Feature Fusion Using Stacked Denoising Auto-Encoder and GBDT for Wi-Fi Fingerprint-Based Indoor Positioning

Indoor localization is a research field that has drawn the attention of many researchers over the last years. In some application scenarios, for example when assisting visually impaired persons with finding their way around within indoor environments, the aspect of dynamic obstacle detection plays an important role. In this work, an optimized localization algorithm based on field strength measurements is presented. Furthermore, methods for dynamic obstacle detection are developed and evaluated. For this purpose, a low-cost infrared array sensor is used. Since most of the encountered dynamic obstacles in indoor environments are generally persons, the major advantage of the proposed concept is that the anonymity of the detected persons is preserved and there are no conflicts with data protection regulations. It will be shown that reliable and accurate indoor localization based on field strength measurements within a wireless mesh network can be carried out. Also the detection of moving persons is possible and even the movement patterns of two persons simultaneously crossing the field of view of the sensor can be displayed. The achieved results are verified by extensive measurements and descriptive statistics.

Distribution network topology identification based on gradient boosting decision tree and attribute weighted naive Bayes

Evaluating athletes' sports skills is significant for optimizing their training strategies and improving competitive outcomes. Although various studies are done to assess athletes' skills, they face challenges because of their limitations in capturing the complex, intricate patterns in athletes' movements. Hence, we proposed an innovative approach utilizing trajectory analysis and machine learning algorithms to evaluate and optimize athletes' sports skills. Initially, the trajectory data is collected using advanced motion tracking technology during either training or competitive sessions of the athletes. The collected data was pre-processed, and then feature extraction was performed to extract the most important and relevant features within the data. Further, the Gradient Boosting Decision Tree (GBDT) was trained using the extracted features to predict the athletes' performances. The GBDT is a powerful machine learning algorithm that can handle complex, non-linear interconnections among variables, enabling it to accurately predict the athletes' performances. Finally, we applied the Whale Optimization Algorithm (WOA) to refine the training process of the GBDT, enabling precise training and accurate predictions. The proposed methodology was implemented in the MATLAB software, and it is validated using the real-world athlete data collected during training sessions. The experimental results are validated in terms of parameters such as accuracy, precision, recall, and f-measure. Furthermore, we made a comparative study with the existing methods to validate the effectiveness and robustness of the proposed technique.

People generally find themselves lost while visiting a new indoor location because they are unaware of the building’s architecture, especially when it is a large one or a shopping mall. The Global Position System (GPS) does not work properly in the indoor environment because of the satellite signal attenuation. In this paper, to assist people in finding their path, a Li-Fi (Light-Fidelity) based Indoor Positioning System (IPS) is proposed. A framework is developed based on a Li-Fi LED lamp transmitter and a dongle receiver connected to an Android smartphone to decode the received sequence. The pathfinding graph-based algorithm is proceeded in a REST architecture by a Web service consulting the graph-path database both installed on a Raspberry pi 4. The proposed solution is a low cost and does not require any additional infrastructure. It is easy to implement in most indoor environments like hospitals, buildings, and campuses. A short survey of techniques and algorithms for indoor positioning and navigation with the help of smartphones is also presented.

In this paper, an algorithm based on Bi-directional Gated Recurrent Unit (Bi-GRU) and Gradient Boosting Decision Tree (GBDT) is proposed to extract the entity relationship of Traditional Chinese Medicine influenza. Firstly, the word vector is used as the input data set and the word vector is constructed by the word embedding model Word2Vec tool. Then the sentence feature is extracted by Bi-GRU and the attention mechanism is integrated to improve the accuracy of feature extraction. Finally, the feature vector is input into the GBDT algorithm for classification training and prediction to complete the Traditional Chinese Medicine influenza entity relationship extraction. In this paper, a variety of different entity relation extraction algorithms are compared with this algorithm to verify the effectiveness of the algorithm. This algorithm improves the stability of the model and effectively solves the problem of insufficient generalization ability of the model. Therefore, when studying the relationship extraction of Traditional Chinese Medicine texts, we can give priority to using Bi-GRU+GBDT model. Also through the experiment to adjust the model parameters and comparison to get the optimal parameters of Traditional Chinese Medicine influenza relationship extraction experiment.

