Papers by Amjed Abbas Ahmed Al time-me
IEEE ACCESS, 2024
During the years 2018-2024, considerable advancements have been achieved in the use of deep learn... more During the years 2018-2024, considerable advancements have been achieved in the use of deep learning for side channel attacks. The security of cryptographic algorithm implementations is put at risk by this. The aim is to conceptually keep an eye out for specific types of information loss, like power usage, on a chip that is doing encryption. Next, one trains a model to identify the encryption key by using expertise of the underpinning encryption algorithm. The encryption key is then recovered by applying the model to traces that were obtained from a victim chip. Deep learning is being used in many different fields in the past several years. Convolutional neural networks and recurrent neural networks, for instance, have demonstrated efficacy in text generation and object detection in images, respectively. Deep learning has been effective throughout the side-channel analysis field. Until 2024, there was no deep learning layers made especially for SCAs. In this article, a systematic review on hybrid deep learning models for enhancing encryption techniques against side channel attacks is presented here.
springer, 2024
Recent Deep-Learning Side-Channel Attacks (DLSCAs) utilize networks of neural that have been skil... more Recent Deep-Learning Side-Channel Attacks (DLSCAs) utilize networks of neural that have been skilled over part of tracings that merely includes actions relevant to subkeys that were targeted. These attacks are known as "side-channel attacks." However, due to the restricted number of training traces that are available for deep learning models, such as in the case of the ASCAD database, there is a risk of overfitting occurring. A technique known as "data augmentation" is an example of a data-level method. This technique makes use of additional traces that have been synthetically altered to act as a regularizer and provide deep learning models with a greater capacity for generalization. Within the scope of this investigation, we provide a cross-subkey training approach that acts as an addition to traces. In order to train deep learning models, we use segments for the other 15 subkeys of AES-128 in addition to a trace segment that includes the SBox operation of the target subkey. We show that developing a network model with a mixture of numerous subkeys is superior to the more conventional method of training a network model with a single subkey by using two well-known datasets.
Kaunas University of Technology., 2023
Deep learning (DL) is a new option that has just been made available for side-channel analysis. D... more Deep learning (DL) is a new option that has just been made available for side-channel analysis. DL approaches for profiled side-channel attacks (SCA) have dominated research till now. In this attack, the attacker has complete control over the profiling device and can collect many traces for a range of critical parameters to characterise device leakage before the attack. In this study, we apply DL algorithms to non-profiled data. An attacker can only retrieve a limited number of side-channel traces from a closed device with an unknown key value in non-profiled mode. The authors conducted this research. Key estimations and deep learning measurements can reveal the secret key. We prove that this is doable. This technology is excellent for non-profits. DL and neural networks can benefit these organisations. Neural networks can provide a new technique to verify the safety of hardware cryptographic algorithms. It was recently suggested. This study creates a non-profiled SCA utilising convolutional neural networks (CNNs) on an AVR microcontroller with 8 bits of memory and the AES-128 cryptographic algorithm. We used aligned power traces with several samples to demonstrate how challenging CNN-based SCA is in practise. This will help us reach our goals. Here is another technique to create a solid CNN data set. In particular, CNN-based SCA experiment data and noise effects are examined. These experiments employ power traces with Gaussian noise. The CNN-based SCA works well with our data set for non-profiled attacks. Gaussian noise on power traces causes many more issues. These results show that our method can recover more bytes successfully from SCA compared to other methods in correlation power analysis (CPA) and DL-SCA without regularisation.
Encryption algorithms and encryption devices both play a key role in ensuring the safety of data ... more Encryption algorithms and encryption devices both play a key role in ensuring the safety of data that has been encrypted. Various types of attacks, such as energy analysis, can be used to assess the reliability of the encryption devices. Since it was originally introduced, side channel attacks' deep learning-based methodology has drawn plenty of attention. This is one of several different attack strategies. In this paper, a side channel attack method based on the LSTM deep learning network is suggested. The method use Correlation Power Analysis (CPA) to find the relevant information in the side channel power consumption data. The choice of a suitable interest interval to utilize as the feature vector in the creation of the neural network model is then guided by the position of the interest points. The trials' findings show that the LSTM model outperforms both MLP and CNN in terms of how well it executes side channel attacks.
