- Multi - antenna GNSS and INS/Odometer Coupling for Robust Vehicular Navigation, Ali Broumandan, University of Calgary (Canada), Gérard Lachapelle, University of Calgary (Canada), Niranjana Vagle, University of Calgary (Canada)
GNSS provide seamless navigation solutions in open sky environments with good satellite visibility. However, performance is limited in degraded environments such as urban canyons, tunnels, foliage conditions and fails to provide continuous navigation. Apart from the satellite visibility concern, GNSS signals intrinsically have low power, which makes them open to different interference scenarios including deliberate spoofing and jamming. GNSS and Inertial Navigation System (INS) have complementary error characteristics. GNSS has good long term accuracy, whereas INS has high short term performance. INS is self-contained and operates continuously. However, it suffers from accuracy degradation over time due to the integration of bias and drift of the Inertial Measurement Units (IMUs). Along with navigation solutions, attitude information can also be estimated with an INS, which is of importance in some applications. Different interference mitigation methods have been studied in the literature. Most of these techniques are based on different signal processing methods using single antenna. However, antenna array based techniques are known to be the most effective against all types of interference and jamming. For spoofing attacks, most of the research has focused on detecting the spoofing attack and alerting the receiver about the attack. However, to provide continuous navigation solutions in addition to detection, it is necessary to mitigate spoofing. In an antenna array based receiver, signals from different antenna are combined using different beamforming/null-steering techniques. To point beam in the direction of a desired satellite, both platform attitude and direction of the satellite are required. By using ephemeris, information direction of the satellite can be achieved. Platform attitude can be obtained from INS which is mounted on the same platform. Thus, to deliver protection against different types of interference and yet provide distortionless measurements and robust navigation, it is necessary to integrate antenna array based GNSS receivers with INS. A four-element antenna array is considered in this research. GNSS and INS are integrated in a loosely coupled manner. The integrated multi-antenna GNSS-INS can provide several advantages including (a) Robust consistency check between antenna array and INS solutions can be performed, (b) INS provides heading information for the antenna array receiver to perform distortionless beamforming and (c) Robust and continuous navigation solution in the presence of interference signals. Advantages of such integrated system is studied using tactical and automotive grade sensors for different spoofing and jamming scenarios in this paper. Several data sets were collected by mounting an antenna array on the top of a vehicle along with different IMUs. Initial results reveals that in the presence of spoofing signals, the signal antenna and GPS/INS receiver is spoofed whereas the proposed antenna array-INS-integration can detect and mitigate the spoofing attack. In the presence of jamming signals, the performance of signal antenna receiver is significantly degraded due to the jamming effect whereas the antenna array-INS integration provides continues high quality measurements. In the presence of long GPS outages due to interference signals, the performance of GPS/RISS is degraded due to the inherent IMU drifts.
- Position Matching Estimation Using 3D Simulator for GNSS Positioning in Multipath/Non - Line - Of - Sight Environments, Nabil Kbayer, ISAE-SUPAERO/TESA (France), Cedric Rouch, CNES (France), Mohamed Sahmoudi, ISAE - SUPAERO (France)
The well-known Extended Kalman Filter (EKF) is one of the most widely used algorithms in science and particularly in localization with GNSS measurements. However, this estimator is not efficient when the GNSS pseudorange (PR) measurements become contaminated by combined, potentially unbounded and non-Gaussian errors including multipath (MP) and non-line-of-sight (NLOS) biases. On the other hand, this kind of ranging measurements errors occurs generally in urban areas where GNSS-based positioning applications requires more accuracy and reliability.
To overcome this accuracy degradation, much of the previous research has focused on using 3D model for position candidates scoring among an array of candidate positions using different scoring function. In this regard, this research propose a likelihood function based on similarity between a conventional Least-Squares (LS) solution, obtained by a LS-like projection of true received PR measurements, and a virtual LS solution obtained by a LS-like projection of predicted PR measurement, obtained by use of the 3D GNSS simulator SPRING at each candidate point. In this work, we eliminate the receiver clock bias from the problem estimation by differentiation of all ranging measurements across satellites using a reference satellite selected using different indicators such as elevation angles, C/N0 levels and LOS probability. Another innovative aspect is the mathematical derivation of the theoretical performance of this method and the demonstration of it sub-optimal efficiency in case of good bias estimation using 3D GNSS simulator. The proposed method is also hybridized with the well-known Intelligent Urban Positioning (IUP) algorithm implemented by UCL, using a hypothesis-domain integration algorithm at the scoring level. The key strength of this approach is it sub-optimal accuracy, which is proven theoretically and using real GNSS data. In this paper, we propose also new modeling of the uncertainty on the bias prediction by 3D modeling to improve the accuracy of the proposed approach. Analysis on the effect of the considered grid of candidate positions on positioning performance of the proposed method will be detailed in this paper.
