Identification and Characterization of Gravitational Wave Signals using Neural Network Algorithms Implemented on FPGA, in the Context of the Low-Latency Pipeline of the LISA Mission
Abstract
The Laser Interferometer Space Antenna (LISA) is one of the largest missions to be build and launched by the European Space Agency (ESA) with the implication of the National Aeronautics and Space Administration (NASA) and Japan Aerospace Exploration Agency (JAXA) agencies and it will be the first space-based gravitational wave (GW) observatory. It will consist of 3 satellites joined by laser interferometers, placed in a triangle, at a distance of 2.5 million kilometers that will follow the Earth in its orbit around the Sun for an in-depth study of the Gravitational Universe. Two of the keys to successfully detect, identify and characterize the gravitational waves by the LISA spacecraft are: (1) the identification of the protected periods for LISA observations, during which gravitational wave events are more likely to be detected and the instruments should interrupt their normal maintenance schedule, and (2) the generation of alerts of potential gravitational waves events towards other space and earth observatories capable to detect other types of signals (e.g. in other wavelengths) that are produce by the same event that produced a GW signal. All of the above will be done by the Low-Latency Pipeline system within the LISA Mission. In this work, we present the first results of a Neural Network algorithm trained to recognize and characterize gravitational wave patterns in LISA-like Signal + Noise data samples. The Neural Network is implemented using MATLAB's Deep Learning Toolbox™ framework and then coded into an Altera FPGA (ARRIA 10) with the purpose to generate low-latency alerts for other space or Earth observatories in the context of multi-messenger observations. In the future, this could become a fast and reliable data analysis and processing tool for on-board operations that will lead to a large reduction of the amount of data needed to be transferred to the ground-located mission operating center (MOC). The speed advantage of an FPGA is derived from the fact that the programmable hardware is customized to a particular algorithm, e.g., a neural network. Moreover, a FPGA can be reconfigured to accommodate changing requirements and in-flight updates, and it is tolerant to radiation induced effects, proving to be an innovative solution for high computational demands of next-generation spacecrafts. Bibliography: 1. LISA Laser Interferometer Space Antenna - A Proposal in Response to the ESA Call for L3 Mission Concept, Prof. Dr. Karsten Danzmann (Lead Proposer) and LISA Core Team, 2017, https://www.elisascience.org/files/publications/LISA_L3_20170120.pdf 2. A.I. Wiki: A Beginner's Guide to Important Topics, https://pathmind.com/wiki/lstm 3. Time series analysis of Gravitational Wave signals using neural networks, M. Carrillo, J. A González, M. Gracia-Linares, F. S. Guzmán, Journal of Physics: Conference Series 654 012001, 2015 4. Advancing NASA's On-Board Processing Capabilities with Reconfigurable FPGA Technologies, Paula J. Pingree, Chapter 5 of Aerospace Technologies Advancements, Book edited by: Dr. Thawar T. Arif, ISBN 978-953-7619-96-1, pp. 492, INTECH, Croatia, 2010
- Publication:
-
43rd COSPAR Scientific Assembly. Held 28 January - 4 February
- Pub Date:
- January 2021
- Bibcode:
- 2021cosp...43E1243C