Technology and artificial intelligence play a significant role in improving healthcare and enable... more Technology and artificial intelligence play a significant role in improving healthcare and enable tasks to be automated. In addition, the diseases can be better understood and diagnosed faster, saving time and reducing costs. This study examines the impact of transfer learning models on the effectiveness of deep learning models in classifying lung cancer through the analysis of CT scan images. Additionally, it investigates the relative performance of various machine learning and deep learning models, encompassing Support Vector Machine (SVM) and convolutional neural networks (CNN) such as Incep-tionV3, VGG16, Xception, ResNet50, and MobileNetV2, in the early detection of lung cancer based on CT scan images. The SVM model achieved an overall accuracy of 89% after preprocessing, the proposed approach was applied to five pre-trained models (ResNet50, In-ceptionV3, VGG16, Xception, MobileNetV2) using the dataset: Chest CT-Scan; Among the pre-trained CNN models, the Mo-bileNetV2 model achieved the highest accuracy of 98% and the lowest test loss, indicating it performed the best. The Xception model achieved the second-highest accuracy of 97%. The image pre-processing phase plays a significant role in improving system performance in terms of improving image contrast and increasing processing speed.
University of Bahrain, 2024
This work explores a novel method of SCA profiling to address compatibility problems and strength... more This work explores a novel method of SCA profiling to address compatibility problems and strengthen Deep Learning (DL) models. Convolutional Neural Networks are proposed in this research as a countermeasure to misalignment-focused countermeasures. "Time-Delay Convolutional Neural Networks" (TDCNN) is more accurate than "Convolutional Neural Network," yet it's still acceptable. It's true that TDCNNs are neural networks based on convolution learned on single spatial information, just as side-channel tracings. However, given to recent surge in popularity of CNNs, particularly from the year 2012 when CNN framework ("AlexNet") achieved Image Net Large Scale Visual Recognition Competition which is a notable image detection competition, a novel TDCNN has been termed out in DL literature. Currently, it needs to employ the characteristics related to CNN design, including declaring that one input feature equals 1 for instance, to establish a TDCNN in the most widely used DL libraries.
ELEKTRONIKA IR ELEKTROTECHNIKA (ISSN 1392-1215) is a peer-reviewed open access bimonthly research journal of Kaunas University of Technology., 2023
Power side channel analysis signal analysis is automated using deep learning. Signal processing a... more Power side channel analysis signal analysis is automated using deep learning. Signal processing and cryptanalytic techniques are necessary components of power side channel analysis. Chip leakages can be found using a classification approach called deep learning. In addition to this, we do this so that the deep learning network can automatically tackle signal processing difficulties such as realignment and noise reduction. We were able to break minimally protected Advanced Encryption Standard (AES), as well as maskingcountermeasure AES and protected elliptic-curve cryptography (ECC). These results demonstrate that the attacker knowledge required for side channel analysis, which had previously placed a significant emphasis on human abilities, is decreasing. This research will appeal to individuals with a technical background who have an interest in deep learning, side channel analysis, and security.
IEEE ACCESS, 2023
Neural network (NN) accelerators are now extensively utilized in a range of applications that nee... more Neural network (NN) accelerators are now extensively utilized in a range of applications that need a high degree of security, such as driverless cars, NLP, and image recognition. Due to privacy issues and the high cost, hardware implementations contained within NN Propagators were often not accessible for general populace. Additionally with power and time data, accelerators also disclose critical data by electro-magnetic (EM) sided channels. Within this study, we demonstrate a side-channel information-based attack that can successfully steal models from large-scale NN accelerators deployed on real-world hardware. The use of these accelerators is widespread. The proposed method of attack consists of two distinct phases: 1) Using EM side-channel data to estimate networking's underlying architecture; 2) Using margin-dependent, attackers learning actively in estimating parameters, notably weights. Deducing the underlying network structure from EM sidechannel data. Inferring the underlying network structure from EM sidechannel data. Experimental findings demonstrate that the disclosed attack technique can be used to precisely retrieve the large-scale NN via the use of EM side-channel information leaking. Overall, our attack shows how critical it is to conceal electromagnetic (EM) traces for massive NN accelerators in practical settings.
IEEE , 2023
The first non-profiled side-channel attack (SCA) method using deep learning is Timon's Differenti... more The first non-profiled side-channel attack (SCA) method using deep learning is Timon's Differential Deep Learning Analysis (DDLA). This method is effective in retrieving the secret key with the help of deep learning metrics. The Neural Network (NN) has to be trained numerous times since the proposed approach increases the learning cost with the key sizes, making it hard to assess the results from the intermediate stage. In this research, we provide three possible answers to the issues raised above, along with any challenges that could result from trying to solve these issues. We will start by offering an updated algorithm that has been modified to be able to keep track of the metrics during the intermediary stage. Next, we provide a parallel NN structure and training technique for a single network. This saves a lot of time by eliminating the need to repeatedly retrain the same model. The newly designed algorithm significantly sped up attacks when compared to the previous one. Thus, we propose employing shared layers to overcome memory challenges in parallel structure and improve performance. We evaluated our approaches by presenting nonprofiled attacks on ASCAD dataset and a ChipWhisperer-Lite power usage dataset. Power utilisation was studied using both datasets. The shared layers strategy we created was up to 134 times more successful than the prior technique when used to the ASCAD database.
