Study on the Satellite Telemetry Data Classification Based on Self-Learning

Peng Wan,Weiwei Jiang,Yafeng Zhan

doi:10.1109/access.2019.2962235

Abstract

Since great redundancy of telemetry data of spacecraft, telemetry data compression is a good solution for the limited bandwidth and contact wireless links. It is important to obtain accurate data characteristic firstly. State-of-the-art machine learning methods work well on data mining and pattern recognition under conditions of the given test data set, which could be used as the available tools for post-event data processing and analysis, such as trend forecasting and outlier detection, but they have not provided the proper solution from the source on-board. In this paper, four base classes of the telemetry data are suggested and studied through the time series feature and information entropy analysis, then a new on-board lightweight self-learning algorithm named Classification Probability calculation - Window Step optimization (CP-WS) is proposed to obtain the class features and make the decision of each single parameter from the continuous discrete telemetry time series. Simulation results show that, our algorithm correctly classifies the simulation and real mission data into the appropriate base class with advantages of high classification accuracy as 100% and adaptive computational complexity from <inline-formula xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink"> <tex-math notation="LaTeX">$O(L^{2})$ </tex-math></inline-formula> to <inline-formula xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink"> <tex-math notation="LaTeX">$O(L)$ </tex-math></inline-formula> , which could be used in satellite on-board data compression for space-to-ground transmission, especially for the deep space explorers to save important status with less on-board storage space.

Highlights

In current space missions, such as manned-space stations, earth observation satellites, deep space explorers, etc., the onboard telemetry plays an important role in helping Mission Control Center (MCC) to monitor the platform status, discover the abnormal phenomena, and acknowledge the remote control feedbacks
OUR PROPOSED MODEL In this paper, we propose the CP-Window Step (WS) telemetry classification model with the WS (Window-Step) algorithm embedded in the CP (Classification Probability) algorithm, that is, CP classification utilizes the classification test probability to describe the membership level of the current time series belonging to a certain classification; and WS optimization self-learns proper parameters from the continuous telemetry data stream in order to improve the CP classification performance, which could be further used as the data compression parameters in telemetry transmission missions
Exhaustive and heuristic methods are widely used in optimization: the former could obtain global optimal solution with high computational complexity, which is usually used in off-line data processing on ground; the latter could obtain local optimal solution with low computational complexity, such as Simulated Annealing (SA), Genetic Algorithm (GA), Particle Swarm Optimization (PSO), etc., which could be used in off-line data processing and on-line static data processing

Summary

Introduction

In current space missions, such as manned-space stations, earth observation satellites, deep space explorers, etc., the onboard telemetry plays an important role in helping Mission Control Center (MCC) to monitor the platform status, discover the abnormal phenomena, and acknowledge the remote control feedbacks. As a fast-developing technology in recent years, machine learning (ML) related technologies have been studied widely in space missions with telemetry in recent years. Tariq et al [4], Hundman et al [5], and Fuertes et al [6] studied on the spacecraft anomaly detection based on the K-Nearest Neighbor (KNN), Support Vector Machine (SVM), Long Short-Term Memory (LSTM) with the exhaustive testing on the telemetry of Centre National d’Etudes Spatiales (CNES) spacecraft. Iverson et al [7] and Robinson et al [8] studied on the space operation assistant based on the data-driven and model-based monitoring techniques applied in several space missions. The applications of machine learning in space missions mainly focus on the forecasting and outlier detection in order to provide the flight control procedure

Methods

Results

Conclusion