A payload data handling approach for small satellites

Abstract

In the recent years satellite launches have increased signifi cantly. This trend is further pushed by big constellations and the new space movement. With the increasing amount of satellites, the data processing capabilities on ground need to grow accordingly. This thesis conquers the task of designing eff ective satellite payload ground processing networks using state of the art technologies. The focus is on separating the payload data processing algorithms from the underlying software architecture algorithms. With the resulting framework, a fl exible and scalable architecture is designed to host processing algorithms. The target user for the introduced handling system is seen in small institutions, universities and Small and Medium-sized Enterprises (SMEs). The thesis aims to build a knowledge base compared to the proprietary, closed source solutions on the market. This thesis describes payload data from satellite missions, data processing level used and sets the focus on image data, as they are a big driver for ground processing. The steps, necessary to process satellite payload data is illustrated. For the underlying software architecture to process satellite data, commercial processing pipelines are introduced. Next to software basics like data exchange formats, databases and software deployment, software architectures are described. As the Flying Laptop mission is used as an example for the payload datahandling, it is introduced with its optical and ship data receiving payloads. The Payload On-board Computer (PLOC) is characterized next to the ground segment, the payload handling is embedded into. The processing of data is defi ned with six requirements using the Flying Laptop mission as an example. They range from data assembly and data correction to calibration, demosaicing, georeferencing and AIS data decoding. These requirements are proven to be checked by the implemented payload handling. Raw payload products with its anomalies are illustrated in this thesis, while presenting implemented and possible solutions for these anomalies. This includes image readout anomalies as well as brightness anomalies, fi ltered out during calibration. For the AIS ship data of Flying Laptop, a performance analysis of the instrument is presented. Next to the payload data processing, the implemented software framework is described. Requirements, defi ning such a handling system implementation are listed. They include backup capabilities, automatic processing, accessibility with new processing steps and user access. In addition, decentralization, extensibility and usability for other missions is required. The concept of nodes is the central element of the payload data handling framework. A node represents an atomic unit, capable of receiving data, altering, and forwarding them. Diff erent node types to assemble, process and store the data are identifi ed. Used database concepts are described and how the data on diff erent level is stored within. To enable users to gain access to the data, a front end and an API solution is implemented. Both provide diff erent query options to fetch data according to needs. A possible deployment philosophy using system services on Ubuntu machines is introduced. With the payload handling pipeline implemented for the Flying Laptop mission, results are presented. Several thousand images and more than 1.35 million AIS messages are received and processed. The introduced payload handling concepts can streamline the payload processing and enable companies and institutions to focus on their core business and value chain.