This dissertation explores the intersection of data fusion, federated learning, and Bayesian methods, with a focus on their applications in indoor localization, GNSS, and image processing. Data fusion involves integrating data and knowledge from multiple sources. It becomes essential when data is only available in a distributed fashion or when different sensors are used to infer a quantity of interest. Data fusion typically includes raw data fusion, feature fusion, and decision fusion. In this thesis, we will concentrate on feature fusion. Distributed data fusion involves merging sensor data from different sources to estimate an unknown process. Bayesian framework is often used because it can provide an optimal and explainable feature by preserving the full distribution of the unknown given the data, called posterior, over the estimated process at each agent. This allows for easy and recursive merging of sensor data with prior knowledge without the need for storage. However, there are challenges in data fusion, such as the multiple counting problem, which occurs when data is utilized numerous times without the user's knowledge. In a Bayesian setting, a priori information of the unknown quantities is available and can be present among different distributed estimators. When the local estimates are fused, the prior knowledge used to construct several local posteriors might be overused unless the fusion node accounts for this and corrects it. In this thesis, we analyze the effects of shared priors in Bayesian data fusion contexts. Our analysis helps to understand the performance behavior as a function of the number of collaborative agents and different types of priors, depending on different common fusion rules. To perform this analysis, we use two divergences that are common in Bayesian inference. The generality of the results allows us to analyze very generic distributions. These theoretical results are corroborated through experiments in a variety of estimation and classification problems, including linear and nonlinear models, and federated learning schemes in a Bayesian perspective. Federated Learning is a discipline of Distributed Data Fusion that can be used for model training on local data without exchanging the data itself. This technique is useful in scenarios like image classification, indoor positioning, and jammer signal classification, where privacy concerns make data crowdsourcing difficult. In this thesis, we explore FL in two specific applications: Jammer signal and indoor positioning. Jamming signals can disrupt the operation of GNSS receivers, making jamming mitigation and localization techniques essential. Jammer classification can help with these techniques. For indoor positioning, local data points are collected to create a local fingerprinting database. Data-driven machine learning methods are used to predict the location based on this model learned without data sharing. However, Federated Learning faces several challenges, including the non-IID (non-identically distributed) problem. This problem occurs when the data distribution varies across different clients, which can lead to a decrease in performance. In this thesis, we will discuss various FL frameworks and algorithms that can help address this problem. A FL method from Bayesian perspective are proposed to deal with the problem in indoor positioning problem. Additionally, we will delve into the concepts of personalized and clustered Federated Learning. The personalized approach tailors the learning process to individual clients, which enhances the system's performance by leveraging the unique data and characteristics of each client. We will discuss the application of this approach in indoor positioning and jammer signal classification, highlighting its potential improvements and challenges. Expanding on the personalized approach, we introduce and discuss clustered Federated Learning. This learning method groups clients with similar characteristics or data patterns. Each client is typically associated with one data distribution and participates in training a model for that distribution along with their cluster peers. We propose a Bayesian framework for clustered FL, which associates clients to clusters, along with several practical algorithms for efficient learning, considering the trade-offs between performance and computational complexity. In conclusion, this thesis provides a detailed investigation and evaluation of the integration of data fusion, federated learning, and Bayesian methods. The insights and conversations presented in this study are anticipated to enhance the progress and implementation of these technologies in various domains, particularly in indoor positioning, and GNSS.--Author's abstract

In this study, an electronic nose (E-nose) consisting of seven metal oxide semiconductor sensors is developed to identify milk sources (dairy farms) and to estimate the content of milk fat and protein which are the indicators of milk quality. The developed E-nose is a low cost and non-destructive device. For milk source identification, the features based on milk odor features from E-nose, composition features (Dairy Herd Improvement, DHI analytical data) from DHI analysis and fusion features are analyzed by principal component analysis (PCA) and linear discriminant analysis (LDA) for dimension reduction and then three machine learning algorithms, logistic regression (LR), support vector machine (SVM), and random forest (RF), are used to construct the classification model of milk source (dairy farm) identification. The results show that the SVM model based on the fusion features after LDA has the best performance with the accuracy of 95%. Estimation model of the content of milk fat and protein from E-nose features using gradient boosting decision tree (GBDT), extreme gradient boosting (XGBoost), and random forest (RF) are constructed. The results show that the RF models give the best performance (R2 = 0.9399 for milk fat; R2 = 0.9301 for milk protein) and indicate that the proposed method in this study can improve the estimation accuracy of milk fat and protein, which provides a technical basis for predicting the quality of milk.