IEEE, 2023
Depending on a device's encryption mechanism, a wide variety of tangible details could be exposed... more Depending on a device's encryption mechanism, a wide variety of tangible details could be exposed. These leaks are used in side-channel analysis, which is used to get keys. Due to deep learning's sensitivity to the characteristics of the data being processed, using such algorithms can significantly improve the accuracy and efficiency of side channel analysis. However, classic neural networks are now used for the vast majority of the work that is being done. When the number of nodes in a network grows, so does the efficiency with which key recovery can function. However, the method's computing complexity grows in direct proportion. Overfitting, inadequate capacity for feature extraction, and inefficient training are all potential issues. In this study, we develop a compact convolutional neural network by enhancing a previously existing combination of neural networks. Novel neural network along with previous neural network both have their own implementations of the side-channel analysis used in comparative trials. Statistically, the new network has better accuracy, quicker convergence, and more robustness. As part of the research, heatmaps were provided as a means of data visualisation. The critical interval concentration is higher and the heat value is higher in the new network. Conventional neural networks, which serve as the foundation for various kinds of neural networks, perform much worse than side channel studies based on feature fusion networks.
Elektronika Ir Elektrotechnika, 2023
Encryption algorithms and encryption devices both play a key role in ensuring the safety of data ... more Encryption algorithms and encryption devices both play a key role in ensuring the safety of data that has been encrypted. Various types of attacks, such as energy analysis, can be used to assess the reliability of the encryption devices. Since it was originally introduced, side channel attacks' deep learning-based methodology has drawn plenty of attention. This is one of several different attack strategies. In this paper, a side channel attack method based on the LSTM deep learning network is suggested. The method use Correlation Power Analysis (CPA) to find the relevant information in the side channel power consumption data. The choice of a suitable interest interval to utilize as the feature vector in the creation of the neural network model is then guided by the position of the interest points. The trials' findings show that the LSTM model outperforms both MLP and CNN in terms of how well it executes side channel attacks.
Iraqi Administrative Sciences Journal, 2018
This paper presents a descriptive information for the E-Learning, its design and development as w... more This paper presents a descriptive information for the E-Learning, its design and development as well as the application and its management taking into consideration of its major components. In recent times, E-Learning applications has done one major thing: remove distance as a barrier toward learning. E-Learning applications in form of online or virtual classrooms has been embraced especially in environments that have the facility to accommodate it. However, recent trends has caused a lot of reputable institutions to re-examine their respective model: the mode at which education is disseminated to their student population as Online Classrooms increased competitions among institutions all over the world. The aim of this paper is to examine E-Learning, the effects it already has on learning and future implications it might have on the education industry as its recent solutions provide outstanding features that makes people question the importance and relevance of institutions with phy...
Diyala Journal For Pure Science, 2017
Pattern recognition is a process of identifying vector of correlated/uncorrelated attributes and ... more Pattern recognition is a process of identifying vector of correlated/uncorrelated attributes and discriminate it among other patterns. Pattern recognition is synonymous to machine learning, data mining and Knowledge Discovery in Database (KDD).In this research work we investigate decomposing pattern (i.e., attribute vector) space into subspaces in which patterns cluster around basis of the subspaces. This paper introduces a theory which states that in case of having space of vectors and having basis then Signal Value Decomposition (SVD) can perform excellent in discovering thesis basis, hence, in pattern recognition a space can be decomposed to sub-spaces to reach clustering around basis. Results are collected and discussed and it has proven that SVD and its extension Latent Segment Analysis (LSA) can optimize the process of machine learning and showed a great tendency to converge toward cognitive based recognition.
Periodicals of Engineering and Natural Sciences (PEN), 2021
The task of recognizing arguments and their components in text is known as argument extraction. M... more The task of recognizing arguments and their components in text is known as argument extraction. Most arguments might be broken down into a petition and at least one premise that support it. A method to extract arguments is suggested in this work. The major words which are of high importance in arguments extraction were included in the suggested method on the basis of Arabic lexicon. The lexicon tool was used to apply classic text mining stages. The dataset, which includes over 3000 petitions, was collected from the Citizen Affairs Department in the Ministry of Health-Iraq. In addition, the experimental results exhibit that the suggested method extracts arguments from collected dataset with a 93.5% accuracy ratio.
IEEE Access
The linguistics related research and particularly, sentiment analysis using data-driven approache... more The linguistics related research and particularly, sentiment analysis using data-driven approaches, has been growing in recent years. However, the large number of users and excessive amount of information available on social media, make it difficult to detect extremism text on these platforms. The literature revealed a plethora of research studies focusing the sentiment analysis primarily, for English texts, however, very limited studies are available concerning the Arabic language which is the 4th mostly spoken language in the world. We first time in this study, propose a text detection mechanism for extremism orientations distinction in Arabic language, to improve the comprehension of subjective phrases. The study introduces a novel method based on Rough Set theory to enhance the accuracy of selected models and recognize text orientation reliably. Experimental outcomes indicate that the proposed method outperforms existing algorithms by contributing towards feature discriminations. Our method achieved 90.853%, 81.707% and 71.951% accuracies for unigram, bigram, and trigram representations, respectively. This study significantly contributes to the limited research in the field of machine learning and linguistics in Arabic language.
