Domain-Adversarial Learning for UWB NLOS Identification in Dynamic Obstacle Environments

In the rapidly evolving Internet of Things (IoT) landscape, accurate indoor positioning is increasingly vital. The proposed algorithm synergizes an ultrawideband (UWB) sensor with an inertial measurement unit (IMU) and artificial intelligence to obtain precise positioning in nonline-of-sight (NLOS) scenarios. In the proposed UWB module, a large language model (LLM) such as bidirectional encoder representations from transformers (BERT) algorithm is designed to utilize the channel impulse response (CIR) for effective NLOS identification and UWB ranging trustworthiness evaluation. Concurrently, the IMU module is also designed with BERT to recognize various pedestrian activity states, thereby optimizing positioning. BERT’s self-attention mechanism and deep learning (DL) bidirectional training efficiently extract essential features from sequential data, capturing both local and global information. The integration of both UWB and IMU through a proposed tightly coupled algorithm significantly boosts positioning performance. Experimental campaigns demonstrate an average NLOS identification accuracy, line-of-sight (LOS), and <inline-formula xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink"> <tex-math notation="LaTeX">${F}2$ </tex-math></inline-formula> of 98.8%, 99.4%, and 0.9926, respectively. These performances surpass the state-of-the-art least-squares support vector machine (LS-SVM), convolutional neural network (CNN), and CNN with long short-term memory (CNN-LSTM) up to 17.66% in NLOS identification. In terms of pedestrian activity recognition using BERT, the BERT algorithm achieves a precision (recall) of 99.3% (99.4%), notably outperforming the CNN and CNN-LSTM by 17.9% (16.2%) and 11.9% (10.9%), respectively. Finally, the UWB-IMU algorithm significantly enhances positioning accuracy by 80.5%, outperforming Kalman, LSTM-EKF, and particle filter (PF) methods by 68.1%, 48.3%, and 45.0%, respectively. The proposed approach presents a robust solution for indoor positioning for IoT applications, particularly in challenging NLOS environments.

In Ultra-Wide Band (UWB) positioning, wireless signals are subject to non-line-of-sight (NLOS) propagation due to obstruction by obstacles, which leads to ranging and positioning estimation errors. How to accurately and efficiently identify line-of-sight (LOS) and NLOS propagation paths is a key research task in UWB positioning systems. By effectively integrating the characteristics of global channel impulse response (CIR) sequence features and statistical time-domain features, a dual-branch feature fusion Transformer (DBFF-Transformer) is proposed for NLOS path identification. Firstly, the original CIR sequence data is processed using the Transformer to learn the global feature relationships within the data. Secondly, four key time-domain features are extracted from the CIR sequence: the first-path energy ratio, the root-mean-square time delay spread, the kurtosis and the phase difference. Finally, by integrating the sequence features and the time-domain features, the two features’ branches are fused through a fully connected network. The proposed method is evaluated in two typical indoor scenarios from the latest open-source datasets of the eWINE project. The ablation experiment proves that the fusion of the sequence features and time-domain features of the CIR sequence can effectively improve NLOS identification accuracy. The identification accuracy in the two experimental scenarios is 95.9% and 95.7%, with F1 scores of 97.2% and 97.1% and Recall of 97.4% and 96.4%, respectively. The comparative analysis of the DBFF-Transformer with the state-of-the-art baseline models demonstrates superior accuracy and robustness, which can provide a novel solution for NLOS identification in UWB indoor positioning.

