Academic literature on the topic 'Received signal power (RSS)'

Avec le développement rapide de l'Internet des objets (IoT), le besoin de services de localisation en intérieur, tels que la gestion des actifs, la navigation et le suivi, a également augmenté au fil du temps. Pour la localisation indoor, les systèmes de navigation par satellite tels que le GPS ont un usage limité par l'absence de visibilité directe avec les satellites.Diverses solutions ont été proposées pour la localisation en intérieur, telles que la trilatération, la triangulation et la navigation à l'estime, mais leurs performances sont limitées par les conditions du canal intérieur, telles que masquage et l'évanouissement par trajets multiples. En exploitant la correspondance entre les mesures des caractéristiques du signal sans fil et les positions, les méthodes basées sur l'empreinte digitale ont montré qu'elles pouvaient fournir de bonnes performances de localisation avec suffisamment de données. Cependant, la localisation indoor est toujours confrontée à des défis tels que l'évolutivité, le coût et la complexité, la confidentialité, etc.L'objectif de cette thèse est d'améliorer l'efficience de la localisation indoor en utilisant des techniques d'apprentissage automatique. Nous divisons le processus de localisation en deux phases : la phase de cartographie radio hors ligne et la phase de localisation en ligne. Pendant la phase hors ligne, nous introduisons l'analyse des jeux de données comme une étape intermédiaire entre la création des jeux de données et la localisation. Nous proposons deux indicateurs numériques de qualité des jeux de données qui peuvent fournir un retour d'information pour améliorer la carte radio. De plus, l'extraction de caractéristiques et le traitement des données à l'aide d'outils d'apprentissage automatique sont intégrés dans un but d'efficience, en réduisant la taille des données et la complexité de calcul, tout en améliorant les performances de localisation. Pour être plus précis, nous proposons une méthode de cartographie radio basée sur les k-means qui peut réduire le nombre d'empreintes de plus de 90 % sans perdre d'informations utiles dans la carte radio ni dégrader les performances de localisation. En explorant la nature hiérarchique des grands ensembles de données, nous proposons une méthode d'extraction de caractéristiques hiérarchiques qui peut réduire davantage la complexité de la localisation sans en compromettre les performances.Pour la phase de localisation en ligne, nous explorons à la fois l'apprentissage automatique traditionnel et l'apprentissage profond. Nous présentons d'abord plusieurs méthodes traditionnelles d'apprentissage automatique et comparons les performances de localisation sur des ensembles de données publiques. Nous cherchons à améliorer les performances de localisation des méthodes traditionnelles.Pour faire face aux problèmes de confidentialité et de complexité, nous introduisons un cadre d'apprentissage fédéré pour le problème de la localisation indoor. Dans ce cadre, les clients partagent uniquement leurs modèles locaux avec le serveur central au lieu des données d'empreintes radios. Nous comparons d'abord les performances de l'apprentissage fédéré et centralisé. Ensuite, nous étudions l'impact du nombre de clients et de la taille des données locales. Toujours dans un objectif d'efficience, afin d'économiser les ressources, nous cherchons à réduire le coût de communication pendant le processus d'apprentissage. Nous évaluons différentes techniques, notamment la sélection des clients, l'accumulation du gradient et la compression du modèle. Une méthode de compression efficace est proposée pour compresser les modèles locaux, ce qui permet de réduire le coût de communication de la liaison montante de 91,5 % sans compromettre les performances de localisation. Enfin, nous considérons des limites sur la capacité des liens montants et évaluons différentes stratégies de compression<br>With rapid development of Internet of Things (IoT), the need of indoor location-based services such as asset management, navigation and tracking has also grown overtime. For indoor localization, navigation satellite systems such as GPS has limited usage since a direct line-of-sight to satellites is unavailable.Various solutions have been proposed for indoor localization such as trilateration, triangulation, dead reckoning, but their performance is limited by indoor channel conditions, such as shadowing and multipath fading. By exploiting the mapping between wireless signal feature measurements and positions, fingerprinting based methods have shown the potential to provide good localization performance with sufficient data. However, indoor localization still faces challenges like scalability, cost and complexity, privacy, etc.The focus of this thesis is to improve efficiency of indoor localization using machine learning techniques. We divide the localization process into two phases: offline radio mapping phase and online localization phase. During the offline phase, we introduce dataset analysis as an intermediate step between dataset creation and localization. We propose two numerical dataset quality indicators which can provide feedback to improve the radio map. Moreover, feature extraction and dataset processing using machine learning tools are integrated to improve efficiency by reducing the data size and computation complexity while improving localization performance. We propose a k-means based radio mapping method which can reduce the number of fingerprints by over % without losing useful information in the radio map or degrading localization performance. By exploring the hierarchical nature of large datasets, we propose a hierarchical feature extraction method which can further reduce localization complexity without compromising localization performance.For the online localization phase, we explore both traditional machine learning and deep learning. We first introduce several traditional machine learning methods and compare the localization performance on public datasets. We aim to improve localization performance of traditional methods.To cope with privacy and complexity issue, we introduce federated learning framework for indoor localization problem. In this framework, the clients share only their local models to the central server instead of the fingerprinting data. We first compare the performance with federated and centralized learning. Then, we further study the impact on different client numbers and local data size. To reduce communication cost during the training process, we evaluate different measures including client selection, gradient accumulation and model compression. An efficient compression method is proposed to compress local models which can reduce the uplink communication cost by 91.5% without compromising localization performance. At last, we consider a limit on uplink capacity and evaluate different compression strategies

