Abstract

This article is devoted to the tight integration of global navigation satellite systems (GNSS) and inertial navigation systems (INS) in modern navigation and positioning applications. In conditions of increased requirements for navigation solutions’ accuracy, reliability, and stability, such integration significantly improves the data quality, minimizing the impact of interference and signal loss. The article considers the basic principles of operation of the systems and the advantages and disadvantages of each method. The emphasis is placed on how tightly linking GNSS and INS data provides a more effective solution to positioning problems in complex conditions, such as urban areas or underground spaces.

Here are the sections that will be covered: GNSS and INS Basics, Causes of GNSS errors under challenging conditions and why Loosely and Tight coupling is implemented, GNSS and INS integration Schemes, and Tightly Coupled Sensor Fusion from Inertial Labs. The conclusion will summarize the benefits of GNSS and INS tight-coupling integration.

Section 1. GNSS and INS Basics

               Global Navigation Satellite System (GNSS) – technology is a collection of satellite positioning systems with worldwide coverage [1]. These systems aid in positioning, navigation, and timing (PNT) applications for commercial and military purposes.

GNSS consists of numerous satellites that orbit the Earth in fixed paths. Each satellite continuously broadcasts radio signals, including the signal’s time and current orbital position. A receiving device, such as a smartphone or GPS navigator, simultaneously captures signals from multiple satellites (see Figure 1). To achieve accurate positioning, data from at least four satellites is necessary; more details can be found here [2]. By measuring the time it takes for signals to travel from the satellites to the receiver, the system calculates the distance to each satellite. With information from several satellites, the receiver can determine its precise location using three coordinates: latitude, longitude, and altitude.

Figure 1. How GNSS works.

However, signals can be distorted due to atmospheric conditions, multipath propagation, and other factors, leading to increased errors or the inability to determine the user’s position. Therefore, GNSS is often integrated with an Inertial Navigation System.

Inertial Navigation System (INS) – is a navigation device that uses acceleration sensors (accelerometers), angular velocity sensors (gyroscopes), and a computing device to calculate the angles of orientation continuously, the speed of a moving object without the need for external landmarks [3].

The core components of an Inertial Navigation System (INS) are gyroscopes and accelerometers, with MEMS sensors commonly used today due to their compact size, lightweight, and low cost [4]. INS has become essential for modern navigation systems in mobile objects because it delivers comprehensive information about navigation parameters, including heading, pitch, roll angles, acceleration, speed, and the object’s coordinates. Additionally, INS operates autonomously, requiring no external input. Data from other navigation systems, particularly the Global Navigation Satellite System (GNSS), is integrated to address the accumulating errors in orientation angles and coordinates. This data fusion typically employs an algorithm based on the Kalman filter [5].

Since GNSS signals may not always be available to determine the user’s position, various integration schemes are implemented, which will be discussed in more detail below. In the next section, we will take a closer look at the causes of GNSS errors.

Section 2. What causes GNSS errors under challenging conditions, and why are loose and Tight coupling implemented?

In cities with tall buildings, the GNSS signal can be reflected. This phenomenon is called multipath propagation [6]. Multipath (echo) signals occur when the transmitted signal passes several paths to the receiver. The effect of interference of satellite signals is observed. As a result, several versions of the same signal appear, superimposed at different times and in various phases. This phenomenon can lead to delays and errors in the navigation system. Figure 2 clearly shows these situations.

Figure 2. Multipath propagation.

Modern multi-frequency receivers can use signal filtering and analyze signal timestamps to identify and correct delays caused by multipath propagation. A multi-frequency GNSS receiver provides more reliable position tracking in urban environments where single-frequency systems have significant errors.

Another approach uses DGPS, particularly Real-Time Kinematic (RTK) [7]. RTK allows for centimeter-level accuracy under poor ionospheric conditions by coordinating signals between a static GNSS base station and the user’s receiver. Communication between the receiver and the base station is implemented via an NTRIP Caster – a device that selects the desired base station near the receiver and transmits correction data [8]. Despite these advantages, RTK has limitations:

– Corrections will not be transmitted to the user’s receiver in poor visibility conditions, such as in a tunnel.

– RTK does not work when visibility is less than 5 of the identical GPS satellites at the same time at the base and rover [9].

– Stable RTK operation is not guaranteed further than 20-30 km from the base

– During geomagnetic storms, there may be no fixed solution (fixed solution – all phase ambiguities are resolved – an integer number of wavelengths on the satellite-receiver line) [10] since the RTK method is based on phase measurements of pseudoranges, even under ideal satellite visibility conditions and a short base-rover distance.

