This work analyses the performance of linear and different quadratic interpolators (in terms of e... more This work analyses the performance of linear and different quadratic interpolators (in terms of estimation error) for FFT frequency estimation of single tones under the effects of multiplicative noise. This method finds a quadratic fit in the neighborhood of the maximum of FFT with the three points, then apply different approximation methods: maximum of FFT, barycentric, and Quinn's Estimator. Numerical results showed that barycentric method is the best estimator under Gaussian multiplicative noise in terms of minimum mean squared estimation error, especially at high signal-to-noise ratios.
— Space-time geometrical channel models proved to be of significance in wireless communications a... more — Space-time geometrical channel models proved to be of significance in wireless communications as they provide spatial information (i.e. DOA, TOA) by which the performance of wireless communication systems and space-time systems (i.e., smart antennas, beamformers) can be analyzed. This paper presents a comparative study of space-time geometrical channel models. Based on the methods through which the models incorporate Doppler fading, we divide these models into two categories: models with oscillated scattering objects and models with unoscillated scattering objects. The main characteristics of each category have been studied using the existing models. I.
Indonesian Journal of Electrical Engineering and Computer Science, 2022
Recent research has demonstrated the effectiveness of utilizing neural networks for detect tamper... more Recent research has demonstrated the effectiveness of utilizing neural networks for detect tampering in images. However, because accessing a database is complex, which is needed in the classification process to detect tampering, reference-free steganalysis attracted attention. In recent work, an approach for least significant bit (LSB) steganalysis has been presented based on analyzing the derivatives of the histogram correlation. In this paper, we further examine this strategy for other steganographic methods. Detecting image tampering in the spatial domain, such as image steganography. It is found that the above approach could be applied successfully to other kinds of steganography with different orders of histogram-correlation derivatives. Also, the limits of the ratio stego-image to cover are considered, where very small ratios can escape this detection method unless modified.
In this work we present a study on the performance of signal similarity measures under non-Gaussi... more In this work we present a study on the performance of signal similarity measures under non-Gaussian noise. Pink noise has been considered, with 1/f power spectral density. This kind of noise has been generated by filtering Gaussian noise through an FIR filter. One-dimensional and two-dimensional signals have been considered. We tested 2D image similarity using the well-known similarity measures: Structural Similarity Index Measure (SSIM), modified Feature-based Similarity Measure (MFSIM), and Histogram-based Similarity Measure (HSSIM). Also, we tested 1D similarity measures: Cosine Similarity, Pearson Correlation, Tanimoto similarity, and Angular similarity. Results show that HSSIM and MFSIM outperform SSIM in low PSNR under pink noise and Gaussian noise. For 1D similarity, it is shown that Cosine Similarity and Pearson Correlation outperform other 1D similarity, especially at low SNR.
This paper represents a new approach for face recognition that incorporates Prewitt edge detectio... more This paper represents a new approach for face recognition that incorporates Prewitt edge detection, Gabor filter and Zernike moments to transform the image into a unified domain. On this joint domain, five distance metrics are constructed using Schoenberg transform for the purpose of defining efficient similarity measures for holistic face recognition. The proposed Schoenberg similarity applies Schoenberg transform to the logarithm of five existing distance metrics: Minkowski, City-Block, Euclidean, Soergel and Lorentzian metrics. The constructed Schoenberg logarithmic metrics are called SL-Minkowski, SL-City-Block, SL-Euclidean, SL-Soergel and SL-Lorentzian. These distance metrics are utilized as similarity measures after being normalized over the range [0,1] for fair comparison with existing measures. The proposed Schoenberg system can resist three problems: Change in illumination, pose and facial expression. Simulation results show that the proposed distance measures have superio...
