Biblio

In today's world, there is rapid progress in the field of artificial intelligence and image captioning. It becomes a fascinating task that has saw widespread interest. The task of image captioning comprises image description engendered based on the hybrid combination of deep learning, natural language processing, and various approaches of machine learning and computer vision. In this work authors emphasize on how the model generates a short description as an output of the input image using the functionalities of Deep Learning and Natural Language Processing, for helping visually impaired people, and can also be cast-off in various web sites to automate the generation of captions reducing the task of recitation with great ease.

Residue number systems provide efficient techniques for speeding up calculations and/or protecting against side channel attacks when used in the context of cryptographic engineering. One of the interests of such systems is their scalability, as the existence of large bases for some specialized systems is often an open question. In this paper, we present highly optimized methods for generating large bases for residue number systems and, in some cases, the largest possible bases. We show their efficiency by demonstrating their improvement over the state-of-the-art bases reported in the literature. This work make it possible to address the problem of the scalability issue of finding new bases for a specific system that arises whenever a parameter changes, and possibly open new application avenues.

Logic locking is a holistic design-for-trust technique that aims to protect the design intellectual property (IP) from untrustworthy entities throughout the supply chain. Functional and structural analysis-based attacks successfully circumvent state-of-the-art, provably secure logic locking (PSLL) techniques. However, such attacks are not holistic and target specific implementations of PSLL. Automating the detection and subsequent removal of protection logic added by PSLL while accounting for all possible variations is an open research problem. In this paper, we propose GNNUnlock, the first-of-its-kind oracle-less machine learning-based attack on PSLL that can identify any desired protection logic without focusing on a specific syntactic topology. The key is to leverage a well-trained graph neural network (GNN) to identify all the gates in a given locked netlist that belong to the targeted protection logic, without requiring an oracle. This approach fits perfectly with the targeted problem since a circuit is a graph with an inherent structure and the protection logic is a sub-graph of nodes (gates) with specific and common characteristics. GNNs are powerful in capturing the nodes' neighborhood properties, facilitating the detection of the protection logic. To rectify any misclassifications induced by the GNN, we additionally propose a connectivity analysis-based post-processing algorithm to successfully remove the predicted protection logic, thereby retrieving the original design. Our extensive experimental evaluation demonstrates that GNNUnlock is 99.24% - 100% successful in breaking various benchmarks locked using stripped-functionality logic locking [1], tenacious and traceless logic locking [2], and Anti-SAT [3]. Our proposed post-processing enhances the detection accuracy, reaching 100% for all of our tested locked benchmarks. Analysis of the results corroborates that GNNUnlock is powerful enough to break the considered schemes under different parameters, synthesis settings, and technology nodes. The evaluation further shows that GNNUnlock successfully breaks corner cases where even the most advanced state-of-the-art attacks [4], [5] fail. We also open source our attack framework [6].

A new method for judging degree sequence is shown by means of perfect ice-flower systems made by operators - stars (particular complete bipartite graphs), and moreover this method can be used to build up degree sequences and perfect ice-flower systems. Graphic lattice, graph-graphic lattice, caterpillar-graphic lattice and topological coding lattice are defined. We establish some connections between traditional lattices and graphic lattices trying to provide new techniques for Lattice-based cryptosystem and post-quantum cryptography, and trying to enrich the theoretical knowledge of topological coding.

Graphs have proved to be a promising data structure to solve complex problems in various domains. Graphs store data in an associative manner which is analogous to the manner in which humans store memories in the brain. Generathe chatbots lack the ability to recall details revealed by the user in long conversations. To solve this problem, we have used graph-based memory to recall-related conversations from the past. Thus, providing context feature derived from query systems to generative systems such as OpenAI GPT. Using graphs to detect important details from the past reduces the total amount of processing done by the neural network. As there is no need to keep on passingthe entire history of the conversation. Instead, we pass only the last few pairs of utterances and the related details from the graph. This paper deploys this system and also demonstrates the ability to deploy such systems in real-world applications. Through the effective usage of knowledge graphs, the system is able to reduce the time complexity from O(n) to O(1) as compared to similar non-graph based implementations of transfer learning- based conversational agents.