Named Entity Recognition (NER) is a fundamental task in information extraction from unstructured text. Most previous machine-learning-based NER systems are domain-specific, which implies that they may only perform well on some specific domains (e.g., Newswire ) but tend to adapt poorly to other related but different domains (e.g., Weblog ). Recently, transfer learning techniques have been proposed to NER. However, most transfer learning approaches to NER are developed for binary classification, while NER is a multiclass classification problem in nature. Therefore, one has to first reduce the NER task to multiple binary classification tasks and solve them independently. In this article, we propose a new transfer learning method, named Transfer Joint Embedding (TJE), for cross-domain multiclass classification, which can fully exploit the relationships between classes (labels), and reduce domain difference in data distributions for transfer learning. More specifically, we aim to embed both labels (outputs) and high-dimensional features (inputs) from different domains (e.g., a source domain and a target domain) into a unified low-dimensional latent space, where 1) each label is represented by a prototype and the intrinsic relationships between labels can be measured by Euclidean distance; 2) the distance in data distributions between the source and target domains can be reduced; 3) the source domain labeled data are closer to their corresponding label-prototypes than others. After the latent space is learned, classification on the target domain data can be done with the simple nearest neighbor rule in the latent space. Furthermore, in order to scale up TJE, we propose an efficient algorithm based on stochastic gradient descent (SGD). Finally, we apply the proposed TJE method for NER across different domains on the ACE 2005 dataset, which is a benchmark in Natural Language Processing (NLP). Experimental results demonstrate the effectiveness of TJE and show that TJE can outperform state-of-the-art transfer learning approaches to NER.

The error compensation and suppression effects of traditional ultrawideband (UWB) ranging in non Line of Sight (NLOS) environments are limited. The contribution of specific occlusions to UWB spatial perception in NLOS is ignored. To achieve comprehensive sensing of spatial information by UWB, we initially analyze the channel impulse response (CIR) and the underlying parameters of registers during UWB communication. By comparing the difference between NLOS and line-of-sight (LOS) environments for each parameter on a continuous time series, a fast discriminative method for UWB environment conversion is proposed. Further, combining the ensemble learning XGBoost classifier, an efficient NLOS occlusion recognition method is proposed. At the same time, an algorithm optimization based on a discrete degree threshold is designed. It is based on loss function probability matrix-weighted predictive labeling. The prediction matrix of the loss function of the XGBoost algorithm is used as label weights. The UWB prediction labels of continuous time series are weighted, which mitigates the effect of low-probability data on the overall prediction results. Finally, UWB spatial sensing experiments are carried out to verify the reliability of the proposed method. The experimental results show that the mutation in the parameter profile can effectively perceive the LOS/NLOS transition. The recognition accuracy of the proposed occlusion recognition method in conditions of human, metal, and wall occlusion is 94.44%, 92.00%, and 95.87%, respectively. In contrast to the origin method, the suggested algorithm’s average recognition accuracy has increased by 16.71%. Its precise recognition accuracy makes UWB spatial sensing more effective.

With the booming demand of indoor position information, Wireless signals (Wi-Fi, Bluetooth, Ultra-Wide-Band etc.) are investigated to construct Indoor Positioning System (IPS). Among these wireless signals, UWB (Ultra-Wide-Band) is recognized as the most promising technology to construct IPS with decimeter-level positioning accuracy. However, there are various objects in indoor environments, UWB signals might be reflected by these surrounding objects. These None-Line-Of-Sight (NLOS) signals will induce additional errors to the distance measurements between the anchor and agent. Therefore, NLOS/LOS (Line-Of-Sight) signals classification should be carried out for identifying the NLOS reception. In UWB based IPS, the distance information is extracted through the Channel Impulse Response (CIR) waveforms. NLOS reception will lead to different CIR waveforms compared with that from the LOS signal, this difference could be characterized for NLOS/LOS signals classification. However, the hardware circuits, signal transmitting path and the NLOS reception will bring about noises to the CIR waveforms, which might influence the signal type identification. Motivated by this problem, in this letter, a reversible transformation method was proposed for de-nosing the CIR data, and a Conventional Neural Network (CNN) was employed to identify the NLOS signal. Raw and de-noised CIR dataset were both input to the CNN for assessing the effect of the de-noising method. With dataset collected from seven different sites, the results showed that the reversible transformation method was able to increase the Signal Noise Ratio (SNR) values of these CIR datasets. After the de-noising, the CNN performed higher classification accuracy with less convolutional layers. Moreover, the de-noising operation brought about higher NLOS signal identification accuracy increase than that of LOS signals.

