Eduardo Antonio Cesar da Costa
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Eduardo Antonio Cesar da Costa has a total of 47 co-authors. Here are the top co-authors by number of shared publications:
- Sergio Bampi
Graduate Program in Microelectronics (PGMICRO), Universidade Federal do Rio Grande do Sul, Porto Alegre, Brazil
21 shared publications - Guilherme Paim
Graduate Program in Microelectronics (PGMICRO), Institute of Informatics (INF), Universidade Federal do Rio Grande do Sul (UFRGS), Porto Alegre, Brazil
17 shared publications - Morgana Macedo Azevedo da Rosa
Universidade Católica de Pelotas (UCPel)
10 shared publications - Rafael Iankowski Soares
Universidade Federal de Pelotas (UFPel)
6 shared publications - Sérgio José Melo de Almeida
Graduate Program on Electronics Engineering and Computing, Catholic University of Pelotas, Pelotas, Brazil
6 shared publications - Brunno Alves de Abreu
Graduate Program in Microelectronics (PGMICRO), Institute of Informatics (INF), Universidade Federal do Rio Grande do Sul (UFRGS), Porto Alegre, Brazil
4 shared publications
Latest Publications
- 1 November 2026
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Circuits and Systems II: Express Briefs
- Vol. 73 (11) , 1174-1178
Approximate NTT/INTT Architectures Through FALSAx-Driven Design
This work presents a novel exploration of energy-efficient architectures for the number-theoretic transform (NTT) and its inverse (INTT), leveraging a suite of sixteen state-of-the-art approximate adders (AxAs). Integrating these arithmetic units into a constant-geometry NTT/INTT pipeline targets post-quantum image encryption and decryption. The design employs a leading-one detector (LOD)-based AxRLL-16 logarithmic unit for modular reduction and adequately approximates the twiddle factors. The FALSAx (framework for accuracy and logic synthesis estimation) guides the design space exploration, enabling accurate area, power, and signal fidelity prediction through machine learning models. Evaluation across 100 images demonstrates that selected AxA configurations achieve up to 38.6% area- and 45% energy-savings while maintaining high visual and structural quality. Visual assessments, accuracy metrics such as normalized cross-correlation (NCC), root mean square error (RMSE), mean absolute error (MAE), bit error rate (BER), and Pareto-optimal analyses confirm the viability of AxA-enhanced NTT/INTT for cryptographic systems.
- 1 November 2026
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Circuits and Systems II: Express Briefs
- Vol. 73 (11) , 1174-1178
Approximate NTT/INTT Architectures Through FALSAx-Driven Design
This work presents a novel exploration of energy-efficient architectures for the number-theoretic transform (NTT) and its inverse (INTT), leveraging a suite of sixteen state-of-the-art approximate adders (AxAs). Integrating these arithmetic units into a constant-geometry NTT/INTT pipeline targets post-quantum image encryption and decryption. The design employs a leading-one detector (LOD)-based AxRLL-16 logarithmic unit for modular reduction and adequately approximates the twiddle factors. The FALSAx (framework for accuracy and logic synthesis estimation) guides the design space exploration, enabling accurate area, power, and signal fidelity prediction through machine learning models. Evaluation across 100 images demonstrates that selected AxA configurations achieve up to 38.6% area- and 45% energy-savings while maintaining high visual and structural quality. Visual assessments, accuracy metrics such as normalized cross-correlation (NCC), root mean square error (RMSE), mean absolute error (MAE), bit error rate (BER), and Pareto-optimal analyses confirm the viability of AxA-enhanced NTT/INTT for cryptographic systems.
