https://scholar.google.com/citations?hl=en&user=ibqBDS4AAAAJ Shawkey

Prof. Dr. https://scholar.google.com/citations?hl=en&user=ibqBDS4AAAAJ Shawkey

Microelectronics department had - Electronics Research Institute
orcid

Research Focus & Metrics

Academic Career Length

19Years (2008-Present)

(Calculated once Scilit publication data is synced: current year − year of first publication + 1)

Representative Publications

View All Publications
16 March 2023 in IEEE Access
Verified
Journal Article
Open access
RSRatshih Sayed
Ratshih Sayed
HAHaytham Azmi
Haytham Azmi
HShttps://scholar.google.com/citations?hl=en&user=ibqBDS4AAAAJ Shawkey
https://scholar.google.com/citations?hl=en&user=ibqBDS4AAAAJ Shawkey
AKA. H. Khalil
A. H. Khalil
MRMohamed Refky
Mohamed Refky
This paper presents an extensive literature review on Binary Neural Network (BNN). BNN utilizes binary weights and activation function parameters to substitute the full-precision values. In digital implementations, BNN replaces the complex calculations of Convolutional Neural Networks (CNNs) with simple bitwise operations. BNN optimizes large computation and memory storage requirements, which leads to less area and power consumption compared to full-precision models. Although there are many advantages of BNN, the binarization process has a significant impact on the performance and accuracy of the generated models. To reflect the state-of-the-art in BNN and explore how to develop and improve BNN-based models, we conduct a systematic literature review on BNN with data extracted from 239 research studies. Our review discusses various BNN architectures and the optimization approaches developed to improve their performance. There are three main research directions in BNN: accuracy optimization, compression optimization, and acceleration optimization. The accuracy optimization approaches include quantization error reduction, special regularization, gradient error minimization, and network structure. The compression optimization approaches combine fractional BNN and pruning. The acceleration optimization approaches comprise computing in-memory, FPGA-based implementations, and ASIC-based implementations. At the end of our review, we present a comprehensive analysis of BNN applications and their evaluation metrics. Also, we shed some light on the most common BNN challenges and the future research trends of BNN.

27 November 2017 in IEEE Transactions on Circuits and Systems I: Regular Papers
Verified
Journal Article
Closed access
Mohamed AliMohamed Ali
Mohamed Ali
HShttps://scholar.google.com/citations?hl=en&user=ibqBDS4AAAAJ Shawkey
https://scholar.google.com/citations?hl=en&user=ibqBDS4AAAAJ Shawkey
Abdelhalim ZekryAbdelhalim Zekry
Abdelhalim Zekry
Mohamad SawanMohamad Sawan
Mohamad Sawan
This paper introduces a fully integrated frequency-modulated ultra-wideband (UWB) transmitter implemented in 130 nm CMOS technology. The 3.5-4 GHz band transmitter involves a phase-locked loop (PLL)-based binary frequency-shift keying subcarrier generator to provide the modulated triangular waveform. The RF signal is generated using a voltage-controlled oscillator and modulated by the triangular signal resulting in a constant-envelope UWB signal. A PLL-based RF calibration scheme is employed to calibrate the lower and the upper frequencies of the modulated RF signal. The purpose of this novel calibration structure is to keep transmitting in a fixed emission band. A Class-AB power amplifier is utilized to amplify the modulated RF signal and drive the antenna through an integrated matching network. The transmitter occupies 0.124 mm² silicon area and consumes 987 μW from a 1.2 V supply. It achieves an energy consumption of 0.987 nJ/bit at 1 Mbps data rate.

12 February 2021 in Microelectronics Journal
Verified
Journal Article
Closed access
MMMarwa Mansour
Marwa Mansour
Abdelhalim ZekryAbdelhalim Zekry
Abdelhalim Zekry
MAMohammed K. Ali
Mohammed K. Ali
HShttps://scholar.google.com/citations?hl=en&user=ibqBDS4AAAAJ Shawkey
https://scholar.google.com/citations?hl=en&user=ibqBDS4AAAAJ Shawkey

