VAP3: IoT Localization and Coverage Assessment
Thursday, 10 June 2021, 9:30-11:00, Zoom Room
Session Chair: Carlos T. Calafate (Univ. Valencia, Spain)
Coherent Multi-Channel Ranging for Precise Localization in Narrowband LPWA Networks: Performance Trials in an Indoor Environment
Vincent Berg (CEA LETI, France); Francois Dehmas (CEA-Leti Minatec, France); Florian Wolf (CEA Grenoble & University of Limoges, France)
Wearable health monitoring is a promising application where Low Power Wide Area (LPWA) connectivity and location services are required. LPWA radios are suited thanks to their small size, low power and low cost, while providing ubiquitous and indoor coverage. However, current LPWA applications do not provide accurate enough localization information for this application. Coherent multi-channel ranging has recently been proposed to improve localization precision. This paper studies the resilience of the new approach in an indoor environment. Field trials performed in office premises confirm the potential of the new technique where performance is significantly better than legacy time-of-flight even in this severe environment.
Improving CSI-Based Massive MIMO Indoor Positioning Using Convolutional Neural Network
Gregor Cerar (Jozef Stefan Institute & Jožef Stefan International Postgraduate School, Slovenia); Ales Svigelj (Jozef Stefan Institute, Slovenia); Mihael Mohorcic (Jozef Stefan Institute & Jozef Stefan International Postgraduate School, Slovenia); Carolina Fortuna and Tomaz Javornik (Jozef Stefan Institute, Slovenia)
Multiple-input multiple-output (MIMO) is an enabling technology to meet the growing demand for faster and more reliable communications in wireless networks with a large number of terminals, but it can also be applied for position estimation of a terminal exploiting multipath propagation from multiple antennas. In this paper, we investigate new convolutional neural network (CNN) structures for exploiting MIMO-based channel state information (CSI) to improve indoor positioning. We evaluate and compare the performance of three variants of the proposed CNN structure to five NN structures proposed in the scientific literature using the same sets of training-evaluation data. The results demonstrate that the proposed residual convolutional NN structure improves the accuracy of position estimation and keeps the total number of weights lower than the published NN structures. The proposed CNN structure yields from 2cm to 10cm better position accuracy than known NN structures used as a reference.
Dissemination of GNSS RTK Using MQTT
Ashwin Rao (University of Helsinki, Finland); Martti Kirkko-Jaakkola (Finnish Geospatial Research Institute FGI, Finland); Laura Ruotsalainen (University of Helsinki, Finland)
Precise positioning using Global Navigation Satellite System (GNSS) requires the GNSS receivers to compensate for the errors caused by distortion in the GNSS signal’s path due to atmospheric conditions. The Real Time Kinematics (RTK) technique uses terrestrial reference stations that continuously monitor the quality of GNSS signals and provide information that can be used by the GNSS receivers in the vicinity of a reference station to compensate for the errors. In this paper, we explore the performance of disseminating the RTK correction information using the Message Queuing Telemetry Transport (MQTT) protocol. Specifically, we compare the indirection costs of using MQTT over Ethernet, Wi-Fi, and 5G links, and we highlight the impact of 5G and Wi-Fi power savings when disseminating GNSS RTK using MQTT.
Coverage of LoRa Links with α-Stable Modeled Interfering Underlying IoT Networks
Romain Chevillon (Université de Nantes, France); Guillaume Andrieux (University of Nantes & IETR Laboratory, France); Jean Francois Diouris (University of Nantes, France)
IoT networks have been omnipresent in urban and sub-urban areas since a few years ago. Even if all the IoT protocols do not use the same characteristics (modulation, resource sharing, transmit power, bandwidth), sometimes they share the same bands, which can inevitably lead to interference on the neighboring networks, and thus to a decrease in the overall coverage. In this paper, we present an analytical study of the coverage for a LoRa network with underlying uncoordinated IoT networks. Thanks to stochastic geometry, we propose closed form expressions and analyze the success and the coverage probabilities for a LoRa network in an infinite area. The interference taken into account in this work come from both the LoRa network itself (so called co-SF interference) and the underlying IoT network, modeled with an α-stable distribution, which parameters are based on recent measured values.
Autoencoder-Based Characterisation of Passive IEEE 802.11 Link Level Measurements
Priyanka Neuhaus (Fraunhofer Institute for Integrated Circuits, Germany); Marcus Henninger (University of Stuttgart & Nokia Bell Labs, Germany); Andreas Frotzscher (Fraunhofer Institute for Integrated Circuits IIS & Design Automation Division EAS, Germany); Ulf Wetzker (Fraunhofer Institute for Integrated Circuits IIS, Germany)
Wireless networks are indispensable in today’s industrial manufacturing and automation. Due to harsh signal propagation conditions as well as co-existing wireless networks, transmission failures resulting in severe application malfunctions are often difficult to diagnose. Remote wireless monitoring systems are extremely useful tools for troubleshooting such failures. However, the completeness of data captured by a remote wireless monitor is highly dependent on the temporal, e.g., short-term interference, and spatial characteristics of its environment. It is necessary to first ensure that the data was completely captured at the remote monitor in order to maintain the integrity of the failure analysis, i.e., to avoid false positives. In this paper, we propose an autoencoder-based framework to evaluate the quality of wireless data captured at a remote wireless monitor. The algorithm is trained using data generated under controlled laboratory conditions and validated on testbed as well as real-world measurement data.