While WiFi-based indoor localization is attractive, the need for a significant degree of pre-deployment effort is a key challenge. In this paper, indoor localization with no pre-deployment effort in an indoor space, such as an office building corridor, with WiFi coverage but no apriori knowledge of the placement of the access points(APs) is implemented for mobile devices. WiFi Received Signal Strength(RSS) in the considered environment is used to build radio maps using WiFi fingerprinting approach. Two architectures are developed based on this localization algorithm. The first one involves a client-server approach where the localization algorithm runs on the server whereas the second one is a standalone architecture and the algorithm runs on the SD card of the mobile device.

Location estimation when deployed on wireless networks supports a range of services including user tracking and monitoring, health care support and push and pull marketing. The main subject of this thesis is improving indoor and outdoor location estimation accuracy using received signal strength (RSS) from neighbouring base stations (BSs) or access points (APs), without using the global positioning system (GPS) or triangulation methods. For the outdoor environment, state-of-the-art deterministic and probabilistic algorithms are adapted to exploit principal components (PCs) and clustering. The accuracy is compared with K-nearest neighbour (KNN) algorithms using different partitioning models. The proposed scheme clusters the RSS tuples based on deviations from an estimated RSS attenuation model and then transforms the raw RSS in each cluster into new uncorrelated dimensions, using PCs. As well as simple global dimensionality reduction using PCs, the data reduction and rotation within each cluster improves estimation accuracy because a) each cluster can model the different local RSS distributions and b) it efficiently preserves the RSS correlations that are observed (some of which are substantial) in local regions and which independence approximations ignore. Different simulated and real environments are used for the comparisons. Experimental results show that positioning accuracy is significantly improved and fewer training samples are needed compared with traditional methods. Furthermore, a technique to adjust RSS data so that radio maps collected in different environmental conditions can be used together to enhance accuracy is also demonstrated. Additionally, in the radio coverage domain, a non-parametric probability approach is used for the radio reliability estimation and a semi-supervised learning model is proposed for the monitoring model training and evolution according to real-time mobile users’ RSS feedback. For the indoor environment, an approach for a large multi-story indoor location estimaiii tion using clustering and rank order matching is described. The accuracies using WiFi RSS alone, cellular GSM RSS alone and integrated WiFi and GSM RSS are presented. The methods were tested on real indoor environments. A hierarchical clustering method is used to partition the RSS space, where a cluster is defined as a set of mobile users who share exactly the same strongest RSS ranking set of transmitters. The experimental results show that while integrating of WiFi RSS with GSM RSS creates a marginal improvement, the GSM data can be used to ameliorate the loss of accuracy when APs fail.

Traditional satellite positioning systems have limited resolution and have proved inaccuratein areas such as urban canyons where signals are subject to bounce phenomena or indeed may be entirely unavailable. An alternative method of positioning is that of tri/multilateration, which uses known positions and distances from beacon points to locate a receiver. In this project, a software was developed which used DSRC Basic Safety Messages (containing locational information) in combination with Received Signal Strength metrics (translated to distance information) to carry out such positioning in static environments. Initial studies confirmed that a signal received on the Craton 2 hardware was subject to considerable signal strength spread approximating a Gaussian distribution. A software was developed to simulate BSMs, including a measure of perturbation, over TCP. Three different traffic scenarios were constructed. Furthermore, multilaterationsoftware was developed to receive the BSMs and calculate position using three separate algorithms. The performance of these algorithms in the three different traffic scenarios was then evaluated. Lastly, the multilateration software was further developed to allow for the capture and processing of real BSMs sent on the 5.9 GHzband. The multilateration software was capable of determining the location of the receiver to varying degrees of accuracy, depending on the geometrical distribution of surrounding vehicles and the algorithm used to multilaterate. The 3D Linear Least Squares method performed well in situations where beacons were well spaced in three dimensions. Other implemented multilateration algorithms, i.e., a 2D Linear Least Squares method and a 3D Gauss Newton method, performed better in typical traffic scenarios where vehicles tend to be coplanar.The software developed provides a useful starting point for further developmentof static, but also dynamic, multilateration algorithms.

