On-Board Satellite Telemetry Forecasting with RNN on RISC-V Based Multicore Processor

Danilo  Cappellone; S. Di Mascio; Gianluca  Furano; A. Menicucci; Marco  Ottavi

doi:10.1109/DFT50435.2020.9250796

On-Board Satellite Telemetry Forecasting with RNN on RISC-V Based Multicore Processor

Danilo Cappellone, S. Di Mascio, Gianluca Furano, A. Menicucci, Marco Ottavi

Space Systems Engineering

Research output: Chapter in Book/Conference proceedings/Edited volume › Conference contribution › Scientific › peer-review

6 Citations (Scopus)

Abstract

The aim of this paper is to assess the feasibility and on-board hardware performance requirements for on-board telemetry forecasting by implementing a Recurrent Neural Network (RNN) on low-cost multicore RISC-V microprocessor. Gravity field and steady-state Ocean Circulation Explorer (GOCE) public telemetry data was used for training RNNs with different hyperparameters and architectures. The prediction accuracy of these models was evaluated using mean error and R-squared score on the same test dataset. The implementation of the RNN on a RISC-V embedded device, representative of future space-grade hardware, required some adaptations and modifications due to the computational requirements and the large memory footprint. The algorithm was implemented to run in parallel on the 8 cores of the microprocessor and tiling was employed for the weight matrices. Further considerations have also been made for the approximation of sigmoid and hyperbolic tangent as activation functions.

Original language	English
Title of host publication	33rd IEEE International Symposium on Defect and Fault Tolerance in VLSI and Nanotechnology Systems, DFT 2020
Editors	Luigi Dilillo, Mihalis Psarakis, Taniya Siddiqua
Publisher	Institute of Electrical and Electronics Engineers (IEEE)
Number of pages	6
ISBN (Electronic)	978-1-7281-9457-8
DOIs	https://doi.org/10.1109/DFT50435.2020.9250796
Publication status	Published - 2020
Event	2020 33rd IEEE International Symposium on Defect and Fault Tolerance in VLSI and Nanotechnology Systems - On-line Virtual Event Duration: 19 Oct 2020 → 21 Oct 2020 Conference number: 33rd

Publication series

Name	33rd IEEE International Symposium on Defect and Fault Tolerance in VLSI and Nanotechnology Systems, DFT 2020

Conference

Conference	2020 33rd IEEE International Symposium on Defect and Fault Tolerance in VLSI and Nanotechnology Systems
Abbreviated title	DFT 2020
Period	19/10/20 → 21/10/20

Keywords

Deep Neural Networks
RISC-V
Space Systems
Artificial Intelligence

Access to Document

10.1109/DFT50435.2020.9250796

Cite this

Cappellone, D., Di Mascio, S., Furano, G., Menicucci, A., & Ottavi, M. (2020). On-Board Satellite Telemetry Forecasting with RNN on RISC-V Based Multicore Processor. In L. Dilillo, M. Psarakis, & T. Siddiqua (Eds.), 33rd IEEE International Symposium on Defect and Fault Tolerance in VLSI and Nanotechnology Systems, DFT 2020 Article 9250796 (33rd IEEE International Symposium on Defect and Fault Tolerance in VLSI and Nanotechnology Systems, DFT 2020). Institute of Electrical and Electronics Engineers (IEEE). https://doi.org/10.1109/DFT50435.2020.9250796

Cappellone, Danilo ; Di Mascio, S. ; Furano, Gianluca et al. / On-Board Satellite Telemetry Forecasting with RNN on RISC-V Based Multicore Processor. 33rd IEEE International Symposium on Defect and Fault Tolerance in VLSI and Nanotechnology Systems, DFT 2020. editor / Luigi Dilillo ; Mihalis Psarakis ; Taniya Siddiqua. Institute of Electrical and Electronics Engineers (IEEE), 2020. (33rd IEEE International Symposium on Defect and Fault Tolerance in VLSI and Nanotechnology Systems, DFT 2020).

@inproceedings{4e4a33c3424c4c0aacc2694bc1beb27b,

title = "On-Board Satellite Telemetry Forecasting with RNN on RISC-V Based Multicore Processor",

abstract = "The aim of this paper is to assess the feasibility and on-board hardware performance requirements for on-board telemetry forecasting by implementing a Recurrent Neural Network (RNN) on low-cost multicore RISC-V microprocessor. Gravity field and steady-state Ocean Circulation Explorer (GOCE) public telemetry data was used for training RNNs with different hyperparameters and architectures. The prediction accuracy of these models was evaluated using mean error and R-squared score on the same test dataset. The implementation of the RNN on a RISC-V embedded device, representative of future space-grade hardware, required some adaptations and modifications due to the computational requirements and the large memory footprint. The algorithm was implemented to run in parallel on the 8 cores of the microprocessor and tiling was employed for the weight matrices. Further considerations have also been made for the approximation of sigmoid and hyperbolic tangent as activation functions.",

keywords = "Deep Neural Networks, RISC-V, Space Systems, Artificial Intelligence",

author = "Danilo Cappellone and {Di Mascio}, S. and Gianluca Furano and A. Menicucci and Marco Ottavi",

year = "2020",

doi = "10.1109/DFT50435.2020.9250796",

language = "English",

series = "33rd IEEE International Symposium on Defect and Fault Tolerance in VLSI and Nanotechnology Systems, DFT 2020",

publisher = "Institute of Electrical and Electronics Engineers (IEEE)",

editor = "Luigi Dilillo and Mihalis Psarakis and Taniya Siddiqua",

booktitle = "33rd IEEE International Symposium on Defect and Fault Tolerance in VLSI and Nanotechnology Systems, DFT 2020",

address = "United States",

note = "2020 33rd IEEE International Symposium on Defect and Fault Tolerance in VLSI and Nanotechnology Systems , DFT 2020 ; Conference date: 19-10-2020 Through 21-10-2020",