Al-Mustansiriyah Journal of Science, 2019
Identification and access have been a concept that has evolved over time as the need to constantl... more Identification and access have been a concept that has evolved over time as the need to constantly identify people and grant them access to sensitive and classified data and information became very important. The effect is felt in most organizations, especially multinational companies that deal in highly classified research that has to do with pharmaceuticals, technology, power as well as the human biology coupled with security. The most common form of implementation of biometrics is facial recognition, fingerprints, iris recognition, a retina scanner, and voice recognition into so many applications and scenarios. The integration of this biometrics has had a rising effect and impact of everyday life and has practically changed some daily routines. This paper will examine future integrations of biometrics and it will in time affect everyday life and routine.
Article in Journal of Advanced Research in Dynamical and Control Systems , 2018
Arabic speaking users in the world is increasing with more significance, higher depth and breadth... more Arabic speaking users in the world is increasing with more significance, higher depth and breadth at more pace, the reason is due to more impact on internet based resources. Due to increase in more Arabic users in the internet, there is necessity of computational techniques to categorize the Arabic text similar to English language. Arabic language is a complex in nature comparing to other speaking and scripting languages, so it requires detail research investigation on analysis of root extraction and text classification approaches for text datasets which are labelled in the form of single or multiple. The main objective of this paper is to propose a new study to improve the automated processing of Arabic texts. There are several machine learning approaches are existing, specifically, for this research, improved naïve Bayes classifier is applied. In first step, the documents which are unclassified are pre-processed by the method of punctuation removal and stop words. Second, after pre-processing, each document is represented by vector of words and frequencies as per the case of Naïve Bayes Classifier approach. Third, the technique of stemming was applied reduce the feature vector dimensionality. Fourth, classification is applied to categorise the Arabic text. The proposed work is an initial study and basic experiment was tested with an in-house Arabic text collection (i.e. selfdeveloped Arabic Corpus). Based on initial study using Naïve Bayes approach for Arabic text categorization, results of the classifier was promising compared to existing classifiers based on accuracy, precession, recall and error rates.
Steganography has attracted an outstanding area of research nowadays. Being the science of hiding... more Steganography has attracted an outstanding area of research nowadays. Being the science of hiding information, it focuses on the concept of hiding a message in plain sight [1]. It can be confusing and frustrating to pointedly understand how image quality is influenced by digital camera noise. The state of art analysis is relying on artificial intelligence and machine language techniques aiming to endow with better capability to differentiate between a carrier image from a hidden message image. Researches conducted in this field manifested that noisy areas in a picture is harder to construct, making it more difficult to detect an embedded message. Herein, this article shed the light on a novel steganography route for hiding information in these areas and bestow more complexation on their detection. The validation of the adopted method has been assessed through harnessing two imaging databases and two recent Steganalysis. The experiments disclosed that the proposed algorithm has significantly improved the statistical undetectability with respect to the LSB matching system for the same capacity of incrustation.
Volume 29, Issue 3, 2018
Identification and access have been a concept that has evolved over time as the need to constantl... more Identification and access have been a concept that has evolved over time as the need to constantly identify people and grant them access to sensitive and classified data and information became very important. The effect is felt in most organizations, especially multinational companies that deal in highly classified research that has to do with pharmaceuticals, technology, power as well as the human biology coupled with security. The most common form of implementation of biometrics is facial recognition, fingerprints, iris recognition, a retina scanner, and voice recognition into so many applications and scenarios. The integration of this biometrics has had a rising effect and impact of everyday life and has practically changed some daily routines. This paper will examine future integrations of biometrics and it will in time affect everyday life and routine.
AIP Conference Proceedings
The emergence of mobile robot cannot be overemphasized as it is seen as essential but extreme tec... more The emergence of mobile robot cannot be overemphasized as it is seen as essential but extreme technology taking into considerations, recent industrialisation. Automation has become a part of the process in order to obtain pathway of collision. The firefly algorithm has been identified to work on the principles of behaviour simulation. Two algorithms are implemented in this regard: Pseudo code of the basic firefly and Pseudo code of the firefly algorithm method that is proposed. To be considered for integration, the firefly algorithm method proposed is taken into consideration. A run of suggested mobile robot environment have been run. In conclusion, metaheuristic algorithms, is inspired by nature and are known as powerful tools especially in tackling global optimisation problem
Uploads
Papers by Amjed Abbas Ahmed Al time-me