Resume

• 1988

  Master's degree

  Industrial Electronics

  Université du Québec à Trois-Rivières

  M.Sc.A.

• 1984

  B.Eng.

  Electrical Engineering

  Philosophy Doctor (Ph.D.)

  Electrical Engineering

• Algorithms

  Electronics

  Digital Signal Processors

  Wireless

  Xilinx

  Electrical Engineering

  FPGA

  Matlab

  Digital Signal Processing

  Telecommunications

  VLSI

  Adaptive Filtering

  Embedded Systems

  Signal Processing

  VHDL

  Wireless Communications Systems

  MIMO

  Microelectronics

  R&D

  Simulations

  Enhanced FBLMS algorithm for ECG and noise removal from sEMG signals

  François Nougarou

  In this paper

  we proposed a Dual-adapted Fast Block Least Mean Squares algorithm (DA-FBLMS) to remove electrocardiogram (ECG) and noise contaminations from surface electromyography signals (sEMG). Based on an adaptive noise cancelation (ANC) structure and artificial input signals

  the ANC integrated proposed algorithm distinguishes itself by the use of an iterative method characterized by a varying number of updates for every different input block

  combined with an adaptive step size guided by a QRS detector and the average error of the corresponding input block. The simulations demonstrate that the proposed DA-FBLMS algorithm presents better performances during the contaminations cancellation compared to a recursive least squares algorithm (RLS) and classic FBLMS algorithm

  especially in noisy and high distortion environments.

  Enhanced FBLMS algorithm for ECG and noise removal from sEMG signals

  Mounir Boukadoum

  This study addresses electrocardiogram (ECG) pulses removal from surface electromyography (sEMG) recordings. We describe a block singular spectrum analysis based adaptive noise canceler (BSSA-ANC) in order to enhance the filtering performance of sEMG signals. The proposed method distinguishes itself by adapting the eigenvalues of every input block data

  using an adaptive noise canceling (ANC) filter based on error gradient minimization

  during the grouping stage of the well-known SSA technique. Using semi-artificially prepared signals

  we demonstrate that the least mean square (LMS) when combined with the proposed technique provides higher filtering performances than with standard mean. The simulation results using real world data confirm the improved noise rejection obtained with the proposed method

  Time Frequency Adaptive Filtering for an Optimized Separation of ECG from Neuromuscular sEMG Signals

  Mounir Boukadoum

  This study addresses the removal of electrocardiogram pulses (ECG) from surface electromyography signals (sEMG) in wireless sEMG electrodes. We describe a wavelet-compression inspired filtering technique in order to minimize the computational complexity required in wireless sEMG electrodes while providing higher filtering performance than when with standard means. Using semi-artificially prepared signals

  we show that the discrete wavelet transform (DWT) enables a partial separation between the sEMG and ECG signals in time domain. Then

  only the highly overlapped parts are fed into an adaptive noise cancellation (ANC) structure for lower computational complexity filtering. The simulation results confirm the improved noise rejection and the gain in computational complexity obtained with the proposed method.

  Wavelet Compression Inspired Implementation for High Performances and Low Complexity ECG Removal in Wireless sEMG Electrodes

  Nonlinear adaptive channel equalization is a well-documented problem. Equalizers based on the complex decision feedback recurrent neural network (CDFRNN) have been intensively studied to address this problem. However

  when trained with conventional training algorithms like the real time recurrent learning (RTRL) technique

  the equalizer suffers from low convergence speed

  \nrequiring very long training sequence to achieve proper performance. In this work

  we propose a new approach to equalize nonlinear channels using genetic algorithms. The proposed Volterra decision feedback genetic algorithm (VDFGA) uses a genetic optimization strategy to estimate Volterra kernels in order to model the inverse of the channel response. Simulation results show very high convergence speed

  which allowed to achieve interesting bit error rate (BER) using relatively short training symbols

  when considering only 8-bits long coded weights.