Another more advanced approach is Sensor Fusion (SF). SF combines and integrates data from several sensors and GNSS + INS, such as a barometer, odometer, magnetic compass, camera, etc. Data fusion improves overall accuracy by using additional information when one part of the system is affected by external factors. For example, a GNSS signal that has been jammed. In this case, the correct vehicle speed data continues to come from the odometer. Thus, SF is essential in both Loosely and Tightly coupling scenarios.

Let’s look at the differences between the GNSS and INS integration schemes.

Section 3. GNSS and INS integration Schemes

Four primary schemes of GNSS and INS integration can be distinguished: [11]:

• Uncoupled
• Loosely Coupled
• Tightly Coupled
• Ultra-Tightly Coupled (Deep Integration)

Let’s look at each of them and figure out the differences. Figure 3 shows the Uncoupled scheme. In the Uncoupled approach, the systems operate independently, generating two different navigation solutions. Typically, one of them, based on GNSS, is considered more accurate and is the primary solution if available. In addition, GNSS data is used to correct or reset the INS solution without analyzing the causes of sensor drift, as in other integration methods.

Figure 3. Uncoupled INS and GNSS integration.

If GNSS data is unavailable, the system relies solely on inertial sensors, which can drift quickly depending on their type. Due to this limitation, the Uncoupled strategy is not widely used.

               The following scheme, Figure 4, implements Loosely Coupled (LC) integration. The LC strategy, or “decentralized,” is implemented using a Kalman filter to integrate INS parameters and GNSS data. A separate block estimates the navigation solution based on raw measurements. These measurements from the GNSS receiver are processed independently, and the estimator accepts only the pseudo-range and Doppler shift as raw data. The difference between the INS and GNSS positions and velocities is input to the Kalman filter.

Figure 4. Loosely coupled INS and GNSS integration.

The scheme shown in Figure 4 operates in a closed loop, which means that the estimated INS errors are returned from the Kalman filter to the mechanization units and the INS itself. The estimated navigation errors are used to correct the INS state, and the sensor errors help to compensate for the raw INS measurements. In an alternative, open loop mode, the INS operates independently of the estimator, and the estimated errors are not fed back to the mechanization units and the INS. In this case, the inertial sensor errors are not corrected, which leads to a rapid increase in the inertial navigation state errors and significant distortions in the determination of position, velocity, and orientation. The open loop mode is only suitable for high-quality inertial sensors with minimal mistakes, while a closed loop strategy is necessary for low-cost MEMS-based systems.

The Tightly Coupled (TC) strategy, known as “centralized,” processes the INS navigation parameters and the raw GNSS data through a single central Kalman filter, Figure 5. Unlike the LC approach, where the raw GNSS measurements are analyzed in a separate filter, they are directly integrated into a single filter in TC. The difference between the pseudoranges and Doppler shifts and the predicted INS measurements are used as input to the Kalman filter.

Figure 5. Tightly Coupled INS and GNSS integration.

The predicted INS measurements (range and range rate) are calculated based on the GNSS satellite position and velocity obtained from the ephemeris and the user position and velocity determined by the INS. The concepts of closed and open loops also apply to the TC strategy.

When using the same system components, tightly coupled INS/GPS integration almost always provides better accuracy and reliability than loosely coupled INS, Figure 6.

Figure 6. Accuracy at Tightly and Loosely coupled.

The LC approach combines GNSS data with inertial information to produce the final integrated system output. In contrast, TC takes the integration deeper by directly combining raw GNSS measurements with INS data in a filter.

LC and TC integrations differ primarily in the type of information transferred between the systems: in the former, processed GNSS data are combined with INS solutions, while in the latter, raw GNSS measurements are coupled with INS predicted values. This creates different architectures: LC uses two separate filters, while TC uses a single centralized filter.

The advantage of the LC approach is that the individual filters are smaller, which speeds up data processing. This method is also more robust since the INS and GNSS operate independently and can continue to provide navigation solutions even if one of the systems fails. However, LC also has a significant drawback: it cannot offer GNSS data output when satellite availability is partial, for example, when fewer than four satellites are visible. Therefore, the TC strategy is often preferred in low-visibility conditions like urban canyons.

With Ultra-Tightly Coupled integration, the GNSS and INS devices no longer function as separate systems: GNSS data is used to estimate INS errors, and INS measurements support the GNSS receiver tracking loops. This form of integration requires deep interaction and access to the receiver firmware, usually implemented by equipment manufacturers or specialized software.

Section 4. Tightly Coupled Sensor Fusion from Inertial Labs

Inertial Labs INS are high-precision positioning systems using TC sensor fusion, Figure 7. In severe multipath interference or GNSS jamming and spoofing environments, traditional positioning, navigation, and timing (PNT) systems face challenges in providing accurate and consistent performance. It is becoming increasingly clear that modern automotive navigation systems must include additional sources independent of infrastructure and robust to external influences.

Figure 7. Sensor Fusion.