Notwithstanding the significantly computational loads, this approach offers valuable insights into robust GNSS positioning in presence of MP/NLOS receptions. First Experimental results using real GPS/GLONASS data in a deep urban area in Toulouse downtown (South-West of France) shows that the proposed method achieves a good positioning accuracy compared to conventional GNSS positioning.
- Bringing Moving Baseline RTK to the Mass – Markets, Cécile Mongrédien, U-Blox (Switzerland), Marten Strom, U-Blox (Finland), Jean-Philippe Doyen, U-Blox (Switzerland), Alex Parkins, U-Blox (United Kingdom)
Building on its recently introduced L1 RTK technology, u-blox has developed a moving baseline L1 RTK solution. While in conventional RTK mode, the reference station must remain stationary at a known location, in moving baseline RTK, both the reference and rover receivers can move while computing a centimeter-level accurate position between them. Moving baseline RTK is ideal for applications where the relative position between two moving antennas is required, for example, when a UAV is programmed to land on a moving platform or when two antennas are rigidly mounted on the same platform to derive accurate attitude information.
Moving baseline RTK is a well-known but complex and expensive technology. This paper focuses on the challenges associated with designing a very precise moving baseline RTK solution while maintaining cost, size, and power consumption as low as possible. Having reviewed the various challenges and discussed potential solutions, this paper presents the results of several test campaigns performed in Finland and Switzerland.
When the reference station moves, then its position changes over time. To ensure that the baseline remains as accurate as possible, the reference station position must be sent for each epoch the reference station measurements are sent. Since the RTCM standard does not include a time-stamped reference position message, it cannot readily be used to support moving baseline RTK. To circumvent this problem, proprietary messages are required. To facilitate integration, u-blox has requested and been granted a new RTCM message type to broadcast its proprietary information. Additionally RTCM3 multi signal messages seven (MSM7) are used. As these messages contain phase range rate information in addition to phase and code range information, they can help limit the performance degradation caused by linearization error when extrapolating the measurement of the moving reference receiver.
Despite these efforts, synchronization requirements between the reference station and the rover remain very stringent. While GNSS time can be determined at each receiver with an accuracy of less than 100 ns, the misalignment between two receivers can reach a few milliseconds. The clock alignment is therefore closely monitored to prevent the non-cancelation of double-difference errors from introducing biases in the baseline estimate, especially in presence of high relative receiver motion.
Since the rover must wait for time-matched reference station correction to compute its position, the moving baseline RTK solution will be greatly affected by the latency and reliability of the communication link. In particular, care must be taken to ensure that any gaps in the incoming correction stream is quickly detected and properly handled.
To assess and validate the performance of the u-blox moving baseline RTK technology, several measurement campaigns were conducted. While UAV applications were extensively covered during these test campaigns, a variety of motion and relative motion was also assessed, including a vehicular dual-antenna set-up for precise heading determination.
The performance of the moving baseline RTK solution is investigated both in the ambiguity and position domains. These results demonstrate the suitability of the u-blox moving baseline solution for mass-market applications.
- Worst Impact of Pseudorange nominal Bias on the Position in a Civil Aviation Context, Jean-Baptiste Pagot, ENAC (France), Yoan Gregoire, CNES (France), Olivier Julien, ENAC (France)
The objective of this paper is to investigate the impact of nominal biases that affect code pseudorange measurements estimated from GNSS signals. This impact is looked at position level considering that a least square algorithm is used to estimate the position from pseudorange measurements and this in a civil aviation context.
As an input of this work, it is assumed that the pseudorange nominal biases can be written as the sum of three components following the proposition made in [Macabiau et al., 2014]:
- Delays induced by the satellite antenna.
- Delays induced by the receiver antenna.
- Distortions induced by the satellite payload, the satellite antenna and the receiver antenna.
One important feature of those models is that they are dependent upon:
- the satellite antenna nadir (angle between the satellite/centre of the Earth line and the satellite/user line) and/or
- the satellite elevation and azimuth with respect to the user.
By consequence pseudorange nominal biases are dependent upon the relative position between the user and satellites.
The two main contributions of this publication are:
- The proposition of three models that are able to characterize the three code pseudorange nominal bias components. These models are based on a wide review of the state-of-the-art regarding each bias source reported on GPS L1 C/A signals.
