Showing papers presented at "IEEE/ION Position, Location and Navigation Symposium in 2020"

WiFi-RTT Indoor Positioning

Abstract: Global navigation satellite systems (GNSSs) can deliver very good position estimates under optimum conditions. However, especially in urban and indoor scenarios with severe multipath propagation and blocking of satellites by buildings the accuracy loss can be very large. Using WiFi for indoor positioning is a common approach because WiFi infrastructure is widely deployed. Recently the WiFi IEEE 802.11-2016 standard was released, which includes a fine timing measurement (FTM) protocol, more commonly known as WiFi-round-trip-time (WiFi-RTT) protocol, for WiFi ranging. This paper researches timing based positioning algorithms, in this case using WiFi-RTT distance estimates. Based on two measurement campaigns, in an antenna measurement chamber and in a typical indoor environment, a WiFi-RTT distance error model is derived. Both measurement campaigns show, that the distance is underestimated, hence, the estimated distance is lower than the true distance. The WiFi-RTT distance error model is included in the likelihood function of a particle filter (PF) and the positioning performances is evaluated in an indoor scenario. These evaluations show clearly the possibility of using WiFi-RTT distance estimates for indoor positioning.

Satellite Navigation for the Age of Autonomy

Abstract: A bstract-Global Navigation Satellite Systems (GNSS) brought navigation to the masses. Coupled with smartphones, the blue dot in the palm of our hands has forever changed the way we interact with the world. Looking forward, cyber-physical systems such as self-driving cars and aerial mobility are pushing the limits of what localization technologies including GNSS can provide. This autonomous revolution requires a solution that supports safety-critical operation, centimeter positioning, and cyber-security for millions of users. To meet these demands, we propose a navigation service from Low Earth Orbiting (LEO) satellites which deliver precision in-part through faster motion, higher power signals for added robustness to interference, constellation autonomous integrity monitoring for integrity, and encryption / authentication for resistance to spoofing attacks. This paradigm is enabled by the ‘New Space’ movement, where highly capable satellites and components are now built on assembly lines and launch costs have decreased by more than tenfold. Such a ubiquitous positioning service enables a consistent and secure standard where trustworthy information can be validated and shared, extending the electronic horizon from sensor line of sight to an entire city. This enables the situational awareness needed for true safe operation to support autonomy at scale.

Navigation With Differential Carrier Phase Measurements From Megaconstellation LEO Satellites

Abstract: An opportunistic framework to navigate with differential carrier phase measurements from megaconstellation low Earth orbit (LEO) satellite signals is proposed. A computationally efficient integer ambiguity resolution algorithm is proposed to reduce the size of the integer least-squares (ILS) problem, whose complexity grows exponentially with the number of satellites. The Starlink constellation is used as a specific megaconstellation example to demonstrate the efficacity of the proposed algorithm, showing a 60% reduction in the size of the ILS problem. The joint probability density function of the megaconstellation LEO satellites' azimuth and elevation angles is derived for efficient and accurate performance characterization of navigation frameworks with LEO satellites, and to facilitate system parameter design to meet desired performance requirements. Experimental results are presented showing an unmanned aerial vehicle (UAV) navigating for 2.28 km exclusively using signals from only two Orbcomm LEO satellites via the proposed framework, achieving an unprecedented position root mean squared error of 14.8 m over a period of 2 minutes.

Mobile Positioning with Signals of Opportunity in Urban and Urban Canyon Environments