Despite the increasing role of machine learning in various fields, very few works considered arti... more Despite the increasing role of machine learning in various fields, very few works considered artificial intelligence for frequency estimation (FE). This work presents comprehensive analysis of a deep-learning (DL) approach for frequency estimation of single tones. A DL network with two layers having a few nodes can estimate frequency more accurately than well-known classical techniques can. While filling the gap in the existing literature, the study is comprehensive, analyzing errors under different signal-to-noise ratios (SNRs), numbers of nodes, and numbers of input samples under missing SNR information. DL-based FE is not significantly affected by SNR bias or number of nodes. A DL-based approach can properly work using a minimal number of input nodes N at which classical methods fail. DL could use as few as two layers while having two or three nodes for each, with the complexity of O{N} compared with discrete Fourier transform (DFT)-based FE with O{Nlog2 (N)} complexity. Furtherm...
Frequency estimation of a single sinusoid in colored noise has received a considerable amount of ... more Frequency estimation of a single sinusoid in colored noise has received a considerable amount of attention in the research community. Taking into account the recent emergence and advances in compressive covariance sensing (CCS), the aim of this work is to combine the two disciplines by studying the effects of compressed measurements of a single sinusoid in moving-average colored noise on its frequency estimation accuracy. CCS techniques can recover the second-order statistics of the original uncompressed signal from the compressed measurements, thereby enabling correlation-based frequency estimation of single tones in colored noise using higher order lags. Acceptable accuracy is achieved for moderate compression ratios and for a sufficiently large number of available compressed signal samples. It is expected that the proposed method would be advantageous in applications involving resource-limited systems such as wireless sensor networks.
Meaningful information sharing between the sensors of a wireless sensor network (WSN) necessitate... more Meaningful information sharing between the sensors of a wireless sensor network (WSN) necessitates node localization, especially if the information to be shared is the location itself, such as in warehousing and information logistics. Trilateration and multilateration positioning methods can be employed in two-dimensional and threedimensional space respectively. These methods use distance measurements and analytically estimate the target location; they suffer from decreased accuracy and computational complexity especially in the three-dimensional case. Iterative optimization methods, such as gradient descent (GD), offer an attractive alternative and enable moving target tracking as well. This chapter focuses on positioning in three dimensions using time-of-arrival (TOA) distance measurements between the target and a number of anchor nodes. For centralized localization, a GD-based algorithm is presented for localization of moving sensors in a WSN. Our proposed algorithm is based on s...
Five power transform (a.k.a snowflake) distance metrics are constructed for the purpose of defini... more Five power transform (a.k.a snowflake) distance metrics are constructed for the purpose of defining similarity measures for holistic face recognition. The power transform metrics are based on the logarithm of five distance metrics: Euclidean, City-Block, Soergel, Lorentzian and Minkowski metrics. The constructed power logarithmic metrics have been called PL-Euclidean, PL-City-Block, PL-Soergel, PL-Lorentzian and PL-Minkowski. Distance is considered in the Zernike domain of face images, a domain that contains a vector of Zernike moments for each face image. These distance metrics are further modified as similarity measures then normalized to be restricted over the range (0,1) so that a fair comparison could be made with the well-known image structural similarity measure (SSIM) and Feature Similarity Index for Image Quality Assignment (FSIM). The purpose is discovering the similarity of a given face image with a face-database to check for best match. A measure for quality of recogniti...
A comprehensive study on the performance of image similarity techniques for face recognition is p... more A comprehensive study on the performance of image similarity techniques for face recognition is presented in this work. Adverse conditions on the reference image are considered in this work for the practical importance of face recognition under non-ideal conditions of noise and / or incomplete image information. |This study presents results from experiments on the effect of burst noise has on images and their structural similarity when transmitted through communication channels. Also addressed in this work is the effect incomplete images have on structural similarity including the effect of intensive burst noise on the missing parts of the image. The AT&T face image database was used in this work which consisted of images with dimensions 92x112 pixels and 256 grey levels per pixel. To quantify the error and evaluate system performance the Structural Similarity Index Measure (SSIM), Feature Similarity Index Measure (FSIM) and the Sjhcorr2 algorithms are considered. Peak signal to noi...