}

Cappellone, D, Di Mascio, S, Furano, G, Menicucci, A & Ottavi, M 2020, On-Board Satellite Telemetry Forecasting with RNN on RISC-V Based Multicore Processor. in L Dilillo, M Psarakis & T Siddiqua (eds), 33rd IEEE International Symposium on Defect and Fault Tolerance in VLSI and Nanotechnology Systems, DFT 2020., 9250796, 33rd IEEE International Symposium on Defect and Fault Tolerance in VLSI and Nanotechnology Systems, DFT 2020, Institute of Electrical and Electronics Engineers (IEEE), 2020 33rd IEEE International Symposium on Defect and Fault Tolerance in VLSI and Nanotechnology Systems , 19/10/20. https://doi.org/10.1109/DFT50435.2020.9250796

On-Board Satellite Telemetry Forecasting with RNN on RISC-V Based Multicore Processor. / Cappellone, Danilo ; Di Mascio, S.; Furano, Gianluca et al.
33rd IEEE International Symposium on Defect and Fault Tolerance in VLSI and Nanotechnology Systems, DFT 2020. ed. / Luigi Dilillo; Mihalis Psarakis; Taniya Siddiqua. Institute of Electrical and Electronics Engineers (IEEE), 2020. 9250796 (33rd IEEE International Symposium on Defect and Fault Tolerance in VLSI and Nanotechnology Systems, DFT 2020).

Research output: Chapter in Book/Conference proceedings/Edited volume › Conference contribution › Scientific › peer-review

TY - GEN

T1 - On-Board Satellite Telemetry Forecasting with RNN on RISC-V Based Multicore Processor

AU - Cappellone, Danilo

AU - Di Mascio, S.

AU - Furano, Gianluca

AU - Menicucci, A.

AU - Ottavi, Marco

N1 - Conference code: 33rd

PY - 2020

Y1 - 2020

N2 - The aim of this paper is to assess the feasibility and on-board hardware performance requirements for on-board telemetry forecasting by implementing a Recurrent Neural Network (RNN) on low-cost multicore RISC-V microprocessor. Gravity field and steady-state Ocean Circulation Explorer (GOCE) public telemetry data was used for training RNNs with different hyperparameters and architectures. The prediction accuracy of these models was evaluated using mean error and R-squared score on the same test dataset. The implementation of the RNN on a RISC-V embedded device, representative of future space-grade hardware, required some adaptations and modifications due to the computational requirements and the large memory footprint. The algorithm was implemented to run in parallel on the 8 cores of the microprocessor and tiling was employed for the weight matrices. Further considerations have also been made for the approximation of sigmoid and hyperbolic tangent as activation functions.

AB - The aim of this paper is to assess the feasibility and on-board hardware performance requirements for on-board telemetry forecasting by implementing a Recurrent Neural Network (RNN) on low-cost multicore RISC-V microprocessor. Gravity field and steady-state Ocean Circulation Explorer (GOCE) public telemetry data was used for training RNNs with different hyperparameters and architectures. The prediction accuracy of these models was evaluated using mean error and R-squared score on the same test dataset. The implementation of the RNN on a RISC-V embedded device, representative of future space-grade hardware, required some adaptations and modifications due to the computational requirements and the large memory footprint. The algorithm was implemented to run in parallel on the 8 cores of the microprocessor and tiling was employed for the weight matrices. Further considerations have also been made for the approximation of sigmoid and hyperbolic tangent as activation functions.

KW - Deep Neural Networks

KW - RISC-V

KW - Space Systems

KW - Artificial Intelligence

UR - http://www.scopus.com/inward/record.url?scp=85097656469&partnerID=8YFLogxK

U2 - 10.1109/DFT50435.2020.9250796

DO - 10.1109/DFT50435.2020.9250796

M3 - Conference contribution

T3 - 33rd IEEE International Symposium on Defect and Fault Tolerance in VLSI and Nanotechnology Systems, DFT 2020

BT - 33rd IEEE International Symposium on Defect and Fault Tolerance in VLSI and Nanotechnology Systems, DFT 2020

A2 - Dilillo, Luigi

A2 - Psarakis, Mihalis

A2 - Siddiqua, Taniya

PB - Institute of Electrical and Electronics Engineers (IEEE)

T2 - 2020 33rd IEEE International Symposium on Defect and Fault Tolerance in VLSI and Nanotechnology Systems

Y2 - 19 October 2020 through 21 October 2020

ER -

Cappellone D, Di Mascio S, Furano G, Menicucci A, Ottavi M. On-Board Satellite Telemetry Forecasting with RNN on RISC-V Based Multicore Processor. In Dilillo L, Psarakis M, Siddiqua T, editors, 33rd IEEE International Symposium on Defect and Fault Tolerance in VLSI and Nanotechnology Systems, DFT 2020. Institute of Electrical and Electronics Engineers (IEEE). 2020. 9250796. (33rd IEEE International Symposium on Defect and Fault Tolerance in VLSI and Nanotechnology Systems, DFT 2020). doi: 10.1109/DFT50435.2020.9250796

On-Board Satellite Telemetry Forecasting with RNN on RISC-V Based Multicore Processor

Abstract

Publication series

Conference

Keywords

Access to Document

Other files and links

Fingerprint

Best Paper Award

Cite this

On-Board Satellite Telemetry Forecasting with RNN on RISC-V Based Multicore Processor

Abstract

Publication series

Conference

Keywords

Access to Document

Other files and links

Fingerprint

Prizes

Best Paper Award

Cite this