Abstract: This paper presents field test results of mobile positioning with signals of opportunity (SOOP) in urban and urban canyon environments and the lessons learned. The particular SOOP considered in this paper is the digital television (DTV) signals available in the United States, namely, ATSC-8VSB. The field tests include runs in downtown San Mateo and San Francisco Financial District, representing typical urban and urban canyon environments. DTV signals from six DTV stations together with IMU and GPS data are recorded aboard a ground vehicle. The field tests show that DTV signals are abundant in urban and urban canyon environments, contrary to a popular concern about signal availability due to blockage by high-risers. Positioning geometry is also not a problem even though the number of transmitters may be limited because the ranges to signal sources are relatively short and the resulting geometric dilution of precision (GDOP) is acceptable especially for the two-dimensional solutions. However, multipath is omnipresent, having two detrimental effects on ranging and positioning. Severe mobile fading is frequent that disrupts continuous tracking and non-line of sight (NLOS) signals introduce large errors to correlation peak based timing. To process multipath-dominant signals, a signal parameter estimation methodology is developed. In this approach, periodic signal patterns (field sync segments at 41 Hz) are searched for via correlation using a constant false alarm rate (CFAR) detector and their times of arrival (TOA) are extracted using the orthogonal matched pursuit (OMP) algorithm amidst multipath to form pseudorange measurements. Both standalone SOOP solutions and integrated SOOP/IMU solutions are generated for the test trajectories, which are then compared to GPS for performance evaluation. The study indicates that the processing algorithms presented in the paper can successfully detect the first arrival of SOOP in urban and urban canyon environments and extract its TOA precisely when the signals are available (out of fading or blockage). However, the timing information carried by such first arrivals, good enough to serve the primary purpose of communications, may not be so for ranging because of NLOS. As such, timing of first arrivals is not sufficient by itself for positioning and has to be used in conjunction with other data such as IMU, which is used in this paper. Yet, distinct and stable multipath signatures may be exploited, together with an environment map for instance, for persistent positioning in urban and urban canyon environments, which is a direction of our future research.

Economical Fused LEO GNSS

Abstract: In addition to Internet service, new commercial broadband low-Earth-orbiting (LEO) satellites could provide a positioning, navigation, and timing (PNT) service far more robust to interference than traditional Global Navigation Satellite Systems (GNSS). Previous proposals for LEO PNT require dedicated spectrum and hardware: a transmitter, antenna, and atomic clock on board every broadband satellite. This paper proposes a highperformance, low-cost alternative which fuses the requirements of PNT service into the existing capabilities of the broadband satellite. A concept of operations for so-called fused LEO GNSS is presented and analyzed both in terms of positioning performance and in terms of the economy of its use of constellation resources of transmitters, bandwidth, and time. This paper shows that continuous assured PNT service over ±60° latitude (covering 99.8% of the world's population) with positioning performance exceeding traditional GNSS pseudo ranging would cost less than 2 % of system capacity for the largest new constellations, such as SpaceX's Starlink or Amazon's Project Kuiper.

Comparing the ‘Big 4’ - A User's View on GNSS Performance

Abstract: With BeiDou-3 and Galileo complementing the legacy systems GPS and GLONASS, a total of four global navigation satellite systems (GNSS) has now become available that offer free and ubiquitous access to accurate positioning, navigation, and timing (PNT). Following an overview of the system status and capabilities, we compare the Big 4 GNSSs in terms of signal and clock characteristics. The signal-in-space range error (SISRE) are assessed and related to the achievable single-point positioning accuracy. Furthermore, service stability and availability aspects are adressed. With respect to geodetic users, we discuss the determination and prediction of GNSS satellite orbits and clocks from measurements of global receiver networks, which forms the basis for precise point positioning applications. Within the International GNSS Service (IGS) various analysis centers independently provide such products based on data of the IGS multi-GNSS network. The challenges in generating precise orbit and clock solutions for the individual constellations are discussed, and the significance of satellite metadata knowledge is highlighted.

Deep Learning-Aided Spatial Discrimination for Multipath Mitigation

Abstract: A deep learning-aided spatial discriminator for multipath mitigation is developed. The proposed system compensates for the limitations of conventional beamforming approaches, especially at the stages of: prefiltering, model order estimation (MOE), and direction-of-arrival (DOA) estimation. Three environments are considered to design and train the proposed deep neural networks (DNNs): indoor office buildings, indoor open ceiling, and outdoor urban area. The performance of the proposed DNN-based MOE is compared to the conventional approaches of minimum description length (MDL) criterion and Akaike information criterion (AIC). The proposed DNN-based MOE is shown to significantly outperform existing approaches and to increase the degrees-of-freedom. Four experiments are presented to assess the performance of the proposed system in multipath-rich environments corresponding to indoor pedestrian navigation and ground vehicle urban navigation with cellular long-term evolution (LTE) signals. The proposed system exhibited a position root mean-squared error (RMSE) of 1.67 m, 3.38 m, 1.73 m, and 2.16 m.

