Diagram of the integrated navigation and guidance architecture

Student Aerospace Challenge

Autonomous Collision Avoidance.

Power-aware GNSS navigation and MPC guidance for LEO CubeSats

GNSS receiver energy saved: –69.8%
Keep-out zone never violated: 150 m
Total Δv for avoidance: ~2 m/s
IEEE Aerospace Conference paper: 2027

A navigation and guidance system that keeps a CubeSat safely clear of debris in low Earth orbit while switching its GNSS receiver on only when needed, to save power.

The problem

Low Earth orbit is getting crowded, and satellites increasingly need to avoid debris on their own. That takes an accurate position estimate, but running a GNSS receiver continuously uses a large share of a CubeSat’s power budget. Within the Student Aerospace Challenge, I worked on the guidance, navigation and control (GNC) of such a mission (work package 7).

What I did

I developed an integrated navigation system in which an Unscented Kalman Filter fuses GNSS readings, taken only when needed, with auxiliary sensors, while a supervisor decides when to switch the receiver on. On top of it, I built a Model Predictive Controller that plans avoidance manoeuvres over a rolling horizon and widens the keep-out zone around the debris as the predicted navigation uncertainty grows. I validated the full loop in MATLAB/Simulink against real GNSS error profiles from Sentinel-6A, as part of a six-person team.

Highlights

01Unscented Kalman Filter fusing duty-cycled GNSS with auxiliary sensors
02Supervisor logic deciding when the GNSS receiver switches on
03Model Predictive Control widening the keep-out zone as uncertainty grows
04Validated against real Sentinel-6A GNSS error profiles

Results

Spacecraft–debris separation over the encounter, with the safety radius and keep-out zone — Separation over a 6,743 s simulated encounter with real Sentinel-6A GNSS data: the adaptive safety radius keeps the debris outside the 150 m keep-out zone.

Control effort and accumulated Δv during the avoidance manoeuvre — Control effort and accumulated Δv for the avoidance manoeuvre — about 2 m/s in total.

Outcome

We presented this work at the Student Aerospace Challenge final in Paris. A paper describing it, “Robust MPC-Based Collision Avoidance Guidance and Safe Duty-Cycled GNSS Navigation for LEO CubeSats” (Fernández-Acero Campoamor, Valverde Sacristán, Yuste Pubill, Grande González, Soler i Pla, Xu), is in preparation for the IEEE Aerospace Conference 2027.