Non-line-of-sight (NLOS) propagation poses a significant challenge to achieving high-accuracy ultra-wideband (UWB) indoor positioning. To address this issue, this study investigates solutions from two complementary perspectives: NLOS identification and error mitigation. First, an NLOS signal classification model is proposed based on multidimensional statistics of the channel impulse response (CIR). The model incorporates an attention mechanism and an improved snake optimization (ISO) algorithm, achieving significantly enhanced classification accuracy and robustness. Building on this foundation, a UKF–BiLSTM dual-directional mutual calibration framework is proposed to compensate for NLOS errors dynamically. The framework embeds the constant turn rate and velocity (CTRV) motion model within an unscented Kalman filter (UKF) to enhance trajectory modeling. It establishes a bidirectional correction loop with a bidirectional long short-term memory (BiLSTM) network. Through the synergy of physical constraints and data-driven learning, the framework adaptively suppresses NLOS errors. Experimental results demonstrate that the proposed classification model and positioning framework significantly outperform state-of-the-art methods, thereby providing a systematic solution for high-precision and robust UWB positioning in complex indoor environments.

This work proposes channel impulse response (CIR) prediction for time-varying ultra-wideband (UWB) channels by exploiting the fast movement of channel taps within delay bins. Considering the sparsity of UWB channels, we introduce a window-based CIR (WB-CIR) to approximate the high temporal resolutions of UWB channels. A recursive least square (RLS) algorithm is adopted to predict the time evolution of the WB-CIR. For predicting the future WB-CIR tap of window wk, three RLS filter coefficients are computed from the observed WB-CIRs of the left wk−1, the current wk and the right wk+1 windows. The filter coefficient with the lowest RLS error is used to predict the future WB-CIR tap. To evaluate our proposed prediction method, UWB CIRs are collected through measurement campaigns in outdoor environments considering line-of-sight (LOS) and non-line-of-sight (NLOS) scenarios. Under similar computational complexity, our proposed method provides an improvement in prediction errors of approximately 80% for LOS and 63% for NLOS scenarios compared with a conventional method.

The radio channel modeling and location system are the focus of many research fields. However, their performance is largely limited by line-of-sight (LOS) and non-line-of-sight (NLOS) conditions. Currently, the most popular LOS/NLOS identification algorithm is based on the channel impulse response (CIR) feature. The decision threshold is applied to identify between LOS and NLOS conditions on this basis. However, the features around the threshold are usually very similar, which leads to low LOS/NLOS accuracy. In this letter, the feature extracted from the CIR is used to fit its distribution with the Gaussian mixture function. Then, a soft decision algorithm is used to classify the signal. In the experimental simulations, the algorithm is tested according to the measured data. The results reveal that by choosing different feature combinations, the algorithm effectively improves the accuracy of LOS and NLOS identification.

Abstract. Ultra-wideband (UWB) technology stands out among numerous indoor positioning techniques due to its high operating frequency, low interception capability, resistance to multipath effects, and strong penetration. The UWB uses the time-of-arrival (TOA) to estimate the distance between the transmitter and receiver anchors in centimeter accuracy. However, in complex indoor positioning environments, obstacles such as walls, glass windows, metal plates, and wooden doors may block and reflect signals, inevitably causing non-line-of-sight (NLOS) errors that significantly affect positioning accuracy. The NLOS signal has lower signal energy due to the reflections. Thus, the channel impulse responses (CIR) from NLOS and LOS are different. To address the NLOS signal identification issue in UWB positioning, we utilize UWB CIR data collected from various positioning scenarios as the data source. CIR waveform input features are provided for the NLOS signal recognition model, and four machine learning models—Support Vector Machine (SVM), Multi-Layer Perceptron (MLP), K-Nearest Neighbors (KNN), and XGBoost—are trained and optimized for NLOS signal recognition. The aim of the study is to analyze the performance of different machine learning algorithms for NLOS signal recognition in UWB indoor localization using these features. Experimental results indicate that machine learning-based NLOS signal recognition algorithms can achieve an accuracy of approximately 77.46%, precision of 80.46%, and an F1 score of 0.81. Among the four models, the XGBoost model demonstrates generally better recognition performance.