Pedestrian Inertial Navigation System Augmented by Vision-Based Foot-to-foot Relative Position Measurements

Abstract: In this paper, we investigate how self-contained pedestrian navigation can be augmented by the use of foot-to-foot visual observations. The main contribution is a measurement model that uses Zero velocity UpdaTe (ZUPT) and relative position measurements between the two shoes obtained from shoe-mounted feature patterns and cameras. This measurement model provides directly the compensation measurements for the three position states and three velocity states of a pedestrian. The involved features for detection are independent of surrounding environments, thus, the proposed system has a constant computational complexity in any context. The performance of the proposed system was compared to a standalone ZUPT method and a relative-distance-aided ZUPT method. Simulation results showed an improvement in accumulated navigation errors by over 90%. Real-world experiments were conducted, exhibiting a maximum improvement of 85% in accumulated errors, verifying validity of the approach.

Look Around You: Sequence-based Radar Place Recognition with Learned Rotational Invariance

Abstract: This paper details an application which yields significant improvements to the adeptness of place recognition with Frequency-Modulated Continuous-Wave scanning, 360-degrees field of view radar - a commercially promising sensor poised for exploitation in mobile autonomy. We show how a rotationally-invariant metric embedding for radar scans can be integrated into sequence-based trajectory matching systems typically applied to videos taken by visual sensors. Due to the complete horizontal field of view inherent to the radar scan formation process, we show how this off-the-shelf sequence-based trajectory matching system can be manipulated to detect place matches when the vehicle is travelling down a previously visited stretch of road in the opposite direction. We demonstrate the efficacy of the approach on 26 km of challenging urban driving taken from the largest radar-focused urban autonomy dataset released to date - showing a boost of 30 % in recall at high levels of precision over a nearest neighbour approach.

A Machine Learning Approach for GPS Code Phase Estimation in Multipath Environments

Abstract: A neural network (NN)-based delay-locked loop (DLL) for multipath mitigation in Global Positioning System (GPS) receivers is developed. The NN operates on equally-spaced samples of the autocorrelation function. The NN is trained using a statistical distribution model that takes into consideration multipath time delay and power attenuation. The performance of the proposed method is compared numerically and experimentally with three other conventional techniques: conventional early-minus-late DLL, narrow correlator, and high resolution correlator. It is demonstrated that the NN-based DLL produces smaller code phase root mean squared error compared to the three conventional techniques in high multipath environments.

A Motion Induced Passive Infrared (PIR) Sensor for Stationary Human Occupancy Detection

Abstract: Passive Infrared (PIR) sensors are commonly used in indoor applications to detect human presence. PIR sensors detect human presence by detecting the change in infrared radiation across the polarity of the sensor. Due to this, PIR sensors are unable to accurately detect stationary human subjects, which results in false negatives. In the pursuit of creating a low-cost solution for detecting stationary occupants in a closed space, the novel approach to mount a PIR sensor on a moving platform was developed (MI-PIR). This approach was developed for the system to artificially induce the motion that is necessary for stationary human detection. Utilizing the raw analog output of the PIR sensor and an artificial neural network (ANN), the closed space was accurately classified for room occupancy, the number of occupants, the approximate location of the human targets, and the differentiation of targets. This novel approach provides the advantages of a utilizing a single PIR sensor for human presence detection, while eliminating the major known drawback to this type of sensor. Scanning the room using a PIR sensor also allows for an expanded field of view (FoV) and a simpler deployment, in comparison to other approaches using a PIR sensor. Finally, MI-PIR expands the functionalities of PIR sensors by using an ANN to detect various other occupancy parameters. The experimental results show that the system can detect room classification with 99% accuracy, 91% accuracy in occupancy count estimation, 93% accuracy in relative location prediction, and 93% accuracy in human target differentiation. These results show promise for an application of tracking and monitoring an at-risk patient in an indoor setting.