In this work, we present a comparative study on the performance of Fourier-based OFDM (FFT-OFDM) ... more In this work, we present a comparative study on the performance of Fourier-based OFDM (FFT-OFDM) and wavelet-based OFDM (DWT-OFDM) under compressive sensing (CS). Transmission over FFT-OFDM and DWT-OFDM, which has been made under different baseband modulation schemes such as Binary Phase Shift Keying (BPSK), Quadrature Phase Shift Key (QPSK), Quadrature amplitude modulation (16QAM) and (64QAM) has been considered. From numerical simulation results, it is observed that the Wavelet-based OFDM system outperforms Fourier based OFDM when the Quadrature Amplitude Modulation is 16QAM and 64QAM within the signal to noise ratios range 30 to 40 dB. Although CS is more efficient in compression than classical compression techniques, it introduces more errors over OFDM transmission. Future directions of this work are also suggested.
Recently, efficient direct numerical integrators of Runge-Kutta type (called RKD and RKT methods)... more Recently, efficient direct numerical integrators of Runge-Kutta type (called RKD and RKT methods) for solving third-order ordinary differential equations (ODEs) of the form y ′′′ = f (x, y) have been proposed. In this paper, we investigate the reliability of these RKD and RKT approaches, with focus on their stability and accuracy. We compare the stability regions of RKD and RKT methods. It is found that RKD stability region is adaptable, in the sense that its area can be controlled using a free parameter to get a more stable solution. To test the accuracy of RKD, we present some examples of this approach towards solving third-order ordinary differential equations. Simulation results show that the RKD approach, in addition to outperforming the existing RKT methods in terms of accuracy and time consumption, gives better control over stability region. AMS subject classification:
An image error test measure based on normalized mean squared error (MSE), called (NMSE), is prese... more An image error test measure based on normalized mean squared error (MSE), called (NMSE), is presented. The performance of the proposed error measure has been tested over FFT-OFDM system under the effect of Gaussian noise, impulse noise and Rayleigh noise. A comparison with the well-known structural similarity measure (SSIM) has been made under different baseband modulation schemes: BPSK, QPSK. It is shown that the NMSE measure outperforms SSIM at low SNR where it gives higher similarity for similar images, and test structural similarity measure (SSIM) for face recognition over FFTOFDM system and compare the performance under different kind of noise for different SNR, where in low SNR probably need re-transmission image over OFDM.
Frequency estimation of a single sinusoid in colored noise has received a considerable amount of ... more Frequency estimation of a single sinusoid in colored noise has received a considerable amount of attention in the research community. Taking into account the recent emergence and advances in compressive covariance sensing (CCS), the aim of this work is to combine the two disciplines by studying the effects of compressed measurements of a single sinusoid in moving-average (MA) colored noise on its frequency estimation accuracy. CCS techniques can recover the second-order statistics of the original uncompressed signal from the compressed measurements, thereby enabling correlation-based frequency estimation of single tones in colored noise using higher-order lags. Acceptable accuracy is achieved for moderate compression ratios and for a sufficiently large number of available compressed signal samples. It is expected that the proposed method would be advantageous in applications involving resource-limited systems such as wireless sensor networks.
Training data for a deep learning (DL) neural network (NN) controller are obtained from the input... more Training data for a deep learning (DL) neural network (NN) controller are obtained from the input and output signals of a conventional digital controller that is designed to provide the suitable control signal to a specified plant within a feedback digital control system. It is found that if the DL controller is sufficiently deep (four hidden layers), it can outperform the conventional controller in terms of settling time of the system output transient response to a unit-step reference signal. That is, the DL controller introduces a damping effect. Moreover, it does not need to be retrained to operate with a reference signal of different magnitude, or under system parameter change. Such properties make the DL control more attractive for applications that may undergo parameter variation, like sensor networks.