The concept of Industry 4.0 uses cyber-physical systems and the Internet of Things to create "smart factories" that enable automated and connected production. However, the complex and diverse nature of cyber-physical systems makes them vulnerable to physical threats, where any fault in their physical components can lead to serious consequences, from financial losses to human safety. To reduce the risk of such failures, it is essential to have techniques in place to monitor the condition of these physical components closely. In recent years, data-driven predictive maintenance methods have become common due to the growth in data and advancements in machine learning. This thesis focuses on fault diagnosis, a key part of predictive maintenance, which aims to detect and identify faults in physical components. Incorporating advanced techniques, including machine learning algorithms and sensor data analysis, enhanced fault diagnosis systems' accuracy and efficiency. However, several real-world barriers challenge the effectiveness of these techniques. This research is structured around three primary objectives. The first objective aims to establish an efficient fault diagnosis system that is highly sensitive to fault occurrences while being resilient to uncertainties such as variability in fault patterns and data imbalance. To address variability in fault patterns, we developed a multi-step approach integrating clustering techniques and an adaptive neuro-fuzzy inference system, which significantly improved diagnosis robustness even under noisy conditions. For imbalanced datasets, a two-step method was introduced, combining permutation entropy for fault detection and envelope analysis for fault isolation. This approach was enhanced by an automatic frequency band selection method leveraging wavelet transform and statistical criteria, resulting in high fault diagnosis accuracy and reliability, even with limited labeled faulty data. The second objective focuses on developing fault diagnosis systems that can maintain high performance despite domain shifts and limited training data in both source and target domains. To address the challenge of domain shifts, a transfer learning technique was introduced, and a systematic literature review was conducted. This study bridges a technical gap by providing a comprehensive taxonomy, state-of-the-art methods, and insights into the challenges of applying transfer learning to predictive maintenance. Additionally, a conceptual framework was proposed to explore scenarios for transferring diagnostic knowledge from different source domains to target domains with insufficient training datasets. In a subsequent study, a new diagnostic framework was developed using digital twins and partial transfer learning to address the lack of real training datasets in both source and target domains as well as label space inconsistencies between them. The framework includes a general approach for developing digital twins of real machines to generate synthetic diagnostic data as source domain data. Furthermore, by incorporating a deep learning network and a double-layer attention mechanism, a partial transfer learning diagnostic model was developed to transfer diagnostic knowledge from the digital twin to the real target domain despite label space inconsistencies. The proposed framework was validated through an industrial case study involving rolling bearings in rotating equipment. The third objective focuses on developing an efficient fault diagnosis system capable of maintaining reliability under unseen and evolving working conditions. To achieve this, a novel self-adaptive fault diagnosis system was proposed, incorporating the MAPE-K loop, a domain generalization network model, and digital twin technologies. This self-adaptive fault diagnosis system leverages the MAPE-K loop to adapt in real time to evolving working conditions. Digital twins were employed to continuously generate data aligned with the real system’s operational context, enabling accurate fault diagnosis in dynamic environments. The domain generalization network model enhances generalization to unseen target working conditions by addressing overfitting and feature loss through multi-domain data augmentation, adversarial learning, and domain discrepancy modules. Validation using multiple rotating machinery datasets demonstrated that the developed fault diagnosis system outperformed state-of-the-art methods across various scenarios. In summary, this dissertation proposes a comprehensive and practical approach to addressing variability in fault patterns, imbalance datasets, data scarcity, domain shifts, and unseen and evolving working conditions challenges in existing fault diagnosis systems.

Non-line-of-sight (NLOS) propagation can severely degrade the reliability of communication and localisation accuracy in indoor ultra-wideband (UWB) ‘location-aware’ networks. Link adaptation and NLOS bias mitigation techniques have respectively been proposed to alleviate these effects, but implicitly rely on the ability to accurately distinguish between LOS and NLOS propagation scenarios. A statistical NLOS identification technique based on the hypothesis-testing of received signal parameters in UWB propagation channels is discussed. In contrast to narrowband and wideband signals, UWB signals possess higher temporal resolution and robustness to multipath fading. We show that these characteristics result in differences in the statistics of (a) the time-of-arrival (TOA), (b) the received signal strength (RSS) and (c) the root-mean-squared delay spread (RDS) of the received signals, between LOS and NLOS propagation scenarios, which can be exploited for accurate channel identification. We statistically characterise the ability of TOA, RSS and RDS estimates to distinguish between LOS and NLOS propagation based on an extensive indoor measurement campaign. Our measurement results suggest that the RDS of UWB signals can, even in isolation and without complete statistical information, serve as a robust and computationally efficient indicator of the LOS/NLOS nature of propagation. Finally, we demonstrate the efficacy of the discussed NLOS identification method in a location-tracking application based on indoor UWB measurements.