Compensation of Systematic Errors in ZUPT-Aided Pedestrian Inertial Navigation

Abstract: We present a method to identify and compensate systematic errors in the ZUPT-aided pedestrian inertial navigation. We considered two main categories of systematic errors resulting in an underestimate of the length of the trajectory and a drift in the heading of the trajectory. In this study, we identified the dominant factors resulting in the trajectory length and heading errors to be residual velocity during the stance phase and g-sensitivity error of the gyroscopes, respectively. Magnetic motion tracking system was used to record the velocity of the foot during the stance phase. Rate table, tilt table, and shaker were used to calibrate the IMU g-sensitivity. After compensation, a more than $6\times$ systematic error reduction was demonstrated from 3.24m to 0.50m during a 100m straight line trajectory. To the best of our knowledge, this study is the first attempt to reduce the systematic errors in the ZUPT-aided pedestrian inertial navigation algorithmically, without adding extra sensing modalities.

Machine Learning-Assisted Anomaly Detection in Maritime Navigation using AIS Data

Abstract: The automatic identification system (AIS) reports vessels' static and dynamic information, which are essential for maritime traffic situation awareness. However, AIS transponders can be switched off to hide suspicious activities, such as illegal fishing, or piracy. Therefore, this paper uses real world AIS data to analyze the possibility of successful detection of various anomalies in the maritime domain. We propose a multi-class artificial neural network (ANN)-based anomaly detection framework to classify intentional and non-intentional AIS on-off switching anomalies. The multi-class anomaly framework captures AIS message dropouts due to various reasons, e.g., channel effects or intentional one for carrying illegal activities. We extract position, speed, course and timing information from real world AIS data, and use them to train a 2-class (normal and anomaly) and a 3-class (normal, power outage and anomaly) anomaly detection models. Our results show that the models achieve around 99.9% overall accuracy, and are able to classify a test sample in the order of microseconds.

Demonstration of Fused RTK (Fixed) + Inertial Positioning Using Android Smartphone Sensors Only

Abstract: This paper describes several experiments related to centimeter accurate positioning using the build-in GNSS receiver and inertial measurement unit (IMU) of a dual-frequency commercial smartphone. Using a choke-ring antenna platform to shield the smartphone from the ground multipath we were able to obtain a GNSS carrier phase (GPS+Galileo L1/L5) solution with good fixed ambiguities and approx. 2 centimeter precision. Furthermore, the GNSS antenna phase center (APC) within the smartphone was determined. An Allan variance analysis of the inertial measurement unit shows an unexpected good gyro bias instability of approx. 15 deg/h. An integrated realtime kinematic (RTK) GNSS+IMU solution was computed and a heuristic sensitivity analysis was performed.

Combining TDoA and AoA with a particle filter in an outdoor LoRaWAN network

Abstract: Internet of Things (IoT) applications that value long battery lifetime over accurate location-based services benefit from localization via Low Power Wide Area Networks (LPWANs) such as LoRaWAN. Recent work on Angle Of Arrival (AoA) estimation with LoRa enables us to explore new optimizations that decrease the estimation error and increase the reliability of Time Difference Of Arrival (TDoA) methods. In this paper, particle filtering is applied to combine TDoA and AoA measurements that were collected in a dense urban environment. The performance of this particle filter is compared to a TDoA estimator and our previous grid-based combination. The results show that a median estimation error of 199 m can be obtained with a particle filter without AoA, which is an error reduction of 10 % compared to the grid-based method. Moreover, the median error is reduced with 57 % if AoA measurements are used. Hence, more accurate and reliable localization is achieved compared to the performance of other baseline methods.

