Celestial Navigation

Horizon Estimation Techniques

Optical and Infrared Sensing of the Earth's Limb

In the hierarchy of flight stabilization, maintaining an accurate awareness of the aircraft's attitude, its roll and pitch, is a critical safety requirement. If an autopilot's primary inertial measurement unit faces degradation due to high-vibration profiles or electromagnetic interference, the system must find an independent physical reference to maintain stable flight. One of the most elegant ways to achieve this is by looking outward to detect the boundary where the Earth's surface meets space, a boundary known in aerospace engineering as the Earth's limb or the horizon line.

To detect this line reliably under varied flight conditions, autonomous platforms utilize optical and infrared sensors. While visible-light optical cameras can easily identify the horizon on clear days, they are highly vulnerable to fog, haze, cloud layers, and the transition into night.

To overcome these environmental limitations, it is possible to integrate infrared sensors. Infrared sensors do not rely on visible light; instead, they measure thermal radiation. Because the Earth's surface is relatively warm compared to the freezing void of space, there is a sharp, massive thermal gradient at the horizon.

By utilizing long-wave infrared sensors, the system detects this sudden drop in thermal energy. This thermal interface remains incredibly sharp and distinct even in the presence of heavy atmospheric haze, smoke, or complete darkness, providing an all-weather visual baseline.

Establishing Stable Attitude and Roll-Pitch References

Once the sensor captures the horizon, the software runs real-time processing algorithms to translate this visual boundary into pitch and roll angles. The horizontal span of the detected horizon line across the sensor's field of view acts as an absolute physical coordinate system.

When the aircraft experiences a roll deviation, the horizon line tilts relative to the camera's rectangular pixel grid. The computer calculates the angular slope of this tilt to determine the exact roll angle of the aircraft.

Similarly, when the aircraft pitches up or down, the horizon line moves vertically upward or downward relative to the central optical axis of the sensor. The computer measures this vertical displacement to calculate the pitch angle.

By feeding these passively derived roll and pitch angles directly into the flight control system, the autopilot can maintain a stable local attitude reference. This visual baseline acts as an un-jammable backup that prevents the aircraft from entering a dangerous attitude or losing aerodynamic control if its primary inertial sensors fail.

Positioning With Sun

Solar Azimuth and Elevation Dynamics

For long-range missions where an aircraft must navigate across continents without GNSS, the Sun provides a passive navigation beacon. Utilizing the Sun for positioning requires measuring two primary angles: the solar elevation and the solar azimuth.

Solar elevation is the vertical angle of the Sun above the local horizontal plane, while solar azimuth is the horizontal compass direction of the Sun relative to true geographic North.

To capture these angles in flight, the UAS utilizes specialized optical sensors called sun sensors. These lightweight sensors are mounted on the upper surfaces of the airframe.

By measuring the angle at which sunlight passes through a series of precision microscopic slits or onto a localized pixel array, the sensor determines the vector pointing directly to the center of the Sun. This vector is then translated into precise azimuth and elevation coordinates relative to the aircraft's local reference frame.

Combining Time and Solar Angles for Geographic Coordinates

Knowing the angles of the Sun is only useful if the APNT system knows exactly when the measurements were taken. Therefore, the APNT system must maintain an accurate onboard clock synchronized to Universal Coordinated Time (UTC). The flight computer contains an astronomical almanac, a highly detailed database that maps the exact coordinates of the Sun relative to the Earth for every second of every day of the year. When the sun sensor measures the solar elevation and azimuth at a specific timestamp, the flight computer queries the almanac to find where on Earth the Sun is directly overhead at that exact microsecond. This point is known as the geographic position of the Sun.

Using this geographic point as the center, the measured solar elevation angle defines a circle of equal altitude on the Earth's surface. The aircraft must be located somewhere along the circumference of this circle.