  Nonlinear Adaptive Channel Equalization using Genetic Algorithms

  Mounir Boukadoum

  This study addresses electrocardiogram (ECG) pulses removal from surface electromyography (sEMG) recordings. We describe a block singular spectrum analysis based adaptive noise canceler (BSSA-ANC) in order to enhance the filtering performance of sEMG signals. The proposed method distinguishes itself by adapting the eigenvalues of every input block data

  using an adaptive noise canceling (ANC) filter based on error gradient minimization

  during the grouping stage of the well-known SSA technique. Using semi-artificially prepared signals

  we demonstrate that the least mean square (LMS) when combined with the proposed technique provides higher filtering performances than with standard mean. The simulation results using real world data confirm the improved noise rejection obtained with the proposed method.

  Adaptive Block SSA Based ANC Implementation For High Performances ECG Removal From sEMG Signals

  In this paper

  we propose an FPGA implementation of a genetic algorithm (GA) for linear and nonlinear auto regressive moving average (ARMA) model parameters identification. The GA features specifically designed genetic operators for adaptive filtering applications. The design was implemented using very low bit-wordlength fixed-point representation

  where only 6-bit wordlength arithmetic was used. The implementation experiments show high parameters identification capabilities and low footprint.

  FPGA Based Implementation of a Genetic Algorithm for ARMA Model Parameters Identification

  Genetic algorithms are increasingly being used to address adaptive filtering problems. The interest lies in their ability to find the global solutions for linear and nonlinear problems. However

  all the work available in the literature use software implementations running on sequential processors. This work proposes a hardware architecture of a real-time genetic algorithm for adaptive filtering applications. Specifically designed genetic operators are proposed to improve processing performance and robustness to the quantization effect

  making low bitwordlength fixed-point arithmetic implementation possible

  which permit hardware cost saving. The proposed architecture is modeled in VHDL and implemented in FPGA using 6-bits wordlength

  addressing linear and\nnonlinear auto regressive moving average (ARMA) model parameters identification problem. The implementation experiments show high signal processing performance and low resources cost.

  Hardware Implementation of a Real-time Genetic Algorithm for Adaptive Filtering Applications

  Genetic algorithms (GAs) have been successfully applied to resolve adaptive filtering problems. The main advantage of using such algorithms over conventional adaptive filtering techniques

  is their ability to deal with nonlinear systems. However

  intensive computations are needed to achieve proper performances

  which can be very critical when limited-resource devices are considered for hardware implementation. This work proposes a low computation load genetic algorithm for adaptive filtering applications. Very low bit-wordlength fixed-point arithmetic is used in all operations to minimize the algorithm footprint. We compare the proposed method with the least mean square (LMS) and theatrical performances in identifying the parameters of an auto regressive moving average (ARMA) model. Simulation results show the high performances of the algorithm to deal with linear and nonlinear environments where only 6-bit wordlength is used.

  Towards Hardware Implementation of a Real-time Genetic Algorithm for Adaptive Filtering Applications

  UQTR - Université du Québec à Trois-Rivières

  Axiocom

  ENSSAT

  Universidad de Chile

  Lab-STICC

  Universite de Bretagne Sud

  Visiting Researcher

  Chile

  Universidad de Chile

  UQTR - Université du Québec à Trois-Rivières

  ENSSAT

  Visiting researcher

  France

  UQTR

  President Founder and CTO

  Axiocom

  Lab-STICC

  Universite de Bretagne Sud

  Visiting Researcher

  France

  Groupe de recherche en électronique industrielle

  Le Groupe de recherche en électronique industrielle (GRÉI) a été créé et accrédité le 1er juin 1988 pour répondre à des besoins industriels et de recherche dans les domaines la conversion

  gestion et utilisation de l'énergie

  des télécommunications

  des technologies émergentes

  incluant notamment la microélectronique

  micro et nanosystèmes.

  Directeur du Groupe de recherche en électronique industrielle

  UQTR - Université du Québec à Trois-Rivières