This master thesis investigates the possibility of locating LTE base stations, known as eNodeBs, using signal measurements collected by routers on trains. Four existing algorithms for transmitter localization are adopted: the centroid, strongest signal, Monte Carlo path loss simulation and power difference of arrival (PDoA) methods. An improved version of Monte Carlo path loss simulation called logloss fitting is proposed. Furthermore, a novel localization method called sector fitting is presented, which operates solely on the cell identity and geographical distribution of the measurements. The methods are evaluated for a set of manually located eNodeBs, and the results are compared to other external systems that can be used to locate eNodeBs. It is found that the novel sector fitting algorithm is able to considerably improve the accuracy of the logloss fitting and PDoA methods, but weighted centroid is overall the most accurate of the considered methods, providing a median error of approximately 1 km. The Google Geolocation API and Mozilla Location Service still provides estimates that are generally closer to the true location than any of the considered methods. However, for a subset of eNodeBs where measurements from all sectors are available, the novel sector fitting algorithm combined with logloss fitting outperforms the external systems. Therefore, a hybrid approach is suggested, where sector fitting combined with logloss fitting or weighted centroid is used to locate eNodeBs that have measurements from all sectors, while Google Geolocation API or Mozilla Location Service is used to locate the remaining eNodeBs. It is concluded that while the localization performance for those eNodeBs that have measurements from all sectors is relatively good, further improvements to the overall results can likely be obtained in future work by considering environmental factors, the angular losses introduced by directional antennas, and the effects of downlink power control.

Inter-Cell Interference (ICI) will be one of main problems for degrading the performance of future wireless networks at cell edge. This adverse situation will become worst in the presence of dense deployment of micro and macro cells. In this context, the Coordinated Multi-Point (CoMP) technique was introduced to mitigate ICI in Next Generation Wireless Networks (NGWN) and increase their network performance at cell edge. Even though the CoMP technique provides satisfactory solutions of various problems at cell edge, nevertheless existing CoMP handover schemes do not prevent unnecessary handover initialisation decisions and never discuss the drawbacks of CoMP handover technique such as excessive feedback and resource sharing among UEs. In this research, new CoMP-based handover schemes are proposed in order to minimise unnecessary handover decisions at cell edge and determine solution of drawbacks of CoMP technique in conjunction with signal measurements such as Reference Signal Received Power (RSRP) and Received Signal Received Quality (RSRQ). A combination of calculations of RSRP and RSRQ facilitate a credible decision making process of CoMP mode and handover mode at cell edge. Typical numerical experiments indicate that by triggering the CoMP mode along with solutions of drawbacks, the overall network performance is constantly increase as the number of unnecessary handovers is progressively reduced.

The ultra-wideband (UWB) technology has a vast unlicensed frequency spectrum, which can support precise indoor positioning in orders of centimeters. The features of UWB signals can be utilized for variety of applications. In this project first we present an empirical channel models to analyze the localization accuracy of the UWB technology for interactive electronic gaming (Ping-Pong) in Line-of-Sight (LOS) and Obstructed LOS (OLOS) scenarios. Then we introduce a new concept that we refer to as micro-gesture detection to handle the more refined motions of the hand, such as rotation, while one antenna is held by the user using features of UWB signal. We use four specific features of the UWB signals: time of arrival, power of the first peak, total power, and the Root-Mean Square (RMS) of the delay spread, for this purpose. As the hand rotates the position of the antenna in the hand and the external antenna changes from LOS to OLOS. We demonstrate that features of the UWB signals are more useful than the RSS signal of the Wi-Fi to detect this class of micro-gestures. We foresee this micro-gesture detection capabilities become helpful for the people with limited ability or visually impaired for implementation of simplified sign language to communication with electronic devices located away from a person. We compare gesture detection using multiple features of the UWB signal with traditional gesture detection using the received signal strength (RSS) of the Wi-Fi signal.