Text analytics directly on compression (TADOC) has proven to be a promising technology for big data analytics. GPUs are extremely popular accelerators for data analytics systems. Unfortunately, no work so far shows how to utilize GPUs to accelerate TADOC. We describe G-TADOC, the first framework that provides GPU-based text analytics directly on compression, effectively enabling efficient text analytics on GPUs without decompressing the input data. G-TADOC solves three major challenges. First, TADOC involves a large amount of dependencies, which makes it difficult to exploit massive parallelism on a GPU. We develop a novel fine-grained thread-level workload scheduling strategy for GPU threads, which partitions heavily-dependent loads adaptively in a fine-grained manner. Second, in developing G-TADOC, thousands of GPU threads writing to the same result buffer leads to inconsistency while directly using locks and atomic operations lead to large synchronization overheads. We develop a memory pool with thread-safe data structures on GPUs to handle such difficulties. Third, maintaining the sequence information among words is essential for lossless compression. We design a sequence-support strategy, which maintains high GPU parallelism while ensuring sequence information. Our experimental evaluations show that G-TADOC provides 31.1× average speedup compared to state-of-the-art TADOC.

Homomorphic Encryption (HE) comes as a sophisticated and powerful cryptography system that can preserve the privacy of data in all cases when the data is at rest or even when data is in processing and computing. All the computations needed by the user or the provider can be done on the encrypted data without any need to decrypt it. However, HE has overheads such as big key sizes and long ciphertexts and as a result long execution time. This paper proposes a novel solution for big data analytic based on clustering and the Elliptical Curve Cryptography (ECC). The Extremely Distributed Clustering technique (EDC) has been used to divide big data into several subsets of cloud computing nodes. Different clustering techniques had been investigated, and it was found that using hybrid techniques can improve the performance and efficiency of big data analytic while at the same time data is protected and privacy is preserved using ECC.

Web server is an appealing target for attackers since it may be exploited to gain access to an organization’s internal network. After compromising a web server, the attacker can construct a webshell to maintain a long-term and stealthy access for further attacks. Among all webshell-based attacks, command injection is a powerful attack that can be launched to steal sensitive data from the web server or compromising other computers in the network. To monitor and analyze webshell-based command injection, we develop a hybrid webshell honeypot framework called HoneyBog, which intercepts and redirects malicious injected commands from the front-end honeypot to the high-fidelity back-end honeypot for execution. HoneyBog can achieve two advantages by using the client-server honeypot architecture. First, since the webshell-based injected commands are transferred from the compromised web server to a remote constrained execution environment, we can prevent the attacker from launching further attacks in the protected network. Second, it facilitates the centralized management of high-fidelity honeypots for remote honeypot service providers. Moreover, we increase the system fidelity of HoneyBog by synchronizing the website files between the front-end and back-end honeypots. We implement a prototype of HoneyBog using PHP and the Apache web server. Our experiments on 260 PHP webshells show that HoneyBog can effectively intercept and redirect injected commands with a low performance overhead.

As a decoy for hackers, honeypots have been proved to be a very valuable tool for collecting real data. However, due to closed source and vendor-specific firmware, there are significant limitations in cost for researchers to design an easy-to-use and high-interaction honeypot for industrial control systems (ICSs). To solve this problem, it’s necessary to find a cost-effective solution. In this paper, we propose a novel honeypot architecture termed HoneyVP to support a semi-virtual and semi-physical honeypot design and implementation to enable high cost performance. Specially, we first analyze cyber-attacks on ICS devices in view of different interaction levels. Then, in order to deal with these attacks, our HoneyVP architecture clearly defines three basic independent and cooperative components, namely, the virtual component, the physical component, and the coordinator. Finally, a local-remote cooperative ICS honeypot system is implemented to validate its feasibility and effectiveness. Our experimental results show the advantages of using the proposed architecture compared with the previous honeypot solutions. HoneyVP provides a cost-effective solution for ICS security researchers, making ICS honeypots more attractive and making it possible to capture physical interactions.

The COVID19 pandemic situation has opened a wide range of opportunities for cyber-criminals, who take advantage of the anxiety generated and the time spent on the Internet, to undertake massive phishing campaigns. Although companies are adopting protective measures, the psychological traits of the victims are still considered from a very generic perspective. In particular, current literature determines that the model proposed in the Big-Five personality traits (i.e., Openness, Conscientiousness, Extraversion, Agreeableness, and Neuroticism) might play an important role in human behaviour to counter cybercrime. However, results do not provide unanimity regarding the correlation between phishing susceptibility and neuroticism. With the aim to understand this lack of consensus, this article provides a comprehensive literature review of papers extracted from relevant databases (IEEE Xplore, Scopus, ACM Digital Library, and Web of Science). Our results show that there is not a well-established psychological theory explaining the role of neuroticism in the phishing context. We sustain that non-representative samples and the lack of homogeneity amongst the studies might be the culprits behind this lack of consensus on the role of neuroticism on phishing susceptibility.

