Departamento de Física
URI permanente desta comunidadehttps://repositorio.fei.edu.br/handle/FEI/785
Navegar
11 resultados
Resultados da Pesquisa
- Evaluating the Reliability of Different Voting Schemes for Fault Tolerant Approximate Systems(2023-06-20) BALEN, T. R.; GONZALEZ, C. J.; OLIVEIRA, I. F. V.; DA ROSA JR. L. S.; SOARES, R. I.; SCHVITTZ, R. B.; ADDED, N.; MACCHIONE, E. L. A.; AGUIAR, V. A. P.; Marcilei Aparecida Guazzelli; MEDINA, N. H.; BUTZEN, P. F.© 2023, The Author(s), under exclusive licence to Springer Science+Business Media, LLC, part of Springer Nature.This work presents a study on the reliability of voters for approximate fault tolerant systems in the context of single event effects and electromagnetic interference. A first case study analyses different topologies of single-bit majority voters for logic circuits employing fault injection by simulation. In these simulations, an analysis is first performed to identify the critical diffusion areas of the physical implementation according to the voter input vector. Additionally, as a second case study, practical heavy ion experiments on different architectures of software-based approximate voters for mixed-signal applications are also presented, and the cross section of each voter is evaluated. The system comprising the voters was irradiated in two distinct experiments with an 16O ion beam, producing an effective LET at the active region of 5.5 MeV/mg/cm 2 . As a complementary study, a conducted electromagnetic interference injection was also performed, considering two distinct voting schemes. Results of the case-studies allow identifying the most tolerant voter architectures (among the studied ones) for approximate computing applications under single event effects and electromagnetic interference.
- Unsupervised machine learning application to identify single-event transients (SETs) from noise events in MOSFET transistor ionizing radiation effects(2023) ALLEGRO, P. R. P.; TOUFEN, D. L.; AGUIAR, V. A. P.; SANTOS, L. S. A.; OLIVEIRA W. N.; ADDED, N.; MEDINA, N. H.; MACCHIONE, E. L. A.; ALBERTON , S. G.; Marcilei Aparecida Guazzelli; MELO, M. A. A.; OLIVEIRA, J. A.© 2023This article presents a novel application of the k-means unsupervised machine learning algorithm to the problem of identifying single event transient (SET) events from noise during heavy-ion irradiation experiments of an electronic device. We explore the performance of the k-means algorithm by analyzing experimental datasets of SET events produced by several heavy-ions irradiations of a MOSFET transistor. Data anomalies and effectiveness of the chosen features (mean, standard deviation, skewness, and kurtosis) were investigated using the Isolation Forest and Random Forest algorithms, respectively. The results show a high capability of the K-means algorithm to identify SET events from noise using the first four statistical moments as features, allowing in the future the use of this method for in situ event detection and diagnosis without previous algorithm training or pre-analysis of the experimental data.
- Heavy-Ion-Induced Avalanche Multiplication in Low-Voltage Power VDMosfet(2021-09-13) ALBERTON. S. G.; MEDINA, N. H.; ADDED, N.; AGUIAR, V. A. P.; Marcilei Aparecida Guazzelli; Roberto Santos© 2021 IEEE.Lackner's theory for avalanche multiplication in semiconductor devices provides physical interpretation for the model parameters and obtaining them through experimental methods is necessary. In this work, the charge collection mechanisms of heavy-ion-induced avalanche multiplication in power MOSFETs are studied based on Lackner's impact ionization model. The heavy-ion-induced impact ionization coefficients were estimated by comparing collected charge values obtained from computational simulations and experimental measurements.
- Isomeric state in the odd-odd 68Ga nucleus(2022-01-05) ESCUDEIRO, R.; VASCONCELOS, C. E. C; ALLEGRO, P. R. P.; MEDINA, N. H.; TUFEN, D. L.; ADDED, N.; AGUIAR, V. A. P.; ALBERTON. S. G. P. N.; ALCANTARA-NUNEZ, J.; Marcilei Aparecida Guazzelli; MACCHIONE. E. L. A.; OLIVEIRA, J. R. B.; RIBAS, R. V.; SCARDUELLI, V. B.© 2022 Institute of Physics Publishing. All rights reserved.The half life of the 7− isomeric state of the odd-odd 68Ga nucleus was measured using a particle-γ delayed coincidence technique. The 68Ga nuclei were produced using the fusion-evaporation reaction 55Mn(16O, 2pn)68Ga at 55 MeV incident beam energy. The beam was produced by the 8 MV Pelletron accelerator of the Nuclear Physics Open Laboratory of the University of São Paulo. The half life was measured using the Isomeric State Measurement System (SISMEI). The obtained value was 60.83(25) ns, compatible with previous measurements. The 68Ga excited states were well described with the Large Scale Shell Model using the JUN45 residual interaction.