Precise positioning through a loosely-coupled sensor fusion of GNSS-RTK, INS and LiDAR for autonomous driving

Abstract: In this paper we describe the integration done between GNSS-RTK/INS/LiDAR in a loosely coupled Kalman Filter in the context of autonomous driving applications. Specifically, we focus in the assessment of potential LiDAR updates by comparing the velocity profile obtained by the GNSS/INS integration solution and the LiDAR observations. The results from a test drive are shown to provide an insight of the advantages of using the LiDAR updates in GNSS denied environments.

Optimal GPS Integrity-Constrained Path Planning for Ground Vehicles

Abstract: Path planning for a ground vehicle in an urban environment is considered. The vehicle is equipped with a GPS receiver and a road map. The vehicle desires to take the shortest path to reach a target destination, while guaranteeing that integrity monitoring-based measures are satisfied along its traversed path. A path planning algorithm is proposed that yields the optimal path to follow as well as suboptimal feasible paths. The integrity monitoring-based measure considered in this paper is the horizontal protection level (HPL), which refers to the statistical bound around the vehicle that guarantees the probability of the absolute position error exceeding a desired threshold is not larger than the integrity risk. Experimental results are presented showing that choosing the optimal path from the proposed algorithm reduces the average and maximum HPL by 2 m and 20.2 m, respectively, compared to choosing the shortest-time path, while introducing a negligible additional path length.

Solution Separation-Based FD to Mitigate the Effects of Local Threats on PPP Integrity

Abstract: In order to obtain protection levels for PPP in multipath prone environments, we formulate a coarse threat model for urban and suburban environments. Based on this threat model, we determine analytically the limitations a of solution separation approach. In particular, we derive for which range of fault rates and fault lag this approach is likely to be feasible. We then describe a solution separation fault detection algorithm adapted to this threat model. Using data collected in urban and suburban environments, we evaluate the benefits of using up to 3 GNSS constellations and dual frequency in PPP combined with an FD algorithm based on solution separation, and in particular its ability to protect against measurement outliers.

Deep Neural Network Approach to GNSS Signal Acquisition

Abstract: This paper investigates the use of data-driven models, popular in the machine learning literature, as an alternative to well-engineered signal processing blocks used in state-of-the-art GNSS receivers. Acknowledging that the latter are optimally designed and extensively tested, it is also agreed that when the nominal models do not hold the performance of the receiver might degrade. Particularly, we investigate the use of data-driven models in the signal acquisition stage of the receiver by addressing a classification problem from Cross Ambiguity Function (CAF) delay/Doppler maps. A discussion on the training of such models and future perspectives is provided. The detection results in nominal situations are then compared to the theoretical bound in the receiver operating characteristic (ROC) plots.

Adaptive Loop-Bandwidth Control Algorithm for Scalar Tracking Loops

Abstract: This paper presents a loop-bandwidth control algorithm for adaptive scalar tracking loops used in modern digital global navigation satellite system (GNSS) receivers. This algorithm modifies the noise bandwidth of the loop filter. The updated loop-bandwidth balances the signal dynamics and noise through a weighting function. The agility of the estimators defines the sensitivity of the algorithm against dynamics. This algorithm is applicable to the delay-, frequency- or phase-locked-loop (DLL, FLL, PLL) and to any order loop-filter, making it simpler to incorporate than other methods. The algorithm is first analyzed and evaluated in a software receiver. Second, it is implemented in an open software interface GNSS hardware receiver for testing in simulated scenarios with real-world conditions. The scenarios represent different dynamics and noise cases. The results show the algorithm's generic usability and advantage over fixed loop settings, while preserving minimum tracking jitter and stability.