- The estimation of the impact on the position of code pseudorange biases considering proposed nominal bias models. To take into account that pseudorange nominal biases are dependent upon the relative position between the user and the satellites, this estimation is made at different locations around the world and at different epochs using realistic orbital parameters to reproduce constellation geometries. A worst case is provided for each location, and corresponds to the maximum absolute position error obtained during a period of 24 hours (position biases are assessed every 2 minutes).
- Cycle – slip Detection and Repair Using a Low Cost Single Frequency Receiver with inertial – aiding, Yu Wang, ENAC (France), Olivier Julien, ENAC (France)
The general objective of this contribution is to investigate the navigation performance one can expect from a low-cost architecture (single-frequency receiver with low cost IMU) using carrier phase measurements in an urban canyon, where the frequency of occurrence of strong multipath environment, masking, Non-Line-of-Sight (NLOS) signal tracking, interference, etc.… is quite high and hard to mitigate.
To achieve a sub-meter precise positioning, GNSS carrier phase measurements should be used since the tracking errors associated to these measurements are significantly lower than those of code pseudo-ranges. However, they also suffer from two main drawbacks: the presence of an unknown integer number of carrier phase cycles and the lack of robustness of the Phase Lock Loop (PLL) resulting in frequent measurement losses and cycle slips, especially in urban areas.
As a consequence, until recently GNSS carrier phase measurements were only used by applications that were taking place mostly in a benign open-space environment. For instance, Real-Time-Kinematic (RTK) methodology has been widely approved to achieve precise positioning. However, recent work has started investigating the use of carrier phase measurements in urban and sub-urban areas, relying on the fact that new GNSS receivers use multiple constellations, thus providing more ‘good’ measurements that can be exploited by the receiver while strengthening the satellite geometry.
Another common methodology to improve positioning is the integration of an IMU together with GNSS, as the benefits and drawbacks of both sensors are very complementary. Many contributions haven proven this interest.
The present article aims at investigating what performance can be expected from a low-cost navigation platform (single frequency multi-constellation GNSS receiver with a low-cost IMU) in a urban and sub-urban environment when trying to exploit GNSS carrier-phase measurements. Starting from the work performed by Carcanague et al in 2012, a series of improvements are proposed:
- The use of all available GNSS systems.
- The integration of a low cost IMU
- A new cycle slip detection and repair mechanism to provide better and more reliable information to the PVT Kalman Filter (in - order to keep benefitting the high accuracy of GNSS phase measurable);
- An extra measurement selection based on the PVT Kalman filter innovations (measurements predictions from INS) is performed;
- A more robust carrier phase integer ambiguity validation procedure is used other than using an empirical fixed threshold.
This proposed algorithm is finally tested based on data collected on the Toulouse ring road as well as Toulouse city center. The use of this improved algorithm provides very interesting results in terms of position quality (less blunders) and accuracy, in particular regarding the percentage of time when the position error is sub-50cm.
- Framework and Performance Evaluation of a Ray Tracing - Software Defined Radio Method for GNSS Positioning in an Urban Canyon Environment, Rei Furukawa, Tokyo University of Marine Science and Technology (Japan), Nobuaki Kubo, Tokyo University of Marine Science and Technology (Japan), Takuji Ebinuma, Chubu University (Japan), Yoshimi Fujii, Kozo Keikaku Engineering.Inc,JP (Japan)
In recent years, positioning using Global Navigation Satellite System (GNSS) are widely being used with the spread of car navigation systems and smart phones. Furthermore, the use of applications that require stable high-precision positioning, such as those for information-oriented construction is also increasing.
In GNSS positioning, various factors exist between the satellite and GNSS receiver which causes degradation in positioning accuracy. Urban areas are particularly susceptible to multipath errors due to shielding and reflection from buildings, which causes a large deterioration in positioning accuracy . It is important to be able to study these influences theoretically to investigate and verify countermeasures against multipath errors. In this study, GNSS receiver outputs (signal level, pseudorange fluctuations, positioning results, etc.) will be evaluated. In many cases, positioning algorithms of GNSS receivers are not disclosed. For this reason, simulation of GNSS receivers using numerical model is difficult. We therefore use an actual GNSS receiver and emulated GNSS signal as input to the receiver. Moreover, it is appropriate to generate the GNSS signal considering the multipath radio wave propagation situation, and let the receiver process it to calculate the position. By utilizing software defined radio , flexible generation of GNSS signals are possible. For the effective prediction of multipath radio wave propagation in urban areas, ray tracing method using 3D building models is used. The ray tracing method is used to estimate multipath profiles, including the amplitude, delay, and phase due to the surrounding building environment.
