Using machine learning for anomaly detection on a system-on-chip under gamma radiation
2
조회
0
다운로드
0
인용
0
참고문헌
키워드
초록
The emergence of new nanoscale technologies has imposed significant challenges to designing reliable electronic systems in radiation environments. A few types of radiation like Total Ionizing Dose (TID) can cause permanent damages on such nanoscale electronic devices, and current state-of-the-art technologies to tackle TID make use of expensive radiation-hardened devices. This paper focuses on a novel and different approach: using machine learning algorithms on consumer electronic level Field Programmable Gate Arrays (FPGAs) to tackle TID effects and monitor them to replace before they stop working. This condition has a research challenge to anticipate when the board results in a total failure due to TID effects. We observed internal measurements of FPGA boards under gamma radiation and used three different anomaly detection machine learning (ML) algorithms to detect anomalies in the sensor measurements in a gamma-radiated environment. The statistical results show a highly significant relationship between the gamma radiation exposure levels and the board measurements. Moreover, our anomaly detection results have shown that a One-Class SVM with Radial Basis Function Kernel has an average recall score of 0.95. Also, all anomalies can be detected before the boards are entirely inoperative, i.e. voltages drop to zero and confirmed with a sanity check.
Abstract
The emergence of new nanoscale technologies has imposed significant challenges to designing reliable electronic systems in radiation environments. A few types of radiation like Total Ionizing Dose (TID) can cause permanent damages on such nanoscale electronic devices, and current state-of-the-art technologies to tackle TID make use of expensive radiation-hardened devices. This paper focuses on a novel and different approach: using machine learning algorithms on consumer electronic level Field Programmable Gate Arrays (FPGAs) to tackle TID effects and monitor them to replace before they stop working. This condition has a research challenge to anticipate when the board results in a total failure due to TID effects. We observed internal measurements of FPGA boards under gamma radiation and used three different anomaly detection machine learning (ML) algorithms to detect anomalies in the sensor measurements in a gamma-radiated environment. The statistical results show a highly significant relationship between the gamma radiation exposure levels and the board measurements. Moreover, our anomaly detection results have shown that a One-Class SVM with Radial Basis Function Kernel has an average recall score of 0.95. Also, all anomalies can be detected before the boards are entirely inoperative, i.e. voltages drop to zero and confirmed with a sanity check.
참고문헌 (43)
- "Nuclear Provision: explaining the cost of cleaning up Britain's nuclear legacy"[URL]
- O. Gonçalves, "Radiation hardened mram-based fpga", IEEE Transactions on Magnetics, 49(7), pp.4355~4358 (2013)
- S. Ramaswamy, "A radiation hardened reconfigurable fpga", 2009 IEEE Aerospace Conference, pp.1~10 (2009)
- A. Lindoso, "A hybrid fault-tolerant LEON3 soft core processor implemented in low-end SRAM FPGA", IEEE Transactions on Nuclear Science, 64(1), pp.374~381 (2017)
- H. Quinn, "Using benchmarks for radiation testing of microprocessors and FPGAs", IEEE Transactions on Nuclear Science, 62(6), pp.2547~2554 (2015)
- S. M. Trimberger, "Three ages of fpgas: a retrospective on the first thirty years of fpga technology", Proceedings of the IEEE, 103(3), pp.318~331 (2015)
- Tomas Fried, "Radiation testing of low cost, commercial off the shelf microcontroller board", Nuclear Engineering and Technology, 53(10), pp.3335~3343 (2021), 한국원자력학회[DOI]
- Andrew West, "Radiation tolerance of a small COTS single board computer for mobile robots", Nuclear Engineering and Technology, 54(6), pp.2198~2203 (2022), 한국원자력학회[DOI]
- Joerg Leukel, "Adoption of machine learning technology for failure prediction in industrial maintenance : a systematic review", Journal of Manufacturing Systems, 61, pp.87~96 (2021)
- S. Kasap, "Survey of soft error mitigation techniques applied to LEON3 soft processors on SRAMbased FPGAs", IEEE Access, 8, pp.28646~28658 (2020)
- Yufan Lu, "A self-adaptive seu mitigation scheme for embedded systems in extreme radiation environments", IEEE Systems Journal, pp.1~12 (2022)
- Carl Carmichael, "Triple Module Redundancy Design Techniques for Virtex FPGAs" (2006), Xilinx Inc.