La navigation pédestre est un domaine de recherche en pleine croissance qui vise à développer des services assurant le positionnement et la navigation en continu des personnes à l'extérieur comme à l'intérieur de bâtiments. Dans cette thèse, nous proposons un prototype de service pour la navigation pédestre ubiquitaire qui tient compte des préférences de l'utilisateur et de la technologie de positionnement optimale disponible. Notre objectif principal est d'estimer, d'une façon continue, la position d'un piéton muni d'un smartphone. En premier lieu, nous proposons un nouvel algorithme, nommé UCOSA, qui permet de sélectionner la technologie de positionnement à adopter à tout moment le long du processus de navigation. L'algorithme UCOSA commence par inférer la nécessité de déclencher un processus de "handover" (changement de technologie) entre les technologies de positionnement détectées (i.e. quand les zones de couvertures se chevauchent) en utilisant la technique de la logique floue. Ensuite, il sélectionne la technologie optimale à l'aide d'une fonction qui calcule un score pour chaque technologie disponible et qui se compose de deux parties. La première partie représente les poids, calculés en utilisant la méthode d'analyse hiérarchique (AHP). Tandis que, la deuxième partie fournit les valeurs normalisées des paramètres considérés. L'algorithme UCOSA intègre aussi la technique de positionnement à l'estime appelé PDR afin d'améliorer le calcul de la position du smartphone. En second lieu, nous portons l'intérêt à la technique de positionnement par empreintes RSS dont le principe consiste à calculer la position du smartphone en comparant les valeurs RSSs enregistrées, en temps réel, avec les valeurs RSSs stockées dans une base de données (radiomap). La majorité des radiomaps sont représentées sous forme de grilles composées de points de référence (PR). Nous proposons une nouvelle conception de radiomap qui ajoute d'autres PRs au centre de gravité de chaque carré de la grille. En troisième lieu, nous abordons le problème de la construction du graphe modélisant un bâtiment multi-étages. Nous proposons un algorithme qui crée tout d'abord un graphe plan pour chaque étage, séparément, et qui relie ensuite les différents étages par des liens verticaux. En dernier lieu, nous étudions un nouvel algorithme nommé SIONA qui calcule et qui affiche d'une manière continue le chemin entre deux points situés à l'intérieur ou à l'extérieur d'un bâtiment. Plusieurs expériences réelles ont été réalisées pour évaluer les performances des algorithmes proposés avec des résultats prometteurs en termes de continuité et de précision (de l'ordre de 1.8 m) du service de navigation<br>Pedestrian navigation is a growing research field, which aims at developing services and applications that ensure the continuous positioning and navigation of people inside and outside covered areas (e.g. buildings). In this thesis, we propose a ubiquitous pedestrian navigation service based on user preferences and the most suitable efficient available positioning technology (e.g. WiFi, GNSS). Our main objective is to estimate continuously the position of a pedestrian carrying a smartphone equipped with a variety of technologies and sensors. First, we propose a novel positioning technology selection algorithm, called UCOSA for the complete ubiquitous navigation service in indoor and outdoor environments. UCOSA algorithm starts by inferring the need of a handover between the available positioning technologies on the overlapped coverage areas using fuzzy logic technique. If a handover process is required, a score is calculated for each captured Radio Frequency (RF) positioning technology. The score function consists of two parts: the first part represents the user preferences weights computed based on the Analytic Hierarchy Process (AHP). Whereas, the second part provides the user requirements (normalized values). UCOSA algorithm also integrates the Pedestrian Dead Reckoning (PDR) positioning technique through the navigation process to enhance the estimation of the smartphone's position. Second, we focus on the RSS fingerprinting positioning technique as it is the most widely used technique, which principle is to return the smartphone's position by comparing the real time recorded RSS values with the radiomap (i.e. a database of previous stored RSS values). Most of radiomap are organized in a grid, formed or Reference Point (RP): we propose a new design of radiomap which complements the grid with other RPs located at the center of gravity of each grid square. Third, we address the challenge of constructing a graph for a multi-floor building. We propose an algorithm that starts by creating the horizontal graph of each floor, separately, and then, adds vertical links between the different floors. Finally, we implement a novel algorithm, called SIONA that calculates and displays in a continuous manner the pathway between two distinct points being located indoor or outdoor. We conduct several real experiments inside the campus of the University of Passau in Germany to evaluate the performance of the proposed algorithms. They yield promising results in terms of continuity and accuracy (around 1.8 m indoor) of navigation service