- 19 August 2025
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Design & Test
- Vol. 43 (2) , 26-35
Low-Power Truncated Pan–Tompkins Architecture Exploring Approximate Radix-2 m Squaring Units
Editor’s notes: This article introduces an energy-efficient hardware architecture for electrocardiogram QRS detection that integrates approximate Radix-2m squaring units into a truncated Pan–Tompkins algorithm. The proposed design achieves substantial reductions in area and power consumption while preserving detection accuracy, demonstrating the effectiveness of approximate computing for low-power biomedical VLSI systems.—José Augusto Nacif, Federal University of Viçosa, Brazil—Omar Paranaiba Vilela Neto, Federal University of Minas Gerais (UFMG), Brazil
- 19 August 2025
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Design & Test
- Vol. 43 (2) , 26-35
Low-Power Truncated Pan–Tompkins Architecture Exploring Approximate Radix-2 m Squaring Units
Editor’s notes: This article introduces an energy-efficient hardware architecture for electrocardiogram QRS detection that integrates approximate Radix-2m squaring units into a truncated Pan–Tompkins algorithm. The proposed design achieves substantial reductions in area and power consumption while preserving detection accuracy, demonstrating the effectiveness of approximate computing for low-power biomedical VLSI systems.—José Augusto Nacif, Federal University of Viçosa, Brazil—Omar Paranaiba Vilela Neto, Federal University of Minas Gerais (UFMG), Brazil
- 20 June 2025
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Latin America Transactions
- Vol. 23 (7) , 600-608
Local Volt-Var Control Applied in an Islanded Microgrid Using Supervised Learning Techniques
The electrical sector pursuit of technical and ecological alternatives makes it possible to integrate and cooperatively optimize dispersed energy resources, enhancing the stability, dependability, and resilience of contemporary energy systems. Microgrids and artificial intelligence are two ideas that could be included into contemporary power grids in an effort to lower costs and pollution emissions. This work proposes a new energy control and management strategy based on smart devices in this context. It explores machine-learning techniques for implementing supervised learning algorithms to perform automatic volt-var control adjustments and mitigate voltage fluctuations at the point of common coupling using smart inverters. The techniques explored and compared in this study include multilayer perceptron, SVM, and random forest. The results were consistent, with average accuracies above 90%, indicating the relevance of the analyzed models for this application. Thus, this research seeks to improve power quality in islanded microgrids with high penetration of distributed generation and explore the potential of artificial intelligence in decision-making processes.
- 20 June 2025
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Latin America Transactions
- Vol. 23 (7) , 600-608
Local Volt-Var Control Applied in an Islanded Microgrid Using Supervised Learning Techniques
The electrical sector pursuit of technical and ecological alternatives makes it possible to integrate and cooperatively optimize dispersed energy resources, enhancing the stability, dependability, and resilience of contemporary energy systems. Microgrids and artificial intelligence are two ideas that could be included into contemporary power grids in an effort to lower costs and pollution emissions. This work proposes a new energy control and management strategy based on smart devices in this context. It explores machine-learning techniques for implementing supervised learning algorithms to perform automatic volt-var control adjustments and mitigate voltage fluctuations at the point of common coupling using smart inverters. The techniques explored and compared in this study include multilayer perceptron, SVM, and random forest. The results were consistent, with average accuracies above 90%, indicating the relevance of the analyzed models for this application. Thus, this research seeks to improve power quality in islanded microgrids with high penetration of distributed generation and explore the potential of artificial intelligence in decision-making processes.
- Open Access
- 23 December 2024
- Published by Journal of Integrated Circuits and Systems in Journal of Integrated Circuits and Systems
- Vol. 19 (3) , 1-12
AxRMU-2^m: Higher m-bit Approximate Radix-2^m Multiplier Unit
This paper presents the design and evaluation of higher m-bit Approximate Radix-2^m Multiplier Units (AxRMU) with applications in error-tolerant systems. The AxRMU proposal architectures, specifically AxRMU-8 and AxRMU-16, incorporate approximation in generating partial products, reducing the number of logic gates and improving energy efficiency while maintaining acceptable accuracy. A case study using the statistically efficient and fast calibration (StEFCal) algorithm in radio astronomy demonstrates the performance of the AxRMUs in terms of energy, area savings, and computational accuracy. Results show that AxRMU designs outperform state-of-the-art approximate multipliers like DRUM, LoBA, and RoBA, achieving energy savings of up to 98.30% and area savings of up to 91.57%, with minimal mean square error (MSE) trade-offs. These findings highlight the potential of AxRMUs in balancing performance and resource efficiency for applications that can tolerate minor computational inaccuracies.
- 23 December 2024
- Published by Journal of Integrated Circuits and Systems in Journal of Integrated Circuits and Systems
- Vol. 19 (3) , 1-12
Open AccessAxRMU-2^m: Higher m-bit Approximate Radix-2^m Multiplier Unit
This paper presents the design and evaluation of higher m-bit Approximate Radix-2^m Multiplier Units (AxRMU) with applications in error-tolerant systems. The AxRMU proposal architectures, specifically AxRMU-8 and AxRMU-16, incorporate approximation in generating partial products, reducing the number of logic gates and improving energy efficiency while maintaining acceptable accuracy. A case study using the statistically efficient and fast calibration (StEFCal) algorithm in radio astronomy demonstrates the performance of the AxRMUs in terms of energy, area savings, and computational accuracy. Results show that AxRMU designs outperform state-of-the-art approximate multipliers like DRUM, LoBA, and RoBA, achieving energy savings of up to 98.30% and area savings of up to 91.57%, with minimal mean square error (MSE) trade-offs. These findings highlight the potential of AxRMUs in balancing performance and resource efficiency for applications that can tolerate minor computational inaccuracies.