In this paper, we propose a novel protocol to represent the human factor on a blockchain environment. Our approach allows single or groups of humans to propose data in blocks which cannot be validated automatically but need human knowledge and collaboration to be validated. Only if human-based consensus on the correctness and trustworthiness of the data is reached, the new block is appended to the blockchain. This human approach significantly extends the possibilities of blockchain applications on data types apart from financial transaction data.

Reusability is demarcated as the way of utilizing existing software components in software development. It plays a significant role in component-based software engineering. Extracting the components from the source code and checking the reusability factors is the most crucial part. Software Intelligence, a combination of data mining and artificial intelligence, helps to cope with the extraction and detection of reusability factor of the component. In this work prediction of reusability factor is considered. This paper proposes a hybrid PSO-NSGA III approach to detect whether the extracted component is reusable or not. The existing models lack in tuning the hyper parameters for prediction, which is considered in this work. The proposed approach was compared with four models, showing better outcomes in terms of performance metrics.

Cloud technology is advancing rapidly because of it’s capability to replace the traditional computing techniques. Cloud offers various kinds of services for the user that are being used. In this research paper, storage as a service provided by cloud is examined as the data of the owner is being shared to the cloud so we have to ensure that data integrity is being maintained. In order to have a robust mechanism that offers a secure pathway for sharing data different encryption algorithms have been utilized. We investigate all the suitable algorithms with various combinations because any single algorithm is prone to some kind of attack. Testing of these algorithms is done by analyzing the parameters such as time required for execution, use of computational resources, key management, etc. Finally the best one that stands and fulfill all the criteria in a reasonable manner is selected for the purpose of storage.

A new methodology to calculate the hysteresis induced power loss of inductor from the measured waveforms of DC-to-DC converter during the voltage conversion is presented. From this study, we find that the duty cycles (D) of the buck and boost converters used till date for inductance current calculation are not exactly equal to VOUT/VIN and 1-VIN/VOUT as the inductance change induced by the hysteresis effect cannot be neglected. Although the increase in the loading currents of the converter increases the remanence magnetization of inductor at the turn-off time (toff), this remanence magnetization is destroyed by the turbulence induced vortex current at the transistor turn-on transient. So, the core power loss of inductor increases with the loading current of the converter and becomes much larger than other power losses and cannot be neglected for the power efficiency calculation during power stage design.

One of the significant difficulties in network safety is the arrangement of a mechanized and viable digital danger's location strategy. This paper presents an AI procedure for digital dangers recognition, in light of fake neural organizations. The proposed procedure changes large number of gathered security occasions over to singular occasion profiles and utilize a profound learning-based discovery strategy for upgraded digital danger identification. This research work develops an AI-SIEM framework dependent on a blend of occasion profiling for information preprocessing and distinctive counterfeit neural organization techniques by including FCNN, CNN, and LSTM. The framework centers around separating between obvious positive and bogus positive cautions, consequently causing security examiners to quickly react to digital dangers. All trials in this investigation are performed by creators utilizing two benchmark datasets (NSLKDD and CICIDS2017) and two datasets gathered in reality. To assess the presentation correlation with existing techniques, tests are carried out by utilizing the five ordinary AI strategies (SVM, k-NN, RF, NB, and DT). Therefore, the exploratory aftereffects of this examination guarantee that our proposed techniques are fit for being utilized as learning-based models for network interruption discovery and show that despite the fact that it is utilized in reality, the exhibition beats the traditional AI strategies.

Network Address Translation (NAT) is an address remapping technique placed at the borders of stub domains. It is present in almost all routers and CPEs. Most NAT devices implement Port Address Translation (PAT), which allows the mapping of multiple private IP addresses to one public IP address. Based on port number information, PAT matches the incoming traffic to the corresponding "hidden" client. In an enterprise context, and with the proliferation of unauthorized wired and wireless NAT routers, NAT can be used for re-distributing an Intranet or Internet connection or for deploying hidden devices that are not visible to the enterprise IT or under its oversight, thus causing a problem known as shadow IT. Thus, it is important to detect NAT devices in an intranet to prevent this particular problem. Previous methods in identifying NAT behavior were based on features extracted from traffic traces per flow. In this paper, we propose a method to identify NAT devices using a machine learning approach from aggregated flow features. The approach uses multiple statistical features in addition to source and destination IPs and port numbers, extracted from passively collected traffic data. We also use aggregated features extracted within multiple window sizes and feed them to a machine learning classifier to study the effect of timing on NAT detection. Our approach works completely passively and achieves an accuracy of 96.9% when all features are utilized.