- Charge deposition analysis of heavy-ion-induced single-event burnout in low-voltage power VDMOSFET(2022-10-05) ALBERTON, S. G.; AGUIAR, V. A. P.; MEDINA, N. H.; ADDED, N.; MACCHIONE, E. L. A.; MENEGASSO, R.; CESARIO. G. J.; SANTOS, H. C.; SCARDUELLI, V. B.; ALCANTARA-NUNEZ, J. A.; Marcilei Aparecida Guazzelli; Roberto Santos; FLECHAS, D.© 2022 Elsevier LtdThe heavy-ion-induced single-event burnout (SEB) risk in power MOSFETs (metal-oxide-semiconductor field-effect transistors) can be assessed in ground facilities, although it is costly and time-demanding. For this reason, there have been few experimental studies dedicated to investigate the relevant parameter related to the description of ion-induced SEB phenomenon. In this work the heavy-ion-induced SEB in a low-voltage power VDMOSFET (vertical double-diffused MOSFET) is studied using several ion-energy combinations. A self-consistent statistical analysis is carried out in order to elucidate the relationship between charge deposition and SEB triggering. Experimental data is compared to a predictive model from the literature for SEE (single-event effect) worst-case prediction in power MOSFETs, supporting for the first time its relevance to the worst-case prediction in the SEB mechanism.
- Testing a Fault Tolerant Mixed-Signal Design Under TID and Heavy Ions(2021-01-05) GONZALEZ, C. J.; MACHADO, D. N.; VAZ, R. G.; VILAS BOAS, A. C.; GONLALEZ, O. L.; PUCHNER, H.; ADDED, N.; MACCHIONE, E. L. A.; AGUIAR, V. A. P.; KASTENSMIDT, F. L.; MEDINA, N. H.; Marcilei Aparecida Guazzelli; BALEN, T. R.© 2021, Brazilian Microelectronics Society. All rights reserved.— This work presents results of three distinct radiation tests performed upon a fault tolerant data acquisition system comprising a design diversity redundancy technique. The first and second experiments are Total Ionizing Dose (TID) essays, comprising gamma and X-ray irradiations. The last experiment considers single event effects, in which two heavy ion irradiation campaigns are carried out. The case study system comprises three analog-to-digital converters and two software-based vot-ers, besides additional software and hardware resources used for controlling, monitoring and memory manage-ment. The applied Diversity Triple Modular Redundancy (DTMR) technique, comprises different levels of diversity (temporal and architectural). The circuit was designed in a programmable System-on-Chip (PSoC), fabricated in a 130nm CMOS technology process. Results show that the technique may increase the lifetime of the system under TID if comparing with a non-redundant implementation. Considering the heavy ions experiments the system was proved effective to tolerate 100% of the observed errors originated in the converters, while errors in the processing unit present a higher criticality. Critical errors occur-ring in one of the voters were also observed. A second heavy ion campaign was then carried out to investigate the voters reliability, comparing the the dynamic cross section of three different software-based voter schemes im-plemented in the considered PSoC.
- Reducing Soft Error Rate of SoCs Analog-to-Digital Interfaces with Design Diversity Redundancy(2020-03-05) GONZALEZ, C. J.; ADDED, N.; MACCHIONE, E. L. A.; AGUIAR, V. A. P.; KASTENSMIDT, F. G. L.; PUCHNER, H. K.; Marcilei Aparecida Guazzelli; MEDINA, N. H.; BALEN, T. R.© 1963-2012 IEEE.In this article, a commercial programmable system-on-chip (PSoC 5, from Cypress Semiconductor) is tested under heavy-ion irradiation with a focus on the analog-to-digital interface blocks of the system. For this purpose, a data acquisition system (DAS) was programmed into the device under test and protected with a design diversity redundancy technique. This technique implements different levels of diversity (architectural and temporal) by using two different architectures of converters (a Σ Δ converter and two successive approximation register (SAR) converters) operating with distinct sampling rates. The experiment was performed in a vacuum chamber, using a 16O ion beam with 36-MeV energy and sufficient penetration into the silicon to produce an effective linear energy transfer (LET) of 5.5 MeV/mg/cm2 at the active region. The average flux was approximately 350 particles/s/cm2 for 246 min. The individual susceptibility of each converter to single-event effects is evaluated, as well as the whole system cross section. Results show that the proposed technique is effective to mitigate errors originating at the converters since 100% of such errors were corrected by using the diversity redundancy technique. Results also show that the processing unit of the system is susceptible to hangs that can be mitigated using watchdog techniques.