Due to the limitation of Global Navigation Satellite System (GNSS), it cannot work properly without direct line of sight to GNSS satellites such as in urban and indoor environments. Visual processing such as Visual Odometry (VO) and Structure-From-Motion (SFM) is usually used for the indoor navigation or localization. SFM was introduced in 1990 and had been successfully used for reconstruction of unordered images, and in the process of reconstruction, the SFM would estimate the position of the camera for each image at the same time. This paper introduces three most common SFM tools and suggests a positioning method for small UAVs equipped with vision cameras in indoor environment. Each stage of the SFM algorithms is studied to analysis the impact on the final positioning accuracy. The output from different feature detection, feature matching, initial pair setting, and the reconstructions are compared with an indoor reference localization system based on infrared cameras. The objective of this paper is to suggest a positioning and navigation method for small UAVs equipped with vision cameras in indoor environment. In this research, we will study the impact of different stages of SFM algorithm on the final positioning accuracy. Several objects such as chairs or boxes would be placed randomly in a lab room to build a feature-based environment without any special landmarks. In our experiment, an indoor localization system based on infrared cameras will be used as the reference navigation result to evaluate the final performance of the suggested algorithm.

Error compensation in indoor positioning systems based on software defined visible light communication

This paper describes the feasibility of selecting fractal feature vector based on M-band wavelet transform to classify ultrasonic liver images-normal liver, cirrhosis, and hepatoma. The proposed feature extraction algorithm is based on the spatial-frequency decomposition and fractal geometry. Various classification algorithms based on respective texture measurements and filter banks are presented and tested. Classifications for the three sets of ultrasonic liver images reveal that the fractal feature vector based on M-band wavelet transform is trustworthy. A hierarchical classifier, which is based on the proposed feature extraction algorithm is at least 96.7% accurate in the distinction between normal and abnormal liver images and is at least 93.6% accurate in the distinction between cirrhosis and hepatoma liver images. Additionally, the criterion for feature selection is specified and employed for performance comparisons herein.

In the world of aquaculture, understanding the condition of a pond is very important for a farmer in deciding which action should they take to prevent any bad condition occurred. Condition of a pond can be justified by measuring plenty of water parameters which can be divided into 3 categories that are physical, chemical and biological. The physical parameter is any physical quantity that can be measured in the pond. The chemical parameter is any kind of chemical substances that are dissolved in water. The biological parameter is any organic matter that lives in water. However, all of these parameters are not so distinguishable in representing the condition of a pond. Therefore, the farmer experience difficulties in justifying the condition and taking proper action to their pond. Even with the help of the K-Nearest Neighbors (KNN) algorithm combined with grid search optimization to model the data, the result is still not satisfying where the model only achieve accuracy of 0.701 in leave one out validation. To overcome this problem, a kind of feature extraction algorithm is needed to extract more information and make the data become more differentiate in representing the condition of the pond. With the help of our proposed feature extraction algorithm, optimized KNN can model the data easier and achieve higher accuracy. From the experiment results, the proposed feature extraction algorithm gives an impressive performance where it increases the accuracy to 0.741. A comparison with other feature extraction algorithms such as Principal Component Analysis (PCA), Non-negative Matrix Factorization (NMF), and Singular Value Decomposition (SVD) is also conducted to validate how good the proposed feature extraction algorithm is. As a result, the proposed algorithm is surpassing the other algorithms which only achieve the accuracy of 0.707, 0.718, and 0.718, respectively.