With meteoric advancement in quantum computing, the traditional data integrity verifying schemes are no longer safe for cloud data storage. A large number of the current techniques are dependent on expensive Public Key Infrastructure (PKI). They cost computationally and communicationally heavy for verification which do not stand with the advantages when quantum computing techniques are applied. Hence, a quantum safe and efficient integrity verification protocol is a research hotspot. Lattice-based signature constructions involve matrix-matrix or matrix vector multiplications making computation competent, simple and resistant to quantum computer attacks. Study in this paper uses Bloom Filter which offers high efficiency in query and search operations. Further, we propose an Identity-Based Integrity Verification (IBIV) protocol for cloud storage from Lattice and Bloom filter. We focus on security against attacks from Cloud Service Provider (CSP), data privacy attacks against Third Party Auditor (TPA) and improvement in efficiency.

This paper investigates the impact of the delay resulting from a blockchain, a promising security measure, for a hierarchical control system of inverters connected to the grid. The blockchain communication network is designed at the secondary control layer for resilience against cyberattacks. To represent the latency in the communication channel, a model is developed based on the complexity of the blockchain framework. Taking this model into account, this work evaluates the plant’s performance subject to communication delays, introduced by the blockchain, among the hierarchical control agents. In addition, this article considers an optimal model-based control strategy that performs the system’s internal control loop. The work shows that the blockchain’s delay size influences the convergence of the power supplied by the inverter to the reference at the point of common coupling. In the results section, real-time simulations on OPAL-RT are performed to test the resilience of two parallel inverters with increasing blockchain complexity.

Wireless sensor networks (WSNs) have gained increasing popularity in ubiquitous support of sensing system services. Often, WSNs are energy-constrained and they are deployed in harsh and unattended environments. Consequently, WSNs are vulnerable to energy and environmental factors. To ensure secure and reliable operations in safety-critical monitoring WSNs, it is important to guarantee energy-efficient communications, location privacy protection, and reliability. Fake packet-based source location privacy (SLP) protocols are known to be energy-inefficient. Therefore, in this study, we investigate the impact of energy-inefficient communications on the privacy performance of the fake packet-based SLP protocols. Experiment results show that the protocols achieve short-term and less reliable SLP protection.

The Internet obviously has a major impact on the global economy and human life every day. This boundless use pushes the attack programmers to attack the data frameworks on the Internet. Web attacks influence the reliability of the Internet and its administrations. These attacks are classified as User-to-Root (U2R), Remote-to-Local (R2L), Denial-of-Service (DoS) and Probing (Probe). Subsequently, making sure about web framework security and protecting data are pivotal. The conventional layers of safeguards like antivirus scanners, firewalls and proxies, which are applied to treat the security weaknesses are insufficient. So, Intrusion Detection Systems (IDSs) are utilized to screen PC and data frameworks for security shortcomings. IDS adds more effectiveness in securing networks against attacks. This paper presents an IDS model based on Deep Learning (DL) with Convolutional Neural Network (CNN) hypothesis. The model has been evaluated on the NSLKDD dataset. It has been trained by Kddtrain+ and tested twice, once using kddtrain+ and the other using kddtest+. The achieved test accuracies are 99.7% and 98.43% with 0.002 and 0.02 wrong alert rates for the two test scenarios, respectively.

Relational database management systems not only support larger-than-memory data processing and very advanced query optimization, but also offer the benefits of data security, privacy, and consistency. When machine learning on large data sets is processed directly on an existing SQL database server, the data does not need to be exported and transferred to a separate big data processing platform. To achieve this, we implement a linear regression algorithm using SQL code generation such that the computation can be performed server-side and directly in the RDBMs. Our method and its implementation, programmed in Python, solves linear regression (LR) using the ordinary least squares (OLS) method directly in the RDBMS using SQL code generation, leaving most of the processing in the database. Only the matrix of the system of equations, whose size is equal to the number of variables squared, is transferred from the SQL server to the Python client to be solved for OLS regression. For evaluation purposes, our LR implementation was tested with artificially generated datasets and compared to an existing Python library (Scikit Learn). We found that our implementation consistently solves OLS regression faster than Scikit Learn for datasets with more than 10,000 input rows, and if the number of columns is less than 64. Moreover, under the same test conditions where the computation is larger than memory, our implementation showed a fast result, while Scikit returned an out-of-memory error. We conclude that SQL is a promising tool for in-database processing of large-volume, low-dimensional data sets with a particular class of machine learning algorithms, namely those that can be efficiently solved with map-reduce queries such as OLS regression.