This work analyses the performance of linear and different quadratic interpolators (in terms of e... more This work analyses the performance of linear and different quadratic interpolators (in terms of estimation error) for FFT frequency estimation of single tones under the effects of multiplicative noise. This method finds a quadratic fit in the neighborhood of the maximum of FFT with the three points, then apply different approximation methods: maximum of FFT, barycentric, and Quinn's Estimator. Numerical results showed that barycentric method is the best estimator under Gaussian multiplicative noise in terms of minimum mean squared estimation error, especially at high signal-to-noise ratios.
— Space-time geometrical channel models proved to be of significance in wireless communications a... more — Space-time geometrical channel models proved to be of significance in wireless communications as they provide spatial information (i.e. DOA, TOA) by which the performance of wireless communication systems and space-time systems (i.e., smart antennas, beamformers) can be analyzed. This paper presents a comparative study of space-time geometrical channel models. Based on the methods through which the models incorporate Doppler fading, we divide these models into two categories: models with oscillated scattering objects and models with unoscillated scattering objects. The main characteristics of each category have been studied using the existing models. I.
Indonesian Journal of Electrical Engineering and Computer Science, 2022
Recent research has demonstrated the effectiveness of utilizing neural networks for detect tamper... more Recent research has demonstrated the effectiveness of utilizing neural networks for detect tampering in images. However, because accessing a database is complex, which is needed in the classification process to detect tampering, reference-free steganalysis attracted attention. In recent work, an approach for least significant bit (LSB) steganalysis has been presented based on analyzing the derivatives of the histogram correlation. In this paper, we further examine this strategy for other steganographic methods. Detecting image tampering in the spatial domain, such as image steganography. It is found that the above approach could be applied successfully to other kinds of steganography with different orders of histogram-correlation derivatives. Also, the limits of the ratio stego-image to cover are considered, where very small ratios can escape this detection method unless modified.
In this work we present a study on the performance of signal similarity measures under non-Gaussi... more In this work we present a study on the performance of signal similarity measures under non-Gaussian noise. Pink noise has been considered, with 1/f power spectral density. This kind of noise has been generated by filtering Gaussian noise through an FIR filter. One-dimensional and two-dimensional signals have been considered. We tested 2D image similarity using the well-known similarity measures: Structural Similarity Index Measure (SSIM), modified Feature-based Similarity Measure (MFSIM), and Histogram-based Similarity Measure (HSSIM). Also, we tested 1D similarity measures: Cosine Similarity, Pearson Correlation, Tanimoto similarity, and Angular similarity. Results show that HSSIM and MFSIM outperform SSIM in low PSNR under pink noise and Gaussian noise. For 1D similarity, it is shown that Cosine Similarity and Pearson Correlation outperform other 1D similarity, especially at low SNR.
This paper represents a new approach for face recognition that incorporates Prewitt edge detectio... more This paper represents a new approach for face recognition that incorporates Prewitt edge detection, Gabor filter and Zernike moments to transform the image into a unified domain. On this joint domain, five distance metrics are constructed using Schoenberg transform for the purpose of defining efficient similarity measures for holistic face recognition. The proposed Schoenberg similarity applies Schoenberg transform to the logarithm of five existing distance metrics: Minkowski, City-Block, Euclidean, Soergel and Lorentzian metrics. The constructed Schoenberg logarithmic metrics are called SL-Minkowski, SL-City-Block, SL-Euclidean, SL-Soergel and SL-Lorentzian. These distance metrics are utilized as similarity measures after being normalized over the range [0,1] for fair comparison with existing measures. The proposed Schoenberg system can resist three problems: Change in illumination, pose and facial expression. Simulation results show that the proposed distance measures have superio...
