Stellar-40 Tactical-grade inertial navigation system

---

Dead reckoning is a method of navigating without GNSS signals by relying on inertia and other sensor data to maintain continuous tracking. Usually, when GNSS is lost, a system loses its position entirely. With dead reckoning, the system calculates a continuous position that will slowly drift over time. The rate of this drift depends on the quality of the internal IMU and the availability of external aiding sensors.

The Stellar-40 uses our proprietary sbgECom protocol, which utilizes clearly defined message frames. Using those predefined messages, the system can seamlessly ingest airdata, body velocities, and absolute position data from any external sensor capable of providing that information.

No, as it is not required. The Stellar-40’s advanced sensor fusion algorithm automatically calculates and estimates wind conditions in real-time, eliminating the need for you to input it manually.

For land platforms, the Stellar-40 supports a variety of aiding inputs to improve navigation accuracy, especially in GNSS-denied environments. These include hardware odometers (wheel odometry), velocity aidings from external sensors (such as LiDAR or optical sensors), and absolute position aidings.

The right choice depends on your operating environment. We highly recommend the damped version for platforms exposed to harsh, vibrating environments, such as deep strike drones or Unmanned Ground Vehicles (UGVs). If you are unsure which version suits your application, our support team is ready to help you make the right choice.

What is SBAS ?

Satellite-Based Augmentation System (SBAS) is a GNSS augmentation technology that improves the accuracy, integrity, availability, and continuity of satellite positioning. Initially developed to support aviation, SBAS has become a valuable navigation aid for applications such as UAVs, marine navigation, precision agriculture, GIS mapping, autonomous vehicles, and inertial navigation systems (INS). Unlike Real-Time Kinematic (RTK) or Precise Point Positioning (PPP), SBAS provides free meter-level positioning without requiring a local base station or an Internet connection.

A standalone GNSS receiver determines its position by measuring pseudoranges from satellites belonging to constellations such as GPS, Galileo, GLONASS, or BeiDou. However, these measurements are affected by several error sources, including satellite clock drift, orbit (ephemeris) errors, ionospheric delay, tropospheric delay, receiver noise, and multipath. SBAS continuously estimates the largest of these errors and broadcasts real-time corrections that significantly improve the navigation solution.

The SBAS architecture relies on a network of precisely surveyed reference stations distributed across a region. These stations continuously monitor GNSS signals and compare the measured satellite ranges with their known positions to estimate satellite clock, orbit, and ionospheric errors. The measurements are transmitted to master control centers, which combine data from all reference stations to generate correction messages and integrity information. These messages are then uplinked to geostationary satellites that rebroadcast them to SBAS-enabled GNSS receivers. The receiver applies the corrections before computing its position, effectively reducing the dominant error sources affecting the pseudorange measurements.

Mathematically, the corrected pseudorange can be expressed as:

Pcorrected​ = Pmeasured​ − Δclock ​− Δorbit​ − Δiono​

where Pmeasured is the original pseudorange measurement, Δclock represents the satellite clock correction, Δorbit​ is the orbital correction, and Δiono is the ionospheric delay correction.

These corrections typically reduce standalone GNSS horizontal positioning errors from approximately 3–10 meters to around 1–2 meters. Although this level of accuracy is lower than the centimeter-level performance achieved with RTK or PPP, SBAS offers a major advantage: integrity monitoring. Rather than simply improving position accuracy, SBAS continuously evaluates the health of every monitored satellite and rapidly alerts users if a satellite becomes unreliable because of clock anomalies, orbital errors, or signal degradation. This capability is essential for safety-critical applications where users must know whether the navigation solution can be trusted.

Several regional SBAS services operate around the world. These include WAAS in North America, EGNOS in Europe, MSAS in Japan, GAGAN in India, SDCM in Russia, KASS in South Korea, and SouthPAN in Australia and New Zealand. Modern multi-frequency GNSS receivers generally support multiple SBAS services automatically, depending on their geographic location.

For inertial navigation systems, SBAS provides a more accurate and reliable GNSS reference that improves sensor fusion performance. An Extended Kalman Filter (EKF) uses the corrected GNSS position and velocity measurements to periodically bound the drift accumulated by the inertial sensors. Compared with standalone GNSS, SBAS reduces measurement noise and satellite-related biases, resulting in more stable position and velocity updates while providing valuable integrity information for navigation algorithms. However, SBAS does not increase the GNSS update rate and cannot compensate for GNSS outages. During signal interruptions, overall navigation performance still depends primarily on the quality of the IMU and the effectiveness of the INS estimation algorithm.