Traditional grid-centric data storage methods are vulnerable to network attacks or failures due to downtime, causing problems such as data loss or tampering. The security of data storage can be effectively improved by establishing an alliance chain. However, the existing consortium chain consensus algorithm has low scalability, and the consensus time will explode as the number of nodes increases. This paper proposes an improved consensus algorithm (MSBFT) based on multi-signature to address this problem, which spreads data by establishing a system communication tree, reducing communication and network transmission costs, and improving system scalability. By generating schnorr multi-signature as the shared signature of system nodes, the computational cost of verification between nodes is reduced. At the end of the article, simulations prove the superiority of the proposed method.

In-browser cryptomining uses the computational power of a website's visitors to mine cryptocurrency, i.e., to create new coins. With the rise of ready-to-use mining scripts distributed by service providers (e.g., Coinhive), it has become trivial to turn a website into a cryptominer by copying and pasting the mining script. Both legitimate webpage owners who want to raise an extra revenue under users' explicit consent and malicious actors who wish to exploit the computational power of the users' computers without their consent have started to utilize this emerging paradigm of cryptocurrency operations. In-browser cryptomining, though mostly abused by malicious actors in practice, is indeed a promising funding model that can be utilized by website owners, publishers, or non-profit organizations for legitimate business purposes, such as to collect revenue or donations for humanitarian projects, inter alia. However, our analysis in this paper shows that in practice, regardless of their being legitimate or not, all in-browser mining scripts are treated the same as malicious cryptomining samples (aka cryptojacking) and blacklisted by browser extensions or antivirus programs. Indeed, there is a need for a better understanding of the in-browser cryptomining ecosystem. Hence, in this paper, we present an in-depth empirical analysis of in-browser cryptomining processes, focusing on the samples explicitly asking for user consent, which we call permissioned cryptomining. To the best of our knowledge, this is the first study focusing on the permissioned cryptomining samples. For this, we created a dataset of 6269 unique web sites containing cryptomining scripts in their source codes to characterize the in-browser cryptomining ecosystem by differentiating permissioned and permissionless cryptomining samples. We believe that (1) this paper is the first attempt showing that permissioned in-browser cryptomining could be a legitimate and viable monetization tool if implemented responsibly and without interrupting the user, and (2) this paper will catalyze the widespread adoption of legitimate crvptominina with user consent and awareness.

This research covers the problem related to user behavior and its relationship with the protection of computer assets in terms of confidentiality, integrity, and availability. The main objective was to evaluate the relationship between the dimensions of awareness, compliance and appropriation of the information security culture and the asset protection variable, the ISCA diagnostic instrument was applied, and social engineering techniques were incorporated for this process. The results show the levels of awareness, compliance and appropriation of the university that was considered as a case study, these oscillate between the second and third level of four levels. Similarly, the performance regarding asset protection ranges from low to medium. It was concluded that there is a significant relationship between the variables of the investigation, verifying that of the total types of incidents registered in the study case, approximately 69% are associated with human behavior. As a contribution, an information security culture model was formulated whose main characteristic is a complementary diagnostic process between surveys and social engineering techniques, the model also includes the information security management system, risk management and security incident handling as part of the information security culture ecosystem in an enterprise.

Nearest neighbor (NN) search is an essential operation in many applications, such as one/few-shot learning and image classification. As such, fast and low-energy hardware support for accurate NN search is highly desirable. Ternary content-addressable memories (TCAMs) have been proposed to accelerate NN search for few-shot learning tasks by implementing \$L\$∞ and Hamming distance metrics, but they cannot achieve software-comparable accuracies. This paper proposes a novel distance function that can be natively evaluated with multi-bit content-addressable memories (MCAMs) based on ferroelectric FETs (Fe-FETs) to perform a single-step, in-memory NN search. Moreover, this approach achieves accuracies comparable to floating-point precision implementations in software for NN classification and one/few-shot learning tasks. As an example, the proposed method achieves a 98.34% accuracy for a 5-way, 5-shot classification task for the Omniglot dataset (only 0.8% lower than software-based implementations) with a 3-bit MCAM. This represents a 13% accuracy improvement over state-of-the-art TCAM-based implementations at iso-energy and iso-delay. The presented distance function is resilient to the effects of FeFET device-to-device variations. Furthermore, this work experimentally demonstrates a 2-bit implementation of FeFET MCAM using AND arrays from GLOBALFOUNDRIES to further validate proof of concept.