Nicholas P. Bertrand, Northrop Grumman Corporation; Derrick Cheung, Northrop Grumman Corporation; Joy Gu, Northrop Grumman Corporation; Chris McNamara, Northrop Grumman Corporation; Camille Saidnawey, Northrop Grumman Corporation; Ernie Zenker, Northrop Grumman Corporation
Keywords: detection algorithms, hosted payload, space domain awareness, thermal vacuum testing
Abstract:
A key component of space domain awareness (SDA) involves the detection, tracking, cataloging, and characterization of near-Earth space objects. In addition to general purpose monitoring of various orbits, SDA is crucial for increasing the resiliency of space assets against both accidental and adversarial collisions. While many conventional SDA tracking systems are designed for ground-based operation, there are clear advantages to on-orbit tracking payloads such as improved imaging capabilities due to the absence of atmospheric effects, diversity of vantage point, and improved reaction time. However, in common deployment scenarios, the SDA sensing system is a secondary component on satellites with assorted primary missions (e.g., communication, weather, radar, etc.). Thus, the SDA hardware resides as an auxiliary payload which shares residual resources with the host spacecraft, creating stringent constraints on size, weight, power (SWaP), and limits on downlink utilization. Additionally, in-flight SDA systems require the development of algorithms that account for imaging platform movement and operate efficiently on the resource-constrained flight-capable hardware.
In this work, we introduce NGSatSentry, a low SWaP, hosted SDA payload design. Our discussion focuses on NGSatSentry’s suite of on-board space object detection algorithms that adapt ideas from both mature space and ground-based systems and are suitable for deployment in low SWaP flight hardware systems. First, we explore the application of spatial filtering algorithms such as complementary median filtering and subspace projection for suppression of background and stars. Next, we propose a data reduction algorithm that identifies undesired stars in the scene and excludes them from the downlink without the need for an on-board star catalog. We validate our algorithms through experiments on simulated and real ground-based observations of objects in geostationary and other orbits. Finally, we present the results of recent testing on prototype system components in our facility’s thermal vacuum chamber and discuss implications on system design.
Date of Conference: September 14-17, 2021
Track: SSA/SDA