- Evaluating Soft Core RISC-V Processor in SRAM-Based FPGA under Radiation Effects(2020-07-05) OLIVEIRA, A. B.; TAMBARA, L. A.; BENEVENUTI, F.; BENITES, L. A. C.; ADDED, N.; AGUIAR, V. A. P.; MEDINA, N. H.; Marcilei Aparecida Guazzelli; KASTENSMIDT, F. L.© 1963-2012 IEEE.This article evaluates the RISC-V Rocket processor embedded in a Commercial Off-The-Shelf (COTS) SRAM-based field-programmable gate array (FPGA) under heavy-ions-induced faults and emulation fault injection. We also analyze the efficiency of using mitigation techniques based on hardware redundancy and scrubbing. Results demonstrated an improvement of $3\times $ in the cross section when scrubbing and coarse grain triple modular redundancy are used. The Rocket processor presented analogous sensitivity to radiation effects as the state-of-the-art soft processors. Due to the complexity of the system-on-chip, not only the Rocket core but also its peripherals should be protected with proper solutions. Such solutions should address the specific vulnerabilities of each component to improve the overall system reliability while maintaining the trade-off with performance.
- Determination of stable isotope ratios using nuclear reaction analysis coupled with a particle-gamma coincidence method(2021-01-05) ALLEGRO, P.R.P.; RIZZUTTO, M.D.A.; ADDED, N.; AGUIAR, V.A.P.D.; TOUFEN, D.L.; MACCHIONE, E.L.A.; MEDINA, N.H.; RIBAS, R.V.; OLIVEIRA, J.R.B.D.; ESCUDEIRO, R.; ALCANTARA-NUNEZ, J.A.; CURADO, J.F.; SCARDUELLI, V.B.; ZAGATTO, V.A.B.; RODRIGUES, C.L.; TABACNIKS, M.H.© The Royal Society of Chemistry.A non-destructive standard-free analytical method of obtaining stable isotope ratios in materials using nuclear reactions excited by an 11.75 MeV proton beam is demonstrated for a special 63Cu/65Cu target characterized traceably by Rutherford backscattering spectrometry (using 2.2 MeV protons). The method is general and is capable of being a useful alternative to standard ICP-MS methods. Relatively high energy proton beams are required to excite multiple isotopes of most elements of interest, and the many resulting nuclear reaction products are filtered by coincidence methods to characterize the desired isotope ratios. Specifically: the use of particle-gamma coincidences permits identification of specific channels populated in the nuclear reaction and avoids or minimizes typical drawbacks of this method including high backgrounds and interfering reactions. The test Cu target was a sandwich of a natural Cu and an enriched 65Cu foil, and particle-gamma coincidences were measured for the reactions 63Cu(p,p?){63Cu,60Ni} and the corresponding reactions for 65Cu (where "p"represents the energetic charged particle reaction product: proton or alpha). Limitations of the method, and applications in cultural heritage and other fields are discussed.
- Robust convolutional neural networks in sram-based fpgas: A case study in image classification(2021-08-23) BENEVENUTI, F.; KASTENSMIDT, F.; OLIVEIRA, A.; ADDED, N.; AGUIAR, V.; MEDINA, N.; Marcilei Aparecida Guazzelli© 2021, Brazilian Microelectronics Society. All rights reserved.— This work discusses the main aspects of vulnerability and degradation of accuracy of an image classification engine implemented into SRAM-based FPGAs under faults. The image classification engine is an all-convolutional neural-network (CNN) trained with a dataset of traffic sign recognition benchmark. The Caffe and Ristretto frameworks were used for CNN training and fine-tuning while the ZynqNet inference engine was adopted as hardware implementation on a Xilinx 28 nm SRAM-based FPGA. The CNN under test was generated using an evolutive approach based on genetic algorithm. The methodologies for qualifying this CNN under faults is presented and both heavy-ions accelerated irradiation and emulated fault injection were performed. To cross validate results from radiation and fault injection, different implementations of the same CNN were tested using reduced arithmetic precision and protection of user data by Hamming codes, in combination with configuration memory healing by the scrubbing mechanism available in Xilinx FPGA. Some of these alternative implementations increased significantly the mission time of the CNN, when compared to the original ZynqNet operating on 32 bits floating point number, and the experiment suggests areas for further improvements on the fault injection methodology in use.