Despite the increasing role of machine learning in various fields, very few works considered arti... more Despite the increasing role of machine learning in various fields, very few works considered artificial intelligence for frequency estimation (FE). This work presents comprehensive analysis of a deep-learning (DL) approach for frequency estimation of single tones. A DL network with two layers having a few nodes can estimate frequency more accurately than well-known classical techniques can. While filling the gap in the existing literature, the study is comprehensive, analyzing errors under different signal-to-noise ratios (SNRs), numbers of nodes, and numbers of input samples under missing SNR information. DL-based FE is not significantly affected by SNR bias or number of nodes. A DL-based approach can properly work using a minimal number of input nodes N at which classical methods fail. DL could use as few as two layers while having two or three nodes for each, with the complexity of O{N} compared with discrete Fourier transform (DFT)-based FE with O{Nlog2 (N)} complexity. Furtherm...
Frequency estimation of a single sinusoid in colored noise has received a considerable amount of ... more Frequency estimation of a single sinusoid in colored noise has received a considerable amount of attention in the research community. Taking into account the recent emergence and advances in compressive covariance sensing (CCS), the aim of this work is to combine the two disciplines by studying the effects of compressed measurements of a single sinusoid in moving-average colored noise on its frequency estimation accuracy. CCS techniques can recover the second-order statistics of the original uncompressed signal from the compressed measurements, thereby enabling correlation-based frequency estimation of single tones in colored noise using higher order lags. Acceptable accuracy is achieved for moderate compression ratios and for a sufficiently large number of available compressed signal samples. It is expected that the proposed method would be advantageous in applications involving resource-limited systems such as wireless sensor networks.
Meaningful information sharing between the sensors of a wireless sensor network (WSN) necessitate... more Meaningful information sharing between the sensors of a wireless sensor network (WSN) necessitates node localization, especially if the information to be shared is the location itself, such as in warehousing and information logistics. Trilateration and multilateration positioning methods can be employed in two-dimensional and threedimensional space respectively. These methods use distance measurements and analytically estimate the target location; they suffer from decreased accuracy and computational complexity especially in the three-dimensional case. Iterative optimization methods, such as gradient descent (GD), offer an attractive alternative and enable moving target tracking as well. This chapter focuses on positioning in three dimensions using time-of-arrival (TOA) distance measurements between the target and a number of anchor nodes. For centralized localization, a GD-based algorithm is presented for localization of moving sensors in a WSN. Our proposed algorithm is based on s...
Five power transform (a.k.a snowflake) distance metrics are constructed for the purpose of defini... more Five power transform (a.k.a snowflake) distance metrics are constructed for the purpose of defining similarity measures for holistic face recognition. The power transform metrics are based on the logarithm of five distance metrics: Euclidean, City-Block, Soergel, Lorentzian and Minkowski metrics. The constructed power logarithmic metrics have been called PL-Euclidean, PL-City-Block, PL-Soergel, PL-Lorentzian and PL-Minkowski. Distance is considered in the Zernike domain of face images, a domain that contains a vector of Zernike moments for each face image. These distance metrics are further modified as similarity measures then normalized to be restricted over the range (0,1) so that a fair comparison could be made with the well-known image structural similarity measure (SSIM) and Feature Similarity Index for Image Quality Assignment (FSIM). The purpose is discovering the similarity of a given face image with a face-database to check for best match. A measure for quality of recogniti...
A comprehensive study on the performance of image similarity techniques for face recognition is p... more A comprehensive study on the performance of image similarity techniques for face recognition is presented in this work. Adverse conditions on the reference image are considered in this work for the practical importance of face recognition under non-ideal conditions of noise and / or incomplete image information. |This study presents results from experiments on the effect of burst noise has on images and their structural similarity when transmitted through communication channels. Also addressed in this work is the effect incomplete images have on structural similarity including the effect of intensive burst noise on the missing parts of the image. The AT&T face image database was used in this work which consisted of images with dimensions 92x112 pixels and 256 grey levels per pixel. To quantify the error and evaluate system performance the Structural Similarity Index Measure (SSIM), Feature Similarity Index Measure (FSIM) and the Sjhcorr2 algorithms are considered. Peak signal to noi...