Assessment of Real-time Multipath Detection with Android Raw GNSS Measurements by Using a Xiaomi Mi 8 Smartphone

Abstract: In the last years, the domain of Global Navigation Satellite System has been revolutionized by several technological and non-technological milestones. The first dual-frequency GNSS smartphone, Xiaomi Mi 8, was released in May 2018, while the number of mass-market dual-frequency chipsets has been constantly increasing since then. The availability of raw GNSS measurements in Android devices enabled several applications and it brought a set of remarkable tools to the GNSS research community. In this work the multipath effects on a smartphone-based positioning have been deeply examined. A long 22h static test has been performed over a geodetic pillar, thus exploiting the Xiaomi Mi 8 dual-frequency capabilities for evaluating different multipath observables. Both the Code-Minus-Phase (CMP) and the Multipath Linear Combination (MLC) observables have been constructed and investigated, while the Multipath Indicator flag available from the API also considered. The validation of these results is performed in the positioning domain, based on a Single Point Positioning (SPP) algorithm. The signals' robustness to the multipath effects for both GPS and Galileo systems has also been discussed, when considering both L1 and L5 signal-bands.

TEX-CUP: The University of Texas Challenge for Urban Positioning

Abstract: A public benchmark dataset collected in the dense urban center of the city of Austin, TX is introduced for evaluation of multi-sensor GNSS-based urban positioning. Existing public datasets on localization and/or odometry evaluation are based on sensors such as Iidar, cameras, and radar. The role of GNSS in these datasets is typically limited to the generation of a reference trajectory in conjunction with a high-end inertial navigation system (INS). In contrast, the dataset introduced in this paper provides raw ADC output of wideband intermediate frequency (IF) GNSS data along with tightly synchronized raw measurements from inertial measurement units (IMUs) and a stereoscopic camera unit. This dataset will enable optimization of the full GNSS stack from signal tracking to state estimation, as well as sensor fusion with other automotive sensors. The dataset is available at http://radionavlab.ae.utexas.edu under Public Datasets. Efforts to collect and share similar datasets from a number of dense urban centers around the world are under way.

Evaluating the Urban Trench Model For Improved GNSS Positioning in Urban Areas

Abstract: Urban environments still form a challenge for Global Navigation Satellite System (GNSS) positioning as they induce difficult conditions for signal propagation. Previous research showed a successful application of the Urban Trench Model to detect non-line-of-sight (NLOS) satellite signals and improve positioning under urban conditions. In this paper, we adapt the approach to the city of Hannover with a detailed 3D city model and focus on the analysis of receiver specific signal characteristics during line-of-sight (LOS) and NLOS phases, herein also investigating the reflection of a signal on building surfaces. The question will be studied whether it is useful to apply a reflection detour on the signal for an improvement in positioning.

Reinforcement Learning for UAV Autonomous Navigation, Mapping and Target Detection

Abstract: In this paper, we study a joint detection, mapping and navigation problem for a single unmanned aerial vehicle (UAV) equipped with a low complexity radar and flying in an unknown environment. The goal is to optimize its trajectory with the purpose of maximizing the mapping accuracy and, at the same time, to avoid areas where measurements might not be sufficiently informative from the perspective of a target detection. This problem is formulated as a Markov decision process (MDP) where the UAV is an agent that runs either a state estimator for target detection and for environment mapping, and a reinforcement learning (RL) algorithm to infer its own policy of navigation (i.e., the control law). Numerical results show the feasibility of the proposed idea, highlighting the UAV's capability of autonomously exploring areas with high probability of target detection while reconstructing the surrounding environment.

3D Dual-Shell Micro-Resonators for Harsh Environments

Abstract: This paper presents the recent advancements in the development of three-dimensional fused quartz dual-shell microresonators for environmentally-challenging applications, where the precision measurements are made through shock and vibrations. The dual-shell micro-resonators made from fused quartz and demonstrate a mechanical Q-factor of well above 1 million. An integration and assembly process for capacitive actuation and detection of such resonators using a silicon-in-glass electrode substrate was developed, and electrostatic tuning of $\mathbf{n}=\mathbf{2}$ wineglass using out-of-plane electrodes was demonstrated experimentally. We also present a simulation framework based on the Finite Element Method. The modeling approach was used to derive the critical design parameters of the dual-shell micro-gyroscopes for survivability under harsh shock waveforms. The developed 3D dual-shell structure is a potential solution for microresonators and gyroscopes for operation in harsh environments.