Chapter 1 discusses the framework of the GNSS signal emulation using ray tracing and software defined radio (RT-SDR) method. Chapter 2 evaluates the accuracy of our framework with actual measurements, while Chapter 3 discusses satellite selection for feedback to actual measurements. Finally, Chapter 4 concludes our study.
- Assessment of New Tracking Architectures for Future GNSS Receivers in Harsh Environments, Mathilde Dufour, Altran Sud-Ouest (France), Christophe Ouzeau, Altran Sud-Ouest (France)
In GNSS receivers, classical scalar tracking algorithms allows to synchronize incoming signals with their locally generated replicas in terms of code delay, Doppler frequency and carrier phase. In that way, basic DLL, FLL and PLL are used. The signals tracking depends upon their waveforms that impact the corresponding tracking channel setting, in particular the choice of tracking loops characteristics: correlators, discriminators, filter orders and bandwidths.
As a consequence, to improve the receiver robustness, it is of interest to study new signals processing strategy. Amongst the potential receivers’ architectures, it is fed by the authors of the GNSS domain as an important task to investigate vector Kalman-based tracking architectures.
Vector tracking has many advantages over scalar tracking loops. The most commonly cited advantage is an increased immunity to interferences. The minimum carrier to noise power density ratio at which the receiver can operate is lowered by processing the signals in aggregate instead of separately.
In this paper, we suggest testing two Kalman-based FPLL tracking loop models. The purpose is to estimate jointly carrier phase and frequency tracking errors by one Kalman filter. Thus, the state vector is composed of tracking errors and the discriminators outputs are considered as observations.
The main difference between the two models concerns the definition of the filter gain that can be fixed or calculated according to measurements. Computing a variable gain requires tracking loop bandwidths knowledge, dynamics and measurement noise settings. This allows the filter to be robust against the receiver dynamics. In comparison, a fixed gain depends only on tracking loop bandwidths but its computation must be as optimal as possible because it weights all measurements in the same way.
This paper details models implementation: gain computation, choices of measurement matrix and how to estimate measurements. Inspired by two classical models (fixed gain and variable gain) that estimate carrier phase tracking error and frequency, the design of new models implies an adaptation of measurement matrix and gain. In various conditions (presence of feared events, filters setting…), the filters’ estimations are discussed. Finally, vector tracking is discussed.
- Navigation Satellite Fault Detection and Failure Cause Identification Methods Using Inter - satellite Links and Trigonometry Law, Jang JinHyeok, Konkuk University (South Korea), Lee Young Jae, Konkuk University (South Korea), Sung SangKyung, Konkuk University (South Korea)
The Global Navigation Satellite System (GNSS) is the most important navigation system currently used around the world. The U.S. Global Positioning System (GPS) was the first GNSS system to be launched, and various GNSS systems have been developed and in operation in some other countries. The primary objective of GNSS is to provide individual users with location information. This is done by a process termed trilateration, in which the distances between the user and four or more satellites are calculated on the basis of the position information transmitted from these satellites. Positioning accuracy depends therefore on the accuracy of the satellite’s orbit and clock data. In other words, the GNSS user can be provided with accurate positioning information, provided that there is no satellite fault. Positioning errors due to failure can also result in serious accidents in safety-critical areas such as aviation. Therefore, satellite fault detection is essential for GNSS operation.
This paper is intended to present a novel satellite fault detection method. In the proposed method, only satellites are used for fault detection and the inter-satellite distances are measured using inter-satellite links (ISL). Conventional satellite fault detection methods use ground measurements, which are subject to various error factors and must additionally undergo error estimation and correction processes, and accurate error removal is a great challenge. This paper proposes an ISL-based fault detection method to avert such problems. Also proposed is an algorithm that configures triangles using only satellites and detects faults using the trigonometry law paired with ISL measurements. Sensitivity analysis is then performed to compare the performance between proposed and conventional methods.
Satellite failure is caused by either satellite orbit or clock faults. It is of importance to identify the causes of satellite failure on occurrence in order to respond efficiently to fault. Predict the tendency of ISL measures according to the cause of satellite failure and check the change of test statistic according to this tendency. The failure cause identification method proposed in this paper draws on such tendency of test statistic. Finally, the operational concepts of satellite fault detection and satellite failure cause identification methods is explained.
This paper presents a fault detection algorithm improved by not using ground measurementsThis upgraded algorithm detects satellite faults at a finer level of details and confirms the detected faults more rapidly to enable a rapid response to failure. Also presented is a failure cause identification method for efficient failure management. Finally, the operational concepts of the proposed fault detection method using only satellites is explained as a succinct guideline for its applications