Iness Ahriz

High-Performance Indoor Localization with Full-Band GSM Fingerprints

GSM trace mobile measurements are used to study indoor handset localization in an urban apartment setting. Nearest-neighbor, Support Vector Machine (SVM), and Gaussian Process classifiers are compared. A linear SVM is found to provide mean room-level classification efficiency near 100%, but only when the full set of GSM carriers is used. To our knowledge, this is the first study to use fingerprints containing all GSM carriers, and the first to suggest that GSM could be useful for very high-performance indoor localization.

Practical indoor localization using ambient RF

The article presents a simple, practical approach for indoor localization using Received Signal Strength fingerprints from the GSM network, including an analysis of the relationship between signal strength and location, and the evolution of localization performance over time. Support Vector Machine regression applied to very high dimensional fingerprints does not reveal any smooth functional relationship between fingerprints and position. Classification using Support Vector Machines however provides very good results on discriminating different rooms in an indoor environment, albeit with performance that degrades over time. Transductive inference, introduced as a means of updating models to overcome degradation over time, provides hints that accurate indoor localization can be achieved by applying classification methods to cellular Received Signal Strength fingerprints, performance robustness being maintained via model updating and refining.

Hybrid indoor localization using GSM fingerprints, embedded sensors and a particle filter

The article presents an indoor localization scheme for mobile devices based on GSM Received Signal Strength fingerprints combined with embedded sensor information and an area site map. Displacements of a mobile user are first estimated using a sensor dead-reckoning approach that adapts stride length to different users and environments, and a dynamically switched orientation estimation scheme responding to orientation changes of the mobile device. Positions derived from GSM fingerprints, along with constraints imposed by a site map, are then integrated using a particle filter in order to prevent the accumulation of dead-reckoning errors over time. The study demonstrates that a standard handset with cellular network access and embedded inertial sensors can provide a good solution for indoor localization.

4 th order statistics based blind channel estimation for multicarrier transmission

This paper presents a blind channel estimation algorithm for multicarrier systems. The proposed scheme is based on fourth order statistics estimation of received data followed by a Gauss linearization of a non-linear system. The channel estimation is performed over a very short number of symbols in order to stay compliant with the channel coherence time. The proposed approach is well suited to CP-OFDM system transmitting circular M-QAM communication symbols but, being not based on the cyclic prefix properties, it could be applied, with some adjustments, to filterbank multicarrier waveforms transmitting M-OQAM communication symbols. The blind algorithm presented provides finally an increase of the useful throughput and it gives a high flexibility for the waveform usage, avoiding the difficult time and frequency pilot location optimization problem.

Fast, handset-based GSM fingerprints for indoor localization

Accurately localizing users in indoor environments remains an important and challenging task. The article presents new results on room-level indoor localization, using cellular Received Signal Strength fingerprints collected with a standard cellular handset programmed to perform fast scans of the 900 and 1800 Megahertz GSM bands as a user explores an indoor environment at a normal walking pace. Support Vector Machines are used to deal with the high dimensionality of the fingerprints. The study demonstrates that an appropriately programmed standard cellular handset can provide a simple, inexpensive solution for accurate room-level indoor localization.

Relevant CIR parameters selection for fingerprinting based location algorithm

This paper investigates the relevance of Channel Impulse Response (CIR) parameters for fingerprint based localization algorithm. We consider in this study a WLAN environment where a Multi-Layer Neural Network (MLNN) learns off-line the location from the location-dependent parameters of the CIR. Then, it calculates on-line the accurate position of a mobile station from given fingerprints. More importantly, an advanced study is conducted to identify the relevant parameters of CIR to be used as fingerprints which offer lower complexity and high location accuracy.

Greedy probabilistic approach for localization in IoT context

In this paper we propose a greedy probabilistic approach for localization in Wireless Sensors Network (WSN). This topic has received much attention since the WSN are considered as the basis in the emerging area of Internet of Things (IoT). The proposed method aims at increasing the performance of the grid based Compressed Sensing (CS) localization algorithm. This latter is based on the sparse nature of localization problem and select one grid point as user position. The grid point is selected based on correlation property. We propose in this paper to select a grid point based on probabilistic approach where grid point probabilities are calculated from the received signal strength. In a second step we propose to combine the grid positions weighted with their probabilities. The performance of the proposed approaches is evaluated through simulations and compared to CS algorithm results.

Compressed sensing-based centralized multiple targets localization

In this paper, we propose Received Signal Strength (RSS)-based localization method in WiFi network using the Compressed Sensing theory (CS). The main contributions are two-folds. First, we need no additional infrastructure more than the existing WiFi networks for targets localization. Second, in our approach, we have introduced an improvement of the Orthogonal Matching Pursuit (OMP) in order to relax the a priori knowledge of the number of targets to be located. The proposed approach offers an accurate recovery of multiple sparse targets locations using RSSI-measurements (RSS Indicator) at the Access Points (APs). Localization performance of the method was investigated by simulation in a multi target localization scenario. We have evaluated the mean localization error and its empirical Cumulative Density Function (CDF). Obtained results show a mean error about 0.5m in an area of 100m2.

Performance Evaluations in Optical and Wireless Networks for CONDOR project

Abstract Wireless and optical networks are widely used nowadays. These networks offer a high throughput thanks to their optical link and allow the development of multiuser applications. Because the network performance is an important issue to provide services to a great number of users while assuring users’ quality of service requirements, CONDOR (CONtribution a la Diffusion de l’histOiRe du traitement de l’information a l’aide du reseau de demain) project aims to evaluate the wireless and optical networks’ performances in terms of link quality, throughput, jitter and delay. Our results show that a high throughput in optical and wireless networks supports a big load through the launched mobile applications while P2P wireless network connections upset some video applications.

Comparison of similarity approaches for indoor localization

This paper presents a comparison study of different similarity metrics used for RSSI fingerprint based indoor localization. These metrics are used for nearest neighbor search which is a crucial step in fingerprint localization system. Including Euclidean distance, Manhattan distance and Gauss distance, the present study compares the localization error respect to a proposed parameter named “error density”. This latter is related to the location error and the size of the studied area. In addition, different methods of combining the locations of neighbors have been introduced to estimate the current position and their performances have been compared. Extensive implementation details are discussed and simulation is conducted to compare them. Obtained results show that the Kernel method combined with the weighted average exhibits the localization accuracy on the studied dataset.