In this work, we present a comparative study on the performance of Fourier-based OFDM (FFT-OFDM) ... more In this work, we present a comparative study on the performance of Fourier-based OFDM (FFT-OFDM) and wavelet-based OFDM (DWT-OFDM) under compressive sensing (CS). Transmission over FFT-OFDM and DWT-OFDM, which has been made under different baseband modulation schemes such as Binary Phase Shift Keying (BPSK), Quadrature Phase Shift Key (QPSK), Quadrature amplitude modulation (16QAM) and (64QAM) has been considered. From numerical simulation results, it is observed that the Wavelet-based OFDM system outperforms Fourier based OFDM when the Quadrature Amplitude Modulation is 16QAM and 64QAM within the signal to noise ratios range 30 to 40 dB. Although CS is more efficient in compression than classical compression techniques, it introduces more errors over OFDM transmission. Future directions of this work are also suggested.
Recently, efficient direct numerical integrators of Runge-Kutta type (called RKD and RKT methods)... more Recently, efficient direct numerical integrators of Runge-Kutta type (called RKD and RKT methods) for solving third-order ordinary differential equations (ODEs) of the form y ′′′ = f (x, y) have been proposed. In this paper, we investigate the reliability of these RKD and RKT approaches, with focus on their stability and accuracy. We compare the stability regions of RKD and RKT methods. It is found that RKD stability region is adaptable, in the sense that its area can be controlled using a free parameter to get a more stable solution. To test the accuracy of RKD, we present some examples of this approach towards solving third-order ordinary differential equations. Simulation results show that the RKD approach, in addition to outperforming the existing RKT methods in terms of accuracy and time consumption, gives better control over stability region. AMS subject classification:
An image error test measure based on normalized mean squared error (MSE), called (NMSE), is prese... more An image error test measure based on normalized mean squared error (MSE), called (NMSE), is presented. The performance of the proposed error measure has been tested over FFT-OFDM system under the effect of Gaussian noise, impulse noise and Rayleigh noise. A comparison with the well-known structural similarity measure (SSIM) has been made under different baseband modulation schemes: BPSK, QPSK. It is shown that the NMSE measure outperforms SSIM at low SNR where it gives higher similarity for similar images, and test structural similarity measure (SSIM) for face recognition over FFTOFDM system and compare the performance under different kind of noise for different SNR, where in low SNR probably need re-transmission image over OFDM.
Frequency estimation of a single sinusoid in colored noise has received a considerable amount of ... more Frequency estimation of a single sinusoid in colored noise has received a considerable amount of attention in the research community. Taking into account the recent emergence and advances in compressive covariance sensing (CCS), the aim of this work is to combine the two disciplines by studying the effects of compressed measurements of a single sinusoid in moving-average (MA) colored noise on its frequency estimation accuracy. CCS techniques can recover the second-order statistics of the original uncompressed signal from the compressed measurements, thereby enabling correlation-based frequency estimation of single tones in colored noise using higher-order lags. Acceptable accuracy is achieved for moderate compression ratios and for a sufficiently large number of available compressed signal samples. It is expected that the proposed method would be advantageous in applications involving resource-limited systems such as wireless sensor networks.
Training data for a deep learning (DL) neural network (NN) controller are obtained from the input... more Training data for a deep learning (DL) neural network (NN) controller are obtained from the input and output signals of a conventional digital controller that is designed to provide the suitable control signal to a specified plant within a feedback digital control system. It is found that if the DL controller is sufficiently deep (four hidden layers), it can outperform the conventional controller in terms of settling time of the system output transient response to a unit-step reference signal. That is, the DL controller introduces a damping effect. Moreover, it does not need to be retrained to operate with a reference signal of different magnitude, or under system parameter change. Such properties make the DL control more attractive for applications that may undergo parameter variation, like sensor networks.
Uploads