-based linear modal test, and an open-loop sine test at a single frequency and level. For phase detection, an adaptive filter based on the LMS algorithm is used, which is shown to be superior to the synchronous demodulation commonly used. Once the phase has locked, one can directly take the next step along the backbone, eliminating the hold times. The latter are currently used for recording the steady state, and to estimate Fourier coefficients in the post-process, which becomes unnecessary since the adaptive filter yields a highly accurate estimation at runtime. The excellent performance of the proposed approach is demonstrated for a doubly clamped beam undergoing bending–stretching coupling leading to a 20 percent shift of the lowest modal frequency. Even for fixed control parameters, designed for the linear regime, only about 100 periods are needed per backbone point, also in the nonlinear regime. This is much faster than what has been reported in the literature so far.

Diffusion LMS algorithm in the presence of second order nonlinearities with theoretical bounds

In this paper, we propose an algorithm for the estimation and compensation of second-order nonlinearity in Wireless Sensor Networks (WSNs) within a distributed estimation framework. First, we investigate the impact of second-order nonlinearity on the performance of the Diffusion Least Mean Square (DLMS) algorithm and derive an upper bound for the l2-norm of the error caused by nonlinearity. Next, we suggest a distributed algorithm incorporating additional nonlinearity estimation and compensation units. Furthermore, considering second-order nonlinearity, we calculate the Cramer-Rao Bound for estimating both the unknown vector and the nonlinearity coefficient vector, with the Fisher information matrix derived in a closed-form formula. Simulation results demonstrate the effectiveness of the proposed algorithm in enhancing the performance of distributed estimation in the presence of nonlinear sensors in a WSN.

Differential privacy for diffusion LMS algorithm over a distributed network

This article presents a way to improve the Diffusion LMS algorithm in practical scenarios while maintaining data privacy. It focuses on reducing external errors and ensuring accurate estimation of the parameter vector through distributed and adaptive privacy protection. By combining these improvements, the overall performance of the algorithm can be significantly enhanced without compromising Mean Square Deviation (MSD). One important aspect is tuning the privacy noise power based on data sensitivity. This work provides a novel solution for improving the performance and privacy of the Diffusion LMS algorithm with implications for future research. The mean square error and evolution behavior of the algorithms are also analyzed, showing their effectiveness and robustness in preserving privacy and MSD over the network.

IMPROVING 5G NETWORK PERFORMANCE WITH MIMO TECHNOLOGY USING BEAMFORMING ALGORITHM

The advent of 5G technology aims to revolutionize wireless communication by providing significantly higher data rates, ultra- reliable low latency, and enhanced connectivity. However, optimizing 5G network performance remains a challenging task, particularly in terms of Bit Error Rate (BER) and spectral efficiency. Multiple-Input Multiple-Output (MIMO) technology, coupled with beamforming algorithms, offers a promising solution to these challenges. This study investigates the application of the Fractional Least Mean Square (FLMS) algorithm for beamforming in MIMO systems to improve BER and spectral efficiency in 5G networks. The primary problem addressed is the need for effective interference mitigation and signal enhancement in densely populated environments, which traditional methods struggle to handle. The proposed method utilizes the FLMS algorithm to dynamically adjust the beamforming weights, thereby optimizing the signal reception quality. The algorithm's fractional nature allows for finer adjustments and better adaptation to varying channel conditions compared to conventional LMS algorithms. Simulation results show the efficacy of the proposed method. Specifically, the implementation of the FLMS algorithm in a 4x4 MIMO system shows a reduction in BER by 30% and an improvement in spectral efficiency by 25% compared to traditional beamforming techniques. These numerical values highlight the potential of FLMS in enhancing 5G network performance, making it a viable approach for future wireless communication systems.

Transient Analysis of a Selective Partial-Update LMS Algorithm

In applications where large-order filters are needed, the computational load of adaptive filtering algorithms can become prohibitively expensive. In this paper, a comprehensive analysis of a selective partial-update least mean squares, named SPU-LMS-M-min, is developed. By employing the partial-update strategy for a non-normalized adaptive scheme, the designer can choose an appropriate number of update blocks considering a trade-off between convergence rate and computational complexity, which can result in a more than 40% reduction in the number of multiplications in some configurations compared to the traditional LMS algorithm. Based on the principle of minimum distortion, a selection criterion is proposed that is based on the input signal’s blocks with the lowest energy, whereas typical Selective Partial Update (SPU) algorithms use a selection criterion based on blocks with highest energy. Stochastic models are developed for the mean weights and mean and mean squared behaviour of the proposed algorithm, which are further extended to accommodate scenarios involving time-varying dynamics and suboptimal filter lengths. Simulation results show that the theoretical predictions are in good agreement with the experimental outcomes. Furthermore, it is demonstrated that the proposed selection criterion can be easily extended to active noise cancellation algorithms as well as algorithms utilizing variable filter length. This allows for the reduction of computational costs for these algorithms without compromising their asymptotic performance.