Geomagnetic Storm Induced Mid-latitude Ionospheric Plasma Irregularities and Their Implications for GPS Positioning over North America: A Case Study

Abstract: This paper presents mid-latitude ionospheric plasma irregularities associated with a geomagnetic storm and their impacts on high-precision GPS positioning solutions. We focus on the geomagnetic storm on 7–8 September 2017. Our study shows the vulnerability of mid-latitude GPS positioning to this geomagnetic storm event due to the storm-induced plasma irregularities. Results indicate more than 80% GPS stations over North America experienced large position errors (>0.5 m) within 30°−60° latitudes in the earlier period of the storm. Afterwards, the impacts became less significant and the large position error mainly concentrated within 50°−60° latitudes. The considerable degradation in the position accuracy correlates with occurrence of cycle slips that was attributed to ionospheric scintillations of GPS signals. This study allows us to improve knowledge of ionosphere response impacts on GPS at middle latitudes under extreme space weather conditions and increase awareness towards development of mitigation and predication means.

Multipoint Channel Charting With Multiple-Input Multiple-Output Convolutional Autoencoder

Abstract: We study the multipoint channel charting problem, where the channel state information (CSI) from multiple bases is used to generate channel charts for user relative positioning and many other applications. In previous work, only non-parametric methods are considered. In this paper, we fill the gap by proposing a novel neural network architecture - a multiple-input multiple-output (MIMO) convolutional autoencoder (CAE) - to solve the problem. Based on an open-source dataset, we demonstrate that for the use cases of user relative positioning and in region location verification (IRLV), compared with a baseline autoencoder (AE) with all fully-connected layers, the proposed network is able to achieve similar or better performance with a much smaller network size. In addition, we note that the proposed network is more capable of extracting useful features from CSI data and thus more promising for end-to-end learning.

Automotive-Radar-Based 50-cm Urban Positioning

Abstract: Deployment of automated ground vehicles (AGVs) beyond the confines of sunny and dry climes will require sub-lane-level positioning techniques based on radio waves rather than near-visible-light radiation. Like human sight, lidar and cameras perform poorly in low-visibility conditions. This paper develops and demonstrates a novel technique for robust 50-cm-accurate urban ground positioning based on commercially-available low-cost automotive radars. The technique is computationally efficient yet obtains a globally-optimal translation and heading solution, avoiding local minima caused by repeating patterns in the urban radar environment. Performance is evaluated on an extensive and realistic urban data set. Comparison against ground truth shows that, when coupled with stable short-term odometry, the technique maintains 95-percentile errors below 50 cm in horizontal position and 1° in heading.

Quantifying Feature Association Error in Camera-based Positioning

Abstract: Camera-based visual navigation techniques can provide six degrees-of-freedom estimates of position and orientation (or pose), and can be implemented at low cost in applications including autonomous driving, indoor positioning, and drone landing. However, feature matching errors may occur when associating measured features in camera images with mapped features in a landmark database, especially when repetitive patterns are in view. A typical example of repetitive patterns is that of regularly spaced windows on building walls. Quantifying the data association risk and its impact on navigation system integrity is essential in safety critical applications. But, literature on vision-based navigation integrity is sparse. This work aims at quantifying and bounding the integrity risk caused by incorrect associations in visual navigation using extended Kalman filters.

Localization Based on Channel Impulse Response Estimates

Abstract: Location-awareness using wireless signals is a key enabler for numerous emerging applications. Inspired by the recently proposed soft information (SI)-based localization, this paper develops a localization algorithm based on estimates of the channel impulse response (CIR), which inherently contains position information. We propose a delay-origin uncertainty model for describing the conditional distribution of the delays in CIR given node positions. A scalable localization algorithm is designed using belief propagation (BP) on a factor graph that incorporates the uncertainty model. The performance of the developed algorithm is quantified for mmWave signals using QuaDriGa channel simulator, showing decimeter-level localization accuracy in typical indoor environments.