In this paper, a novel fingerprint-based localization technique is proposed, which is applicable for positioning user equipments (UEs) in cellular communication networks such as the long-term-evolution (LTE) system. This technique utilizes a unique mapping between the characteristics of a radio channel formulated as a fingerprint vector and a geographical location. A feature-extraction algorithm is applied to selecting channel parameters with non-redundant information that are calculated from the LTE down-link signals. A feedforward neural network with the input of fingerprint vectors and the output of UEs’ known locations is trained and used by UEs to estimate their positions. The results of experiments conducted in an in-service LTE system demonstrate that by using only one LTE eNodeB, the proposed technique yields a median error distance of 6 and 75 meters in indoor and outdoor environments, respectively. This localization technique is applicable in the cases where the Global Navigation Satellite System (GNSS) is unavailable, e.g., in indoor environments or in dense-urban scenarios with closely spaced skyscrapers heavily blocking the line-of-sight paths between a UE and GNSS satellites.

Neural-Network-Assisted UE Localization Using Radio-Channel Fingerprints in LTE Networks

In wireless sensor network, the exact positions of the sensor nodes is necessary for location-aware services. Traditional approaches are not producing satisfactory results. In this paper we propose the use of Time Difference of Arrival (TDOA) information with Neural network for accurate node localization. We use two artificial neural network models-Back Propagation Network (BPN) and Radial Basis Function (RBF) Network model for Wireless Sensor Network's node localization problem. Time Difference of Arrival (TDOA) data is used to calculate the distance information from anchor nodes to sensor nodes. This distance information was used to train the neural networks' models. Simulation results show the superiority of Radial Basis Function Network over Back Propagation Network in terms of root mean square error when training data density is high.

TDOA Based Node Localization in WSN Using Neural Networks

Radio-based locating systems allow for a robust and continuous tracking in industrial environments and are a key enabler for the digitalization of processes in many areas such as production, manufacturing, and warehouse management. Time difference of arrival (TDoA) systems estimate the time-of-flight (ToF) of radio burst signals with a set of synchronized antennas from which they trilaterate accurate position estimates of mobile tags. However, in industrial environments where multipath propagation is predominant it is difficult to extract the correct ToF of the signal. This article shows how deep learning (DL) can be used to estimate the position of mobile objects directly from the raw channel impulse responses (CIR) extracted at the receivers. Our experiments show that our DL-based position estimation not only works well under harsh multipath propagation but also outperforms state-of-the-art approaches in line-of-sight situations.

A Deep Learning Approach to Position Estimation from Channel Impulse Responses.

Object localization and tracking is essential for many applications including logistics and industry. Many local Time-of-Flight (ToF)-based locating systems use synchronized antennas to receive radio signals emitted by mobile tags. They detect the Time-of-Arrival (TOA) of the signal at each antenna and trilaterate the position from the Time Difference-of-Arrival (TDoA) between antennas. However, in multipath scenarios it is difficult to extract the correct ToA. This causes wrong positions. This paper proposes a signal processing method that uses deep learning to estimate the absolute tag position directly from the raw channel impulse response (CIR) data. We use the CIR together with ground truth positional data to train a convolutional neural network (CNN) that not only estimates non-linearities in the signal propagation space but also analyzes the signal for multipath effects. Our evaluation shows that our position estimation works in multipath environments and also outperforms classical signal processing in line-of-sight situations.

Convolutional Neural Networks for Position Estimation in TDoA-Based Locating Systems

Low-data-rate wireless networks can be deployed for physical intrusion detection and localization purposes. The intrusion of a physical object (or human) will disrupt the radio-frequency magnetic field and can be detected by observing the change in radio attenuation. This gives the basis for the radio tomographic imaging technology, which has recently been developed for passively monitoring and tracking objects. Due to noise and the lack of knowledge about the number and the sizes of intruding objects, multiobject intrusion detection and localization is a challenging issue. This paper proposes an extended variational Bayesian Gaussian mixture model (VB-GMM) algorithm in treating this problem. The extended VB-GMM algorithm applies a Gaussian mixture model to model the changed radio attenuation in a monitored field due to the intrusion of an unknown number of objects and uses a modified version of the variational Bayesian approach for model estimation. Real-world data from both outdoor and indoor experiments (using the radio tomographic imaging technology) have been used to verify the high accuracy and the robustness of the proposed multiobject localization algorithm.

Localizing Multiple Objects Using Radio Tomographic Imaging Technology

In an ultra-wide band (UWB) sensor network signal reflections from objects can be used to accurately determine the location. UWB signals are preferred in these types of sensor networks since they provide a very good resolution due to their fine time granularity. We propose an artificial neural network based localization algorithm to detect single object in a sensor network and compare its performance to Cramer-Rao bound and least squares estimator. Then we propose a two phase algorithm for multiple object detection and evaluate the algorithm for the case when there are two objects in a sensor network with three nodes.

Localization via TDOA in a UWB Sensor Network using Neural Networks

Localization is becoming more important with increasing number of cellular phone users. Due to safety aspects with increased emergency calls from mobile phones, new applications related to location based services, and the network optimization with increasing load, localization draws interest from both the academia and the industry. In this study, we propose a neural network based algorithm that uses multipath strengths to locate a mobile user without a GPS receiver. We validated our algorithm in a commercial network.

O. Dural

Papers

Localization via TDOA in a UWB Sensor Network using Neural Networks

Localization via multipath strengths in a CDMA2000 cellular network using neural networks