Parallel processing for orbital maneuver detection

Abstract

With the increase in the number of objects orbiting Earth, space situational awareness (SSA) has becoming an important area of research in the space sector (Lemmens and Krag, 2014). This paper outlines approaches to increase the performance of SSA algorithms. Parallel processing is implemented in orbital maneuver detection methods with the goal of increasing the computational speed and accuracy of detection by fusing two well-known detection methods: Two-Line Elements (TLE) analysis and state propagation into a hybrid detection method. Speed is an important component when considering the implementation of real time SSA, as well as, with the increase in the number of objects orbiting Earth the throughput of maneuver detection algorithms needs to increase to stay feasible. The results show that implementing parallel processing increased the computational speed of TLE analysis by approximately eight times and of state propagation by approximately four orders of magnitude. In terms of computational speed, the novel method performed worse than TLE analysis and state propagation. However, the method reduced the number of false maneuver detections.