SBAS is therefore an excellent solution for applications requiring free, globally available meter-level positioning with high reliability and integrity. It is particularly well suited for autonomous vehicles, UAVs, marine navigation, agriculture, and industrial systems where robust navigation is more important than achieving centimeter-level accuracy. When applications demand survey-grade positioning, RTK or PPP remain the preferred correction techniques, while SBAS serves as a dependable and widely available enhancement to conventional GNSS navigation.

How Inertial Navigation Systems Maintain Positioning During GNSS Outages

1 – Inertial Navigation Systems and GNSS Integration

An inertial navigation system (INS) determines position, velocity, and orientation from onboard inertial measurements. At its core, an inertial measurement unit (IMU) combines three accelerometers and three gyroscopes. The accelerometers measure specific force, while the gyroscopes measure angular rate along three orthogonal axes.

The navigation computer processes these measurements at a high rate. First, it integrates gyroscope measurements to estimate changes in attitude. It then transforms the measured acceleration into the navigation frame and compensates for gravity and other known effects. Finally, it integrates acceleration to estimate velocity and position. This process allows the INS to propagate a continuous navigation solution without relying on external signals.

When GNSS is available, the INS typically operates as part of an integrated GNSS/INS solution. GNSS provides an external reference for position and velocity, while the IMU provides high-rate motion measurements. The navigation filter combines both sources to estimate the vehicle’s state and constrain inertial errors.

When GNSS signals become unavailable, the system transitions to inertial propagation, often described as dead reckoning. The INS continues integrating gyroscope and accelerometer measurements from the last known navigation state. As a result, it can maintain continuous position, velocity, and attitude outputs during the outage.

However, inertial navigation does not eliminate positioning errors during an outage. Sensor bias, noise, scale-factor errors, misalignment, and other imperfections accumulate through integration. In particular, accelerometer errors affect velocity and then position, while gyroscope errors affect attitude and can introduce additional position errors through incorrect gravity compensation.

Therefore, the duration and magnitude of a GNSS outage directly influence the resulting position uncertainty. Higher-performance inertial sensors reduce error growth and allow the system to bridge longer outages before accumulated drift becomes significant.

With nearly two decades of experience engineering and manufacturing inertial sensors, SBG Systems develops IMU and INS technologies designed to provide reliable navigation performance when GNSS availability is limited or interrupted.

2 – Kalman Filtering and Sensor Fusion

Sensor fusion allows a GNSS/INS system to combine measurements with different characteristics. The IMU provides continuous, high-rate measurements but accumulates errors over time. GNSS provides an absolute position reference but can become unavailable or update less frequently. A navigation filter combines these complementary characteristics to estimate the vehicle’s state.

The Kalman filter is widely used for this type of state estimation. In practical INS applications, the system commonly uses an Extended Kalman Filter (EKF) because the equations describing vehicle motion and attitude are nonlinear. The EKF maintains an estimate of the navigation state and its uncertainty while processing measurements from the available sensors.

The process consists of two fundamental steps: prediction and measurement update.

During prediction, the filter uses the previous navigation state and the latest IMU measurements to estimate the next state. It also propagates the associated uncertainty. During the measurement update, the filter compares available external measurements with the predicted state. It then uses the difference, or innovation, to update the state estimate and reduce its uncertainty.

In a GNSS/INS system, GNSS measurements provide the external observations that constrain inertial drift. The filter can also estimate parameters such as accelerometer and gyroscope biases when the system model and available measurements make those parameters observable.

This distinction is important: a Kalman filter does not simply remove every error from every onboard sensor. Instead, it estimates the navigation state and relevant error states from a mathematical model and the available measurements. Residual errors remain because sensor imperfections, model inaccuracies, and unobserved disturbances cannot be eliminated completely.

When GNSS becomes unavailable, the filter can continue propagating the navigation solution using inertial measurements. However, without an external observation, the system has fewer opportunities to constrain its error states. Consequently, uncertainty increases as the outage continues. When GNSS measurements become available again, the filter compares them with the inertial solution and uses the resulting innovation to correct the estimated state and accumulated errors. This enables the GNSS/INS system to transition smoothly between GNSS-aided navigation and inertial propagation.