20 August 2020 in Electronics
Verified
Journal Article
Open access
PDF file
HShttps://scholar.google.com/citations?hl=en&user=ibqBDS4AAAAJ Shawkey
https://scholar.google.com/citations?hl=en&user=ibqBDS4AAAAJ Shawkey
Dalia ElsheakhDalia Elsheakh
Dalia Elsheakh
A new, compact, on-chip antenna architecture for 5G body-centric networks’ (BCNs) applications is presented in this paper. The integrated antenna combines two turns of dual-meander lines (DML) on two stacked layers and a metal ground layer. The proposed DML antenna structure operated at resonant bands 22 GHz, 34 GHz, 44 GHz, and 58 GHz with an operating bandwidth up to 2 GHz at impedance bandwidth ≤−7.5 dB (VSWR—Voltage Standing Wave Ratio ≤ 2.5) and antenna gain about −20 dBi, −15 dBi, −10 dBi, and −1 dBi, respectively. Then it was compared with conventional single-meander line antenna. The proposed structure decreased the resonant frequency by 22%, increased number of tuning bands, and broadened the operating bandwidth by 25%, 15%, 10%, and 20% for the tuning bands to be a suitable choice for high-data -ate biomedical applications. Furthermore, the proposed antenna was simulated and studied for its performance on and inside the human body to test the integration effect in wearable equipment. The results showed that the antenna had acceptable performance in both locations. All simulations of the proposed antenna were done were done by using Ansys HFSS (high-frequency structure simulator) v.15 (Ansys, Canonsburg, PA, USA). The DML (Digital Microwave Links) antenna was fabricated by using UMC (United Microelectronics Corporation) 180 nm CMOS (Complementary Metal–Oxidesemi–Conductor) technology with a total area of 1150 µm × 200 µm and the results showed a good agreement between measured and simulated results.

4 January 2023 in Active and Passive Electronic Components
Verified
Journal Article
Open access
PDF file
KAKawther I. Arafa
Kawther I. Arafa
DEDina M. Ellaithy
Dina M. Ellaithy
Abdelhalim ZekryAbdelhalim Zekry
Abdelhalim Zekry
MAMohamed Abouelatta
Mohamed Abouelatta
HShttps://scholar.google.com/citations?hl=en&user=ibqBDS4AAAAJ Shawkey
https://scholar.google.com/citations?hl=en&user=ibqBDS4AAAAJ Shawkey
This study presents a survey of the most promising reported SAR ADC designs for biomedical applications, stressing advantages, disadvantages, and limitations, and concludes with a quantitative comparison. Recent progress in the development of a single SAR ADC architecture is reviewed. In wearable and biosensor systems, a very small amount of total power must be devoured by portable batteries or energy-harvesting circuits in order to function correctly. During the past decade, implementation of the high energy efficiency of SAR ADC has become the most necessary. So, several different implementation schemes for the main components of the SAR ADC have been proposed. In this review study, the various circuit architectures have been explained, beginning with the sample and hold (S/H) switching circuits, the dynamic comparator, the internal digital-to-analog converter (DAC), and the SAR control logic. In order to achieve low power consumption, numerous different configurations of dynamic comparator circuits are revealed. At the end of this overview, the evolutions of DAC architecture in distinct biomedical applications today can make a tradeoff between resolution, speed, and linearity, which represent the challenges of a single SAR ADC. For high resolution, the dual split capacitive DAC (CDAC) array technique and hybrid capacitor technique can be used. Also, for ultralow power consumption, various voltage switching schemes are achieved to reduce the number of switches. These schemes can save switching energy and reduce capacitor array area with high linearity. Additionally, to increase the speed of the conversion process, a prediction-based ADC design is employed. Therefore, SAR ADC is considered the ideal solution for biomedical applications.

21 October 2020 in IEEE Access
Verified
Journal Article
Open access
RSRatshih Sayed
Ratshih Sayed
HAHaytham Azmi
Haytham Azmi
ANAmin M. Nassar
Amin M. Nassar
HShttps://scholar.google.com/citations?hl=en&user=ibqBDS4AAAAJ Shawkey
https://scholar.google.com/citations?hl=en&user=ibqBDS4AAAAJ Shawkey
This paper presents a novel framework that automates the creation of a trained-classifier integrated circuit from a dataset. The framework accepts a dataset in a comma-separated value format and performs several processing steps to create a trained model. After creating the model, the framework generates a tree-based machine learning classifier in two formats; Extensible Markup Language (XML) and Verilog. We use the XML representation to present the hierarchy of the generated tree and the Verilog code as a hardware description language representation of the trained model. Our framework uses the Verilog code as an input to a Field programmable Gate Array (FPGA) design validation flow. Then, we automate the Application Specific Integrated Circuit (ASIC) flow implementation and build a customized classifier integrated circuit. The novelty of the proposed framework lies in bridging the gap between machine learning model training and its hardware design when dealing with machine learning classifier’s implementation. Our framework addresses several challenges related to the design automation and implementation of customized machine learning classifier chips from raw dataset files. We discuss these challenges in detail in this paper and explain how researchers can use our proposed framework to build low-cost, high-performance classifier chips. Our framework operates at 100 MHZ and achieves 80.79% average 10-fold cross-validation accuracy across five different datasets.