Harmonic Vibration Suppression of MSCSG Based on Normalized-Variable-Step- Size LMS Algorithm

Harmonic Vibration Suppression of MSCSG Based on Normalized-Variable-Step- Size LMS Algorithm

A high-pass shaped LMS algorithm based predistortion technique for fractional-[formula omitted] BB-DPLLs

In this paper, we prove that rather than the second-order ΔΣ modulator (DSM) as typically believed using the first-order one yields a faster convergence for the linear-piecewise predistortion technique employed in digital/time converter (DTC) based fractional-N Bang-Bang digital phase-locked-loops (BB-DPLLs). We also propose a novel technique that addresses the limit-cycle issue happening in near-integer channels of the conventional BB-DPLLs. With the first-order DSM-based novel technique, simulation results show that the worst fractional spurs are below −83 dBc at 12 kHz offset frequency assuming the DTC’s integral nonlinearity error is as large as 5 least-significant-bits.

Echo cancellation using Artificial Intelligence technique

To improve the audio quality in telecommunication, echo cancellation was proposed to prevent signals from re-appearing. Generally speaking, the echo canceler first recognizes the original transmitted signal that appears again due to delay in the transmitted or recieved signal. Then, the echo canceler subtracts the echo in the signals. Commonly, echo cancelers are calledacoustic echo cancellation (AEC) in general, whether they aim for acoustic echo, line echo, or both. However, in modern times, new problems have been emerging which require more innovative, more advanced methods to tackle the echo issue. Thus, the paper reviews methods regarding Artificial Intelligence (AI) to improve the current means for echo cancellation. In this paper, we mainly discuss the approaches to echo cancellation, the identification of echo using adaptive AI and the usage of adaptive filters and neural network for echo cancellation. For instance, deep learning approaches can facilitate the issues of AEC (active echo control) and ANC (active noise control),the optimization of LMS algorithm can be achieved by combining Swarm Intelligence (SI), combining adaptive digital filter and recurrent neural network for acoustic echo cancellation etc. Using AI in echo cancellation can help us achieve high-quality, full-duplex telecommunication in modern telephony system.

Analysis of a Diffusion LMS Algorithm with Probing Delays for Cyclostationary White Gaussian and Non-Gaussian Inputs

Analysis of a Diffusion LMS Algorithm with Probing Delays for Cyclostationary White Gaussian and Non-Gaussian Inputs

Stability analysis of the bias compensated LMS algorithm

The presence of noise is a pervasive issue that significantly impacts the performance of adaptive filtering algorithms. To address this challenge, the bias compensation technique has recently emerged, involving the incorporation of an additional term into the update equation. Despite the apparent simplicity of the bias-compensated least-mean-squares algorithm, conducting a comprehensive theoretical analysis of its performance is a complex task. In this paper, we undertake an exact expectation analysis to demonstrate the asymptotic unbiasedness of the algorithm, even when disregarding the commonly assumed theory of independence between adaptive coefficients and input data. Furthermore, to improve the understanding of the algorithm's stability, we employ a stochastic model that assumes independence between the radial and angular distributions of the input vector. The resulting model is intricate, necessitating heuristic approximations to derive practical insights. Notably, our analysis reveals that the upper bound on the step size of the bias-compensated least-mean-squares algorithm consistently remains smaller than that of the least-mean-squares algorithm. These findings gain robust support from extensive simulations.

Insights into the Filtered-x LMS Algorithm in the Presence of Frequency Mismatch

Insights into the Filtered-x LMS Algorithm in the Presence of Frequency Mismatch

Speech Network Analysis and Spoken English Teaching for 6G Wireless Transmission Technology

Abstract Recent years have seen intensive research and widespread application of 6G wireless transmission technology, providing a technical bridge for the application of speech network analysis to spoken English teaching. Thus, this paper suggests a method for improving 6G technology. This paper primarily focuses on the following areas: First, to prevent overfitting, the dropout technique under 6G is added to the model, and the speech feature extracted from the data by the coding network is directly utilized. We then utilize the feature for either reconstruction or classification purposes. Initially, we perform unsupervised reconstruction using unlabeled data, and once we obtain the network parameters, we use the labeled data for classification. The next step involves feeding both data simultaneously into the model to calculate the error. Finally, we use gradient descent to minimize the error and optimize the parameters until the model converges. This model can convert sound-related symbols or information into visual symbols, and it can also convert speech into text. Comparative analysis shows that the improved method in this paper is far better than other methods in terms of noise reduction, and the signal-to-noise ratio of this paper is 3.76 dB higher than that of the LMS algorithm and 1.36 dB higher than that of the spectral subtraction method, which can stimulate the students’ interest in learning spoken English and increase the student’s interest in learning by 13.64% as a whole. It can be summarized through the comparison that this paper’s improved method increases the clarity of speech network analysis, which helps students teach spoken English.