Since the well-established navigation systems such as GPS are ineffective in indoor environments, research into developing novel indoor positioning technologies has emerged in recent years. While several technologies are being investigated, a practical and reliable indoor positioning system is yet to emerge. Indoor positioning using light signals holds a great potential to provide a reliable solution to the indoor positioning problem. This paper presents a comprehensive review of the available literature on light-based indoor positioning systems. The paper establishes a classification based on the design of various light-based indoor positioning (LIP) systems, and discusses their key components including hardware, software and algorithms. The limitations and challenges of the LIP systems are discussed in details, and current published solutions are presented. The review also provides a comparison between LIP systems and other optical positioning systems, and identifies the associated research gaps. Finally, based on the identified gaps, the future research directions in both LIP systems and other optical positioning systems are presented.

<p>The interest in indoor positioning systems has been rising in the last couple of decades. This interest was further accelerated by the availability of signals of opportunity. The introduction of signals of opportunity into the indoor environments has meant that these widely available technologies can be utilised to achieve higher levels of accuracy and precision for positioning in parking garages, airports, underground locations, large office buildings, and shopping centres. Indoor positioning can also be utilised for emergency services or for Building Information Models (BIM) for purposes such as construction sites or hospital building management. In addition to the need for indoor and outdoor positioning, there is a need to ensure seamless transitions in and out of these different environments. For example, building information from BIM could be utilised for navigation of BIM users that considers the architectural barriers (e.g., for wheelchairs, baby strollers, and transport of large equipment within the facility). For this, a positioning system capable of seamlessly transitioning in and out of buildings is needed. This paper proposes a multi-sensor and cooperative positioning method of seamless environment transition and presents the first architecture concept. For this purpose, Global Navigation Satellite System (GNSS) data are planned to be fused with Ultra Wide-Band (UWB), Inertial Measurement Unit (IMU), Wireless-Fidelity (Wi-Fi), visual camera and LiDAR (Light Detection and Ranging) data. All sensors required for such a positioning system, are already available on modern smartphones commonly used as low-cost platforms for pedestrian navigation and localisation. Newer smartphones are also equipped with LiDAR and UWB chips, which are a future signal of opportunity, just like Wi-Fi currently is. The transition to different environments will be determined based on the UWB anchors called identical anchors that will be used to transition from the global coordinate system for the outdoor environment to the local building environment for positioning in indoor environments. The proposed multi-sensor and cooperative positioning system is expected to be able to achieve decimetre-level accuracy.</p>

Indoor localization has gained significant attention recently due to the growing demand for location-aware applications within indoor environments. Among various techniques, Wi-Fi fingerprinting combined with the Internet of Things (IoT) has emerged as a promising solution for accurate and cost-effective indoor localization. This paper surveys the existing research related to indoor localization using Wi-Fi fingerprinting with the IoT, aiming to provide a comprehensive understanding of the advancements, challenges, and potential applications in this field. The paper begins by introducing the fundamental concepts of indoor localization and the role of Wi-Fi fingerprinting in achieving accurate position estimation. In addition, this paper focuses on the contributions of Artificial Intelligence (AI), Machine Learning (ML), and Deep Learning (DL) in improving localization accuracy, robustness, and scalability. The reviewed papers have been examined from various aspects, including system architecture, deployment strategies, fingerprint creation techniques, and localization algorithms, with a discussion about the advantages and limitations of each approach. In addition to discussing the state-of-the-art techniques, this paper identifies research gaps and open challenges in indoor localization using Wi-Fi fingerprinting with the IoT. The findings presented in this paper can guide future research efforts, leading to the development of intelligent and context-aware IoT applications within indoor environments.

WiFi fingerprint based indoor localization has become a key research direction in the field of indoor localization due to its high positioning accuracy and low equipment deployment cost. Increasing the number of reference point collected offline can improve the positioning accuracy, however it yields excessive cost of offline collection. Fingerprint augment is an effective solution to reduce the cost while ensuring the positioning accuracy. In this paper, we are pioneering to propose a fingerprint augment framework based on super-resolution (FASR), which achieves the fusion of fingerprint augment and super-resolution based on mutual conversion between fingerprint data and fingerprint image. The processing framework of FASR is formulated and the implementation of three modules in FASR are given, including Fingerprint-To-Image Conversion module, Super-Resolution module and Image-To-Fingerprint Conversion module. Simulated and real data experiments reveal the feasibility and effectiveness of the FASR. In addition, we explore the impact of two key engineering parameters on the performance of the FASR method. Our work demonstrates the new application of super-resolution in image processing field in wireless indoor localization topics.