- D. K. Pradhan, "Roll-forward and rollback recovery: performancereliability trade-off", IEEE Transactions on Computers, 46(3), pp.372~378 (1997)
- L. Parra, "A new hybrid nonintrusive error-detection technique using dual control-flow monitoring", IEEE Transactions on Nuclear Science, 61(6) (2014)
- H. Quinn, "Robust duplication with comparison methods in microcontrollers", IEEE Transactions on Nuclear Science, 64(1), pp.338~345 (2017)
- Kuang-Hua Huang, "Algorithm-based fault tolerance for matrix operations", IEEE Transactions on Computers, 33(6), pp.518~528 (1984)
- J. R. Azambuja, "HETA: hybrid errordetection technique using assertions", IEEE Transactions on Nuclear Science (2013)
- E. Chielle, "S-SETA: selective software-only error-detection technique using assertions", IEEE Transactions on Nuclear Science, 62(6), pp.3088~3095 (2015)
- Harn Hua Ng, "PPC405 Lockstep System on ML310" (2007), Xilinx Inc
- F. Abate, "techniques for improving the performance of the lockstep architecture for SEEs mitigation in FPGA embedded processors", IEEE Transactions on Nuclear Science, 56(4), pp.1992~2000 (2009)
- M. Violante, "A low-cost solution for deploying processor cores in harsh environments", IEEE Transactions on Industrial Electronics, 58(7), pp.2617~2626 (2011)
- H. Pham, "Low-overhead fault-tolerance technique for a dynamically reconfigurable softcore processor", IEEE Transactions on Computers, 62(6) (2013)
- A. B. de Oliveira, "Lockstep dual-core ARM A9:implementation and resilience analysis under heavy ion-induced soft errors", IEEE Transactions on Nuclear Science, 65(8), pp.1783~1790 (2018)
- E.W. W€achter, "Survey of lockstep based mitigation techniques for soft errors in embedded systems", Computer Science and Electronic Engineering Conference (CEEC 2019), pp.124~127 (2019)
- N. Rezzak, "Investigation of tid and dynamic burn-in-induced vT shift on rtg4 flash-based fpga", IEEE Transactions on Nuclear Science, 65(1), pp.64~70 (2018)
- P. Villa, "Analysis of cots fpga seusensitivity to combined effects of conducted-emi and tid", 2017 11th International Workshop on the Electromagnetic Compatibility of Integrated Circuits (EMCCompo), pp.27~32 (2017)
- J. Tarrillo, "Analyzing the effects of tid in an embedded system running in a flash-based fpga", IEEE Transactions on Nuclear Science, 58(6), pp.2855~2862 (2011)
- G. Lentaris, "Tid evaluation system with on-chip electron source and programmable sensing mechanisms on fpga", IEEE Transactions on Nuclear Science, 66(1), pp.312~319 (2019)
- "Minized board product brief"[URL]
- "Zynq-7000 SoC data sheet"[URL]
- "DA9062 data sheet"[URL]
- "Elegoo R3 board"[URL]
- "ATmega data sheet"[URL]
- "PmodTC1 board reference manual"[URL]
- J. Cohen, "Statistical Power Analysis for the Behavioral Sciences" (1977), Lawrence Erlbaum Associates
- Ben Hoyle, "Anomaly detection for machine learning redshifts applied to SDSS galaxies", Monthly Notices of the Royal Astronomical Society, 452(4), pp.4183~4194 (2015)
- Malak Alshawabkeh, "Accelerating the local outlier factor Algorithm on a GPU for intrusion detection systems", Proceedings of the 3rd Workshop on General-Purpose Computation on Graphics Processing Units, pp.104~110 (2010), Association for Computing Machinery
- Katherine A. Heller, "One class support vector machines for detecting anomalous windows registry accesses", Proceedings of the ICDM Workshop on Data Mining for Computer Security (2003)
- Radial Basis Functions, Multi-Variable Functional Interpolation and Adaptive Networks, "David S. Broomhead", Royal Signals and Radar Establishment Malvern, ~ (1988)
- D. M. W. Powers, "Evaluation: From precision, recall and f-measure toroc, informedness, markedness and correlation", Journal of Machine Learning Technologies, 2(1), pp.37~63 (2011)
- "Dalton Cumbrian facility"[URL]
- Zeba Khanam, "Gamma-induced image degradation analysis of robot vision sensor for autonomous inspection of nuclear sites", IEEE Sensors Journal, 2021, pp.3050168~ (2021)[DOI]