- 19 December 2024
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Circuits and Systems I: Regular Papers
- Vol. 72 (4) , 1679-1692
FALSAx: An Integrated Framework for Accuracy and Logic Synthesis Estimation of Approximate Adders
This work proposes an integrated framework for accuracy and logic synthesis (LS) estimation of approximate adders (FALSAx). It represents a versatile and robust framework designed to estimate the accuracy, power, and area of various approximate adders (AxAs) for any input width ( $W$ ) and $K$ bits of approximation using machine learning (ML) models. FALSAx facilitates performance predictions and optimization for different AxAs configurations through meticulously curated datasets and ML-driven analysis. The framework’s capability to automatically generate Pareto fronts from estimated values aids in identifying optimal trade-offs among crucial metrics, providing essential insights for circuit design and optimization. The FALSAx includes four internal frameworks: FrAQ, PILSE, and FELSE, which estimates dynamic power, total leakage power, and area, with frequency variations automatically, and the FALED dataset of the FALSAx. As a case study, this work analyzed 16 types of AxAs on FALSAx: AMA-V, AxPPA, COPY, TRUNC, ETA, LOA, HOERAA, LDCA, LZTA, HEAA, M-HEAA, HERLOA, M-HERLOA, HOAANED, OLOCA, and SETA. The rigorous analysis provided by FALSAx revealed that HERLOA, M-HERLOA, M-HEAA, and AxPPA demonstrated superior accuracy metrics such as SSIM, NCC, MAE, and MRE. Furthermore, power analysis showed that AxPPA exhibited the best power efficiency for lower approximation bits ( $K \leq 3$ ). At the same time, gate-free adders like COPY, TRUNC, AMA-V, LDCA, and LZTA were more power-efficient for higher approximation bits ( $K > 3$ ). Area estimations indicated that AxPPA maintained competitive efficiency for lower approximation bits ( $K \leq 5$ ), while TRUNC and LDCA were more efficient for higher bits ( $K > 5$ ).
- 19 December 2024
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Circuits and Systems I: Regular Papers
- Vol. 72 (4) , 1679-1692
FALSAx: An Integrated Framework for Accuracy and Logic Synthesis Estimation of Approximate Adders
This work proposes an integrated framework for accuracy and logic synthesis (LS) estimation of approximate adders (FALSAx). It represents a versatile and robust framework designed to estimate the accuracy, power, and area of various approximate adders (AxAs) for any input width ( $W$ ) and $K$ bits of approximation using machine learning (ML) models. FALSAx facilitates performance predictions and optimization for different AxAs configurations through meticulously curated datasets and ML-driven analysis. The framework’s capability to automatically generate Pareto fronts from estimated values aids in identifying optimal trade-offs among crucial metrics, providing essential insights for circuit design and optimization. The FALSAx includes four internal frameworks: FrAQ, PILSE, and FELSE, which estimates dynamic power, total leakage power, and area, with frequency variations automatically, and the FALED dataset of the FALSAx. As a case study, this work analyzed 16 types of AxAs on FALSAx: AMA-V, AxPPA, COPY, TRUNC, ETA, LOA, HOERAA, LDCA, LZTA, HEAA, M-HEAA, HERLOA, M-HERLOA, HOAANED, OLOCA, and SETA. The rigorous analysis provided by FALSAx revealed that HERLOA, M-HERLOA, M-HEAA, and AxPPA demonstrated superior accuracy metrics such as SSIM, NCC, MAE, and MRE. Furthermore, power analysis showed that AxPPA exhibited the best power efficiency for lower approximation bits ( $K \leq 3$ ). At the same time, gate-free adders like COPY, TRUNC, AMA-V, LDCA, and LZTA were more power-efficient for higher approximation bits ( $K > 3$ ). Area estimations indicated that AxPPA maintained competitive efficiency for lower approximation bits ( $K \leq 5$ ), while TRUNC and LDCA were more efficient for higher bits ( $K > 5$ ).
