Improvement of satellite tracking at early orbit phase of a clustered launch

A unique characteristic of CubeSats launch is that it is essentially a piggy back launch, satellites would share a similar orbit. This characteristic induces a challenge in the early orbit operation as well as the determination of so-called TLEs (two-line elements). In the paper, an identification procedure is proposed and verified to classify satellites at the early orbit phase under which a cluster of satellites are closely located and beaconing in the same band. Using the rough TLE and the nominal frequency as the reference, the expected reception frequency is estimated through an orbit propagator. A wide-bandwidth radio signal recorder is employed to collect the received signal samples. A time-frequency processing algorithm is then developed to extract and analyze specific features of the satellite transmission data or beacon. Advantages of the proposed approach are the improvement of satellite orbit determination without relying on GPS information, and an early prediction of different satellites at the same time. In the paper, the software radio approach for data collection and processing will be described. Simulation results are then provided to assess the performance and limitation of the method. A field trial by processing collected data at the early orbit phase of a cluster of CubeSats is then reported to verify the proposed method.