By taking multiple sequential measurements as the aircraft flies, or by combining the solar vector with another directional reference (such as the gravity vector or a magnetic heading), the APNT system calculates the intersection of these positioning lines. This intersection resolves the absolute latitude and longitude of the aircraft, providing a completely passive global positioning update that cannot be jammed or spoofed by ground-based radio transmitters.

Star Tracker Instrumentation

High-Sensitivity Camera Architectures and Star Catalogs

For high-altitude, long-endurance platforms operating in the stratosphere, or for tactical aircraft conducting night operations, star tracking offers the ultimate method for drift-free navigation. A star tracker is a highly specialized celestial sensor consisting of a high-sensitivity, low-noise digital camera pointing upward through an optical window in the fuselage.

Because stars are point sources of light with very low physical intensity when viewed through the atmosphere, star tracker cameras must use large aperture lenses and advanced sensors optimized for low-light performance.

Additionally, these cameras utilize active thermoelectric cooling systems to reduce thermal noise within the silicon pixels, ensuring that faint stars can be distinguished from background electronic noise.

Stored inside the star tracker's dedicated processor is a comprehensive stellar catalog containing the exact celestial coordinates, brightness levels, and geometric relationships of thousands of prominent stars.

Absolute Attitude Determination via Pattern Recognition

When the star tracker captures an image of the night sky, it initiates a real-time pattern recognition process. First, image processing software analyzes the frame to locate the exact center point, or centroid, of every visible star.

Once these centroids are mapped, the system runs "lost-in-space" algorithms to identify the stars. These algorithms do not need to know where the aircraft is currently pointing; instead, they analyze the angular distances and geometric triangles formed by the detected stars in the image.

By matching these geometric patterns against the onboard stellar database, the system identifies the exact constellations within its field of view.

Once the stars are identified, the star tracker computes the precise mathematical rotation required to align the camera's current coordinate frame with the known celestial coordinate system. This process resolves the absolute three-axis orientation (roll, pitch, and true heading) of the aircraft with incredible precision, often down to arcseconds.

Because this orientation estimate is anchored directly to the stars, it does not suffer from any drift over time. This makes star trackers the perfect tool to continuously calibrate and reset the drift of high-performance inertial navigation systems during long-range flights.

Daytime Star Tracking

Overcoming Atmospheric Scattering and Solar Glare

While star tracking has historically been restricted to night flight, modern systems are increasingly adopting daytime star tracking. The primary engineering challenge of tracking stars during the day is the intense brightness of the daytime sky.

Solar radiation entering the Earth's atmosphere collides with air molecules, scattering shorter blue wavelengths in all directions. This physical phenomenon, known as Rayleigh scattering, creates a bright background shroud of blue light that completely overpowers the faint light of stars when viewed with standard optical cameras.

To track stars in broad daylight, the system must find a way to increase the contrast between the stars and the bright background sky. This is achieved by shifting the tracking sensor's operational spectrum.

While Rayleigh scattering is extremely strong in the visible light spectrum, its intensity drops off significantly at longer wavelengths. By utilizing sensors that operate in the short-wave infrared (SWIR) spectrum, the system views a sky that is dramatically darker, allowing bright stars to stand out as distinct points of light even in the middle of the day.

Optical Filtering and Real-Time Signal Extraction

In addition to using short-wave infrared sensors, daytime star trackers utilize physical and digital filtering techniques to isolate stellar signals. The camera is equipped with narrow bandpass optical filters that block out the wavelengths dominated by solar scattering while allowing the specific infrared wavelengths emitted by stars to pass through to the sensor.

Once the filtered infrared image is captured, the system runs advanced real-time image processing algorithms. The software uses spatial filtering to analyze the image, suppressing the remaining uniform daylight glare while highlighting localized, high-frequency point sources of light.

By tracking these isolated daytime star centroids and cross-referencing them with the onboard astronomical catalog, the APNT system receives continuous, drift-free orientation and heading updates twenty-four hours a day, ensuring survival in heavily jammed airspace.