- 8 November 2024
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Very Large Scale Integration (VLSI) Systems
- Vol. 33 (3) , 733-745
RCU- 2 m : A VLSI Radix- 2 m Cubic Unit
Cubic operations are among the most used arithmetic operations in many applications that demand higher order simultaneous operand computation, such as cryptography and bicubic polynomial interpolation. This article proposes a novel VLSI radix-$2^{m}$ cubic unit (RCU-$2^{m}$ ) capable of processing cubic operations at m bits simultaneously, with m values of 2 (RCU-4), 3 (RCU-8), and 4 (RCU-16). RCU-16 emerges as the most area-efficient configuration, surpassing RCU-8 and notably outperforming RCU-4. In the 8-bit scenario, RCU-16 achieves remarkable area savings, surpassing the literature’s proposed cubic unit by$11.58\times $ . Across all configurations, RCU-$2^{m}$ consistently outperforms the automatically selected cube unit, with energy savings ranging from$1.04\times $ to$2\times $ . In application specific integrated circuit (ASIC) and field-programmable gate array (FPGA)-based analyses, RCU-16 consistently exhibits superior performance in both area and energy savings compared with RCU-4, RCU-8, and solutions from the literature. These findings emphasize the importance of adopting radix-$2^{m}$ configurations, particularly RCU-16, for optimal energy-constrained VLSI applications.- 8 November 2024
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Very Large Scale Integration (VLSI) Systems
- Vol. 33 (3) , 733-745
RCU- 2 m : A VLSI Radix- 2 m Cubic Unit
Cubic operations are among the most used arithmetic operations in many applications that demand higher order simultaneous operand computation, such as cryptography and bicubic polynomial interpolation. This article proposes a novel VLSI radix-$2^{m}$ cubic unit (RCU-$2^{m}$ ) capable of processing cubic operations at m bits simultaneously, with m values of 2 (RCU-4), 3 (RCU-8), and 4 (RCU-16). RCU-16 emerges as the most area-efficient configuration, surpassing RCU-8 and notably outperforming RCU-4. In the 8-bit scenario, RCU-16 achieves remarkable area savings, surpassing the literature’s proposed cubic unit by$11.58\times $ . Across all configurations, RCU-$2^{m}$ consistently outperforms the automatically selected cube unit, with energy savings ranging from$1.04\times $ to$2\times $ . In application specific integrated circuit (ASIC) and field-programmable gate array (FPGA)-based analyses, RCU-16 consistently exhibits superior performance in both area and energy savings compared with RCU-4, RCU-8, and solutions from the literature. These findings emphasize the importance of adopting radix-$2^{m}$ configurations, particularly RCU-16, for optimal energy-constrained VLSI applications.- 4 September 2024
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Very Large Scale Integration (VLSI) Systems
- Vol. 33 (2) , 558-562
ReAdapt-II: Energy-Quality Optimizations for VLSI Adaptive Filters Through Automatic Reconfiguration and Built-In Iterative Dividers
Adaptive filters using least mean square (LMS) algorithms offer high precision, low complexity, and fast convergence, but choosing the correct algorithm can be difficult and time-consuming. In this brief, we present ReAdapt-II, a VLSI circuit that enhances energy efficiency in adaptive filters through automatic reconfiguration and built-in iterative dividers, optimizing the energy-quality (EQ) tradeoff. This design features a self-selecting, reconfigurable hardware system with four adaptive algorithms, integrating iterative-based dividers and reusing arithmetic operators. Our results show a minimum energy consumption reduction of 39.75%, a 66.61% reduction in the circuit area, and a maximum accuracy increase of 17.07% compared with the previous ReAdapt architecture.
- 4 September 2024
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Very Large Scale Integration (VLSI) Systems
- Vol. 33 (2) , 558-562
ReAdapt-II: Energy-Quality Optimizations for VLSI Adaptive Filters Through Automatic Reconfiguration and Built-In Iterative Dividers
Adaptive filters using least mean square (LMS) algorithms offer high precision, low complexity, and fast convergence, but choosing the correct algorithm can be difficult and time-consuming. In this brief, we present ReAdapt-II, a VLSI circuit that enhances energy efficiency in adaptive filters through automatic reconfiguration and built-in iterative dividers, optimizing the energy-quality (EQ) tradeoff. This design features a self-selecting, reconfigurable hardware system with four adaptive algorithms, integrating iterative-based dividers and reusing arithmetic operators. Our results show a minimum energy consumption reduction of 39.75%, a 66.61% reduction in the circuit area, and a maximum accuracy increase of 17.07% compared with the previous ReAdapt architecture.
