When a vehicle travels at Mach 5 or faster, it generates a dense shockwave and a thermal boundary layer around its body. These aerodynamic effects refract incoming starlight in ways that normal star sensors cannot handle. The same layer also produces strong thermal radiation that can overwhelm the faint signals from stars . Previous attempts at star-based navigation for hypersonic vehicles failed because they could not compensate for this distortion in real time.
The Guangdong team developed a method to predict and compensate for this refraction, enabling the sensor to maintain lock on stellar patterns . They built:
The system achieved recognition rates above 80% even in interference-heavy environments, and over 99% star-pattern recognition in interference-free conditions . The project passed its final expert review on August 10, 2026 .
Celestial navigation uses stars as fixed reference points to determine a vehicle's position and orientation. By tracking the natural, predictable movements of stars, a sensor can calculate the vehicle's location without any external signals . The system requires no satellites, ground stations, or radio emissions — it is entirely passive.
| Feature | Celestial (star-based) — this system | Satellite-based (GPS/BeiDou) | Inertial navigation (INS) |
|---|---|---|---|
| External signal required | None — passive optical sensing of starlight | Requires continuous satellite signals from space | None — self-contained using accelerometers/gyros |
| Immunity to electromagnetic jamming | Very high — relies on light, not radio; cannot be jammed by RF electronic warfare | Low — GPS/BeiDou signals are weak and easily jammed or spoofed | Very high — no external emissions or receptions |
| Autonomy (no external reference) | Fully autonomous once stellar catalog is loaded; no ground or space contact needed | Fully dependent on satellite constellation availability | Fully autonomous |
| Limitations | Requires clear view of sky (may be degraded by clouds, day/night transition, or maneuvers that obscure the sensor); accuracy is good but historically less precise than GPS | Global coverage, but single point of failure under EW attack; easily denied | Drift accumulates over time without external updates; low-cost INS drifts significantly over hypersonic flight distances |
| Synergy potential | Can be used to correct INS drift — giving a hybrid system that is both jam-proof and stable over long flights | Often paired with INS for periodic corrections | Usually paired with GPS or celestial updates to prevent drift |
GPS and BeiDou signals are vulnerable to electronic warfare. They are weak, easily jammed, and can be spoofed by adversaries. Hypersonic weapons — traveling at speeds above Mach 5 — have a short time to target, making accurate navigation critical . A purely inertial navigation system (INS) drifts over time, and at hypersonic speeds that drift translates into significant targeting errors.
The star-based system matches INS in autonomy and immunity to jamming, but adds the ability to correct inertial drift using celestial observations. This provides long-range accuracy without ever needing a satellite signal . The trade-off — reliance on optical conditions and the need to accurately predict atmospheric refraction at Mach 5+ — appears to have been the core technical barrier that this research claims to have solved .
When a vehicle travels at Mach 5 or faster, it generates a dense shockwave and a thermal boundary layer around its body. These aerodynamic effects refract incoming starlight in ways that normal star sensors cannot handle. The same layer also produces strong thermal radiation that can overwhelm the faint signals from stars . Previous attempts at star-based navigation for hypersonic vehicles failed because they could not compensate for this distortion in real time.
The Guangdong team developed a method to predict and compensate for this refraction, enabling the sensor to maintain lock on stellar patterns . They built:
The system achieved recognition rates above 80% even in interference-heavy environments, and over 99% star-pattern recognition in interference-free conditions . The project passed its final expert review on August 10, 2026 .
Celestial navigation uses stars as fixed reference points to determine a vehicle's position and orientation. By tracking the natural, predictable movements of stars, a sensor can calculate the vehicle's location without any external signals . The system requires no satellites, ground stations, or radio emissions — it is entirely passive.
| Feature | Celestial (star-based) — this system | Satellite-based (GPS/BeiDou) | Inertial navigation (INS) |
|---|---|---|---|
| External signal required | None — passive optical sensing of starlight | Requires continuous satellite signals from space | None — self-contained using accelerometers/gyros |
| Immunity to electromagnetic jamming | Very high — relies on light, not radio; cannot be jammed by RF electronic warfare | Low — GPS/BeiDou signals are weak and easily jammed or spoofed | Very high — no external emissions or receptions |
| Autonomy (no external reference) | Fully autonomous once stellar catalog is loaded; no ground or space contact needed | Fully dependent on satellite constellation availability | Fully autonomous |
| Limitations | Requires clear view of sky (may be degraded by clouds, day/night transition, or maneuvers that obscure the sensor); accuracy is good but historically less precise than GPS | Global coverage, but single point of failure under EW attack; easily denied | Drift accumulates over time without external updates; low-cost INS drifts significantly over hypersonic flight distances |
| Synergy potential | Can be used to correct INS drift — giving a hybrid system that is both jam-proof and stable over long flights | Often paired with INS for periodic corrections | Usually paired with GPS or celestial updates to prevent drift |
GPS and BeiDou signals are vulnerable to electronic warfare. They are weak, easily jammed, and can be spoofed by adversaries. Hypersonic weapons — traveling at speeds above Mach 5 — have a short time to target, making accurate navigation critical . A purely inertial navigation system (INS) drifts over time, and at hypersonic speeds that drift translates into significant targeting errors.
The star-based system matches INS in autonomy and immunity to jamming, but adds the ability to correct inertial drift using celestial observations. This provides long-range accuracy without ever needing a satellite signal . The trade-off — reliance on optical conditions and the need to accurately predict atmospheric refraction at Mach 5+ — appears to have been the core technical barrier that this research claims to have solved .