These systems must use alternative PNT solutions to ensure continuous provision of accurate navigation data and time synchronization, especially during extended GPS/GNSS outages. The Inertial Labs Assisted Data Ecosystem offers a range of alternative position, navigation, and timing sources and capabilities to provide highly reliable and accurate estimates of vehicle navigation data streams in GNSS-poor or GNSS-less environments for air, land, and sea platforms [12, 13].

Inertial Labs offers several products that exemplify advancements in sensor fusion technology. Key products include:

• Inertial Measurement Units (IMUs): These devices combine accelerometers, gyroscopes, and magnetometers to measure orientation, velocity, and gravitational forces precisely. Notable models include the IMU-P and IMU-F series.
• Attitude and Heading Reference Systems (AHRS): These systems, such as the AHRS-10, integrate data from multiple sensors to provide accurate orientation and heading information, which is crucial for aviation, marine, and autonomous vehicle applications.
• Inertial Navigation Systems (INS): Products like the INS-D and INS-P series combine IMUs with GPS/GNSS receivers, offering highly accurate positioning and navigation solutions for drones, robotics, and land vehicles.
• Marine and Land Solutions: Inertial Labs’ offerings for marine and land applications, such as the MRU-B and the GPS-Aided INS, enhance navigation and control for ships, underwater vehicles, and land-based systems.

These products demonstrate Inertial Labs’ commitment to leveraging sensor fusion technology to deliver precise, reliable, and versatile solutions across various industries.

As sensor fusion technology continues to evolve, driven by companies like Inertial Labs, we can expect even greater precision, efficiency, and versatility in various technological applications. This technology will ultimately transform industries and improve the accuracy and functionality of multiple systems.

Conclusion

Tight integration of global navigation satellite systems (GNSS) and inertial navigation systems (INS) is critical to improving the accuracy and reliability of navigation solutions in today’s environments. In increasingly complex environments, including urban and forested areas, the ability to process data from these systems together can significantly reduce the impact of interference and time delays. Tight-coupling technology with Sensor Fusion opens new horizons for industries ranging from aviation to autonomous driving, providing more reliable and accurate position data.

Integrating Sensor Fusion from Inertial Labs in the Tightly Coupling scheme improves the quality of navigation solutions. It lays the foundation for innovative applications that will shape the future of mobility and transportation management.

References

[1] Wikipedia. “Satellite Navigation.” Wikipedia, 19 Mar. 2020, en.wikipedia.org/wiki/Satellite_navigation.

[2] Jan Van Sickle. GPS for Land Surveyors. Boca Raton, CRC Press, Taylor & Francis Group, 2015.

[3] Wikipedia Contributors. “Inertial Navigation System.” Wikipedia, Wikimedia Foundation, 21 May 2019, en.wikipedia.org/wiki/Inertial_navigation_system.

[4] “MEMS.” Wikipedia, 26 Apr. 2023, en.wikipedia.org/wiki/MEMS.

[5] Wikipedia Contributors. “Kalman Filter.” Wikipedia, Wikimedia Foundation, 27 Mar. 2019, en.wikipedia.org/wiki/Kalman_filter.

[6] “What Is GNSS Multipath Mitigation?” U-Blox, 13 Sept. 2023, www.u-blox.com/en/technologies/multipath-mitigation.

[7] to Contributors. “Real Time Kinematic.” Wikipedia.org, Фонд Викимедиа, 24 Oct. 2008, ru.wikipedia.org/wiki/Real_Time_Kinematic. Accessed 27 Sept. 2024.

[8] Surveying Hub B.V. “What Is NTRIP, NTRIP Client and a NTRIP Caster?” Surveying Hub Community, 29 Sept. 2021, community.surveyinghub.com/t/what-is-ntrip-ntrip-client-and-a-ntrip-caster/50. Accessed 27 Sept. 2024.

[9] “Wayback Machine.” Archive.org, 2016, web.archive.org/web/20160304133930/www.gichd.org/fileadmin/pdf/LIMA/RTK_survey2.pdf. Accessed 27 Sept. 2024.

[10] “Space Weather and GPS Systems | NOAA / NWS Space Weather Prediction Center.” Www.swpc.noaa.gov, www.swpc.noaa.gov/impacts/space-weather-and-gps-systems.

[11] Angrisano, Antonio. GNSS/INS Integration Methods. 2011.

[12] Mendez, Maria. “Integrating INS with Aiding Data Technologies – the INS Ecosystem RoadMap.” Inertial Labs, 29 Apr. 2024, inertiallabs.com/integrating-ins-with-aiding-data-technologies-the-ins-ecosystem-roadmap/.

[13] Mendez, Maria. “Advancements in Sensor Fusion Technology.” Inertial Labs, 3 June 2024, inertiallabs.com/advancements-in-sensor-fusion-technology/. Accessed 27 Sept. 2024.