ATTITUDE DETERMINATION FOR CUBESATS USING MEASUREMENTS FROM NEIGHBOURING SATELLITES IN LOW EARTH ORBIT

Abstract

Accurate attitude determination is essential for CubeSat navigation, communication, payload pointing, and autonomous formation-flying operations. This study investigated a cooperative attitude-determination approach in which measurements from neighbouring satellites in low Earth orbit (LEO) were integrated with conventional onboard sensors. A simulation-based quantitative research design was employed using a quaternion-based Extended Kalman Filter (EKF) supported by gyroscope, magnetometer, Sun-sensor, and inter-satellite line-of-sight measurements. A representative 3U CubeSat and three neighbouring satellites were modeled under multiple geometric and sensor-degradation scenarios using Monte Carlo simulations. The results indicated that the integration of one neighbouring satellite reduced overall attitude root mean square error (RMSE) from 2.05° to 1.26%, while three neighbouring satellites further reduced RMSE to 0.83°. Convergence time decreased from 342 s for the conventional configuration to 146 s for the three-neighbour configuration. The results also demonstrated that attitude-estimation performance was strongly influenced by neighbouring-satellite geometry and inter-satellite measurement noise. Under degraded conventional sensor conditions, cooperative estimation maintained substantially greater accuracy and robustness than the sensor-only architecture. The findings support the applicability of Wahba's problem as an underpinning theoretical framework for integrating inertial reference vectors and body-frame line-of-sight observations. Overall, cooperative neighbouring-satellite measurements demonstrated significant potential for improving the accuracy, observability, convergence, and fault tolerance of CubeSat attitude determination in future distributed LEO missions.