Breast cancer is widely prevalent in women. To diagnose it, pathologists evaluate breast histopathology images under a microscope in various magnifications. However, the number of pathologists per population is low in many countries, and a mistake is probable. Nowadays, deep learning has become a popular artificial intelligence research trends. Even though deep neural networks have obtained promising results in image processing, the need for extensive training data avoids its use in medical image processing. After the evaluation of various state-of-the-art deep learning methods and algorithms in medical data processing, this study proposes an effective deep transfer learning-based model, relying on pre-trained DCNN using an extensive data of ImageNet dataset which improves state-of the-art systems in both binary and multi-class classification. The weights of the pre-trained DesneNet121 on the Imagenet are transferred as initial weights first, and then the model is fine-tuned with a deep classifier along with data augmentation to distinguish various malignant and benign samples in the two categories of binary and multi-class classification. In the multi-class classification, the proposed model obtained up to 97% image-level accuracy. In binary classification, the model obtained up to 100% image-level accuracy. The achieved-results are outperforming previous studies accuracies in multiple performance metrics in breast cancer CAD systems. Moreover, the proposed method is both flexible and scalable, meaning it can be easily expanded to cover the detection of other types of diseases in the future and be integrated with more CNNs to increase its generalization capabilities.

A deep learning fusion network trained with clinical and high-frequency ultrasound images in the multi-classification of skin diseases in comparison with dermatologists: a prospective and multicenter study

As security threats become more complex, the need for effective intrusion detection systems (IDSs) has grown. Traditional machine learning methods are limited by the need for extensive feature engineering and data preprocessing. To overcome this, we propose two enhanced hybrid deep learning models, an autoencoder–convolutional neural network (Autoencoder–CNN) and a transformer–deep neural network (Transformer–DNN). The Autoencoder reshapes network traffic data, addressing class imbalance, and the CNN performs precise classification. The transformer component extracts contextual features, which the DNN uses for accurate classification. Our approach utilizes an enhanced hybrid adaptive synthetic sampling–synthetic minority oversampling technique (ADASYN-SMOTE) for binary classification and enhanced SMOTE for multi-class classification, along with edited nearest neighbors (ENN) for further class imbalance handling. The models were designed to minimize false positives and negatives, improve real-time detection, and identify zero-day attacks. Evaluations based on the CICIDS2017 dataset showed 99.90% accuracy for Autoencoder–CNN and 99.92% for Transformer–DNN in binary classification, and 99.95% and 99.96% in multi-class classification, respectively. On the NF-BoT-IoT-v2 dataset, the Autoencoder–CNN achieved 99.98% in binary classification and 97.95% in multi-class classification, while the Transformer–DNN reached 99.98% and 97.90%, respectively. These results demonstrate the superior performance of the proposed models compared with traditional methods for handling diverse network attacks.

The softmax and binary classifier are commonly preferred for image classification applications. However, as softmax is specifically designed for categorical classification, it assumes each image has just one class label. This limits its applicability for problems where the number of labels does not equal one, most notably zero- and multi-label problems. In these challenging settings, binary classifiers are, in theory, better suited. However, as they ignore the correlation between classes, they are not as accurate and scalable in practice. In this paper, we start from the observation that the only difference between binary and softmax classifiers is their normalization function. Specifically, while the binary classifier self-normalizes its score, the softmax classifier combines the scores from all classes before normalisation. On the basis of this observation we introduce a normalization function that is learnable, constant, and shared between classes and data points. By doing so, we arrive at a new type of binary classifier that we coin quasibinary classifier. We show in a variety of image classification settings, and on several datasets, that quasibinary classifiers are considerably better in classification settings where regular binary and softmax classifiers suffer, including zero-label and multi-label classification. What is more, we show that quasibinary classifiers yield well-calibrated probabilities allowing for direct and reliable comparisons, not only between classes but also between data points.