- 13 August 2024
- Published by Springer Nature in Circuits, Systems, and Signal Processing
- Vol. 43 (12) , 7750-7780
Exploring Discrete Haar Wavelet and Cosine Transforms for Accuracy-and Energy-Quality VLSI Watermarking Systems Design
- 13 August 2024
- Published by Springer Nature in Circuits, Systems, and Signal Processing
- Vol. 43 (12) , 7750-7780
Exploring Discrete Haar Wavelet and Cosine Transforms for Accuracy-and Energy-Quality VLSI Watermarking Systems Design
- 26 February 2024
- Published by Springer Nature in Circuits, Systems, and Signal Processing
- Vol. 43 (6) , 3649-3678
AxRSU-$2^{m}$: Higher-Order m-Bit Approximate Encoders for Radix-$2^{m}$ Squarer Units- 26 February 2024
- Published by Springer Nature in Circuits, Systems, and Signal Processing
- Vol. 43 (6) , 3649-3678
AxRSU-$2^{m}$: Higher-Order m-Bit Approximate Encoders for Radix-$2^{m}$ Squarer Units- 21 November 2023
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Circuits and Systems I: Regular Papers
- Vol. 71 (3) , 1000-1013
VLSI Architectures of Approximate Arithmetic Units Applied to Parallel Sensors Calibration
Approximate computing maximizes area and energy savings for a trade-off between quality and efficiency. Approximate arithmetic operators have emerged as an efficient alternative to design low-power VLSI circuits. This paper investigates the design of approximate arithmetic operator units used in the calibration procedure for radio astronomy light sensors — the so-called StEFCal (statistically efficient and fast calibration) method. The StEFCal algorithm comprises arithmetic operations like a divider, square-accumulate (SAC), and multiply-accumulate (MAC) units. The StEFCal circuit of this work explores the following arithmetic operators: i) two approximate squarer units from the literature, i.e., radix-4 (AxRSU) and SquASH, ii) two approximate iterative-based Newton-Raphson (NR) and Goldschmidt (GLD) dividers, iii) one approximate parallel prefix adder (AxPPA), and iv) a new approximate radix-4 multiplier (AxRMU), proposed in this work, explored in the StEFCal multiply-accumulate circuit design. The AxRSU utilizes the parameters $K1$ and $K2$ to represent the number of exact encoders for squarer-and conventional-partial products, respectively, subsequently replaced with approximate encoders. The same principle applies to AxRMU, where the parameter $K$ indicates the number of exact encoders for conventional-partial products, subsequently exchanged with approximate encoders. We demonstrate the efficiency of StEFCal using the approximate arithmetic operators from the Pareto-optimal front that expresses the area-and power-quality trade-off. The results show that using the AxRSU with $K1=4$ and $K2=6$ , AxRMU, and AxPPA with $K=16$ and NR with one iteration has an MSE equal to $89.98$ dB and offers up to $158\times$ energy-savings compared to the exact StEFCal, and up to $25\times$ more energy-savings and $3.33\times$ area-savings compared with our previous work, $440\times$ energy-savings compared to the accurate state-of-the-art, and $258\times$ compared with the approximate state-of-the-art.
- 21 November 2023
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Circuits and Systems I: Regular Papers
- Vol. 71 (3) , 1000-1013
VLSI Architectures of Approximate Arithmetic Units Applied to Parallel Sensors Calibration
Approximate computing maximizes area and energy savings for a trade-off between quality and efficiency. Approximate arithmetic operators have emerged as an efficient alternative to design low-power VLSI circuits. This paper investigates the design of approximate arithmetic operator units used in the calibration procedure for radio astronomy light sensors — the so-called StEFCal (statistically efficient and fast calibration) method. The StEFCal algorithm comprises arithmetic operations like a divider, square-accumulate (SAC), and multiply-accumulate (MAC) units. The StEFCal circuit of this work explores the following arithmetic operators: i) two approximate squarer units from the literature, i.e., radix-4 (AxRSU) and SquASH, ii) two approximate iterative-based Newton-Raphson (NR) and Goldschmidt (GLD) dividers, iii) one approximate parallel prefix adder (AxPPA), and iv) a new approximate radix-4 multiplier (AxRMU), proposed in this work, explored in the StEFCal multiply-accumulate circuit design. The AxRSU utilizes the parameters $K1$ and $K2$ to represent the number of exact encoders for squarer-and conventional-partial products, respectively, subsequently replaced with approximate encoders. The same principle applies to AxRMU, where the parameter $K$ indicates the number of exact encoders for conventional-partial products, subsequently exchanged with approximate encoders. We demonstrate the efficiency of StEFCal using the approximate arithmetic operators from the Pareto-optimal front that expresses the area-and power-quality trade-off. The results show that using the AxRSU with $K1=4$ and $K2=6$ , AxRMU, and AxPPA with $K=16$ and NR with one iteration has an MSE equal to $89.98$ dB and offers up to $158\times$ energy-savings compared to the exact StEFCal, and up to $25\times$ more energy-savings and $3.33\times$ area-savings compared with our previous work, $440\times$ energy-savings compared to the accurate state-of-the-art, and $258\times$ compared with the approximate state-of-the-art.