Biblio

Compute-intensive platforms such as micro-servers and embedded systems have already undergone a shift from a single-chip to multichip architecture to achieve better yield and lower cost. However, performance of multichip systems is limited by the latency and power-hungry chip-to-chip wired I/Os. On the other hand, wireless interconnections are emerging as an energy-efficient and low latency interconnect solution for such multichip systems as it can mask long multi-hop off-chip wired I/O communication. Despite efficient communication, the unguided on and off-chip wireless communication introduce security vulnerabilities in the system. In this work, we propose a reconfigurable, secure millimeter-wave (mm-Wave) wireless interconnection architecture (AReS) for multichip systems capable of detecting and defending against emerging threats including Hardware Trojans (HTs) and Denial-of-Service (DoS) using a Machine Learning (ML)-based approach. The ML-based approach is used to classify internal and external attack to enable the required defense mechanism. To serve this purpose, we design a reconfigurable Medium Access Control (MAC) and a suitable communication protocol to enable sustainable communication even under jamming attack from both internal and external attackers. The proposed architecture also reuses the in-built test infrastructure to detect and withstand a persistent jamming attack in a wireless multichip system. Through simulation, we show that, the proposed wireless interconnection can sustain chip-to-chip communication even under persistent jamming attack with an average 1.44xand 1.56x latency degradation for internal and external attacks respectively for application-specific traffic.

Application programs are a possible source of attacks to databases as attackers might exploit vulnerabilities in a privileged database application. They can perform code injection or code-reuse attack in order to steal sensitive data. However, as such attacks very often result in changes in the program's behavior, program monitoring techniques represent an effective defense to detect on-going attacks. One such technique is monitoring the library/system calls that the application program issues while running. In this paper, we propose AD-PROM, an Anomaly Detection system that aims at protecting relational database systems against malicious/compromised applications PROgraMs aiming at stealing data. AD-PROM tracks calls executed by application programs on data extracted from a database. The system operates in two phases. The first phase statically and dynamically analyzes the behavior of the application in order to build profiles representing the application's normal behavior. AD-PROM analyzes the control and data flow of the application program (i.e., static analysis), and builds a hidden Markov model trained by the program traces (i.e., dynamic analysis). During the second phase, the program execution is monitored in order to detect anomalies that may represent data leakage attempts. We have implemented AD-PROM and carried experimental activities to assess its performance. The results showed that our system is highly accurate in detecting changes in the application programs' behaviors and has very low false positive rates.

As the demand of online data availability increases for sharing data, business analytics, security of available data becomes important issue, data needs to be protected from unauthorized access as well as it needs to provide authority that the data is received from a trusted owner. To provide owners identity digital watermarking technique is used since long time for multimedia data. This paper proposed a technique which supports watermarking on database as most of the data available today is in database format. The characters to be entered as watermark are converted into binary values; these binary values are hidden in the database using space character. Each bit is hidden in each tuple randomly. Ant colony optimization algorithm is proposed to select tuples where watermark bits are inserted. The proposed system is enhanced in terms of security due to use of ant colony optimization and resilient because even if some bits are modified the hidden text remains almost same.

Lightweight ciphers are algorithms with low computational and spacial complexity. In the modern world of miniaturization, a lightweight cipher is used in constrained devices such as RFID tags, fire and security detectors, devices for wireless communications and other IoT devices. Stream ciphers are symmetric ciphers which encrypts the plain text bit stream with corresponding key stream to generate cipher text. Hence a stream cipher with low computational complexity and maximum security can be termed as a lightweight stream cipher. Many light weight stream ciphers are already existing. Each has its own vulnerabilities and spacial requirement. This paper has successfully developed, implemented, and analyzed a lightweight stream cipher - A4. Along with low computational cost, A4 also ensures paramount security and is less prone to the emerging cryptographic attacks.

The following topics are dealt with: authorisation; data privacy; mobile computing; security of data; cryptography; Internet of Things; message authentication; invasive software; Android (operating system); vectors.

Expressing Keystroke Dynamics (KD) in form of sound opens new avenues to apply sound analysis techniques on KD. However this mapping is not straight-forward as varied feature space, differences in magnitudes of features and human interpretability of the music bring in complexities. We present a musical interface to KD by mapping keystroke features to music features. Music elements like melody, harmony, rhythm, pitch and tempo are varied with respect to the magnitude of their corresponding keystroke features. A pitch embedding technique makes the music discernible among users. Using the data from 30 users, who typed fixed strings multiple times on a desktop, shows that these auditory signals are distinguishable between users by both standard classifiers (SVM, Random Forests and Naive Bayes) and humans alike.

In the present scenario, security is the biggest concern in any domain of applications. The latest and widely used system for user authentication is a biometric system. This includes fingerprint recognition, retina recognition, and voice recognition. But these systems can be bypassed by masqueraders. To avoid this, a combination of these systems is used which becomes very costly. To overcome these two drawbacks keystroke dynamics were introduced in this field. Keystroke dynamics is a biometric authentication-based system on behavior, which is an automated method in which the identity of an individual is identified and confirmed based on the way and the rhythm of passwords typed on a keyboard by the individual. The work in this paper focuses on identifying the best algorithm for implementing an authentication system with the help of machine learning for user identification based on keystroke dynamics. Our proposed model which uses XGBoost gives a comparatively higher accuracy of 93.59% than the other algorithms for the dataset used.

The open and broadcast nature of wireless channels makes eavesdropping possible, leading to the inherent problem of information leakage. Inherent problems should be solved by endogenous security functions. Accordingly, wireless security problems should be resolved by channel-based endogenous security mechanisms. Firstly, this paper analyzes the endogenous security principle of the physical-layer-secret-key method. Afterward, we propose a novel conjecture that in a fast-fading environment, there must exist wireless systems where the endogenous secret key rate can match the user data rate. Moreover, the conjecture is well founded by the instantiation validation in a wireless system with BPSK inputs from the perspectives of both theoretical analysis and simulation experiments. These results indicate that it is possible to accomplish the one-time pad via endogenous secret keys in wireless communication.

The training process of deep learning model is costly. As such, deep learning model can be treated as an intellectual property (IP) of the model creator. However, a pirate can illegally copy, redistribute or abuse the model without permission. In recent years, a few Deep Neural Networks (DNN) IP protection works have been proposed. However, most of existing works passively verify the copyright of the model after the piracy occurs, and lack of user identity management, thus cannot provide commercial copyright management functions. In this paper, a novel user fingerprint management and DNN authorization control technique based on backdoor is proposed to provide active DNN IP protection. The proposed method can not only verify the ownership of the model, but can also authenticate and manage the user's unique identity, so as to provide a commercially applicable DNN IP management mechanism. Experimental results on CIFAR-10, CIFAR-100 and Fashion-MNIST datasets show that the proposed method can achieve high detection rate for user authentication (up to 100% in the three datasets). Illegal users with forged fingerprints cannot pass authentication as the detection rates are all 0 % in the three datasets. Model owner can verify his ownership since he can trigger the backdoor with a high confidence. In addition, the accuracy drops are only 0.52%, 1.61 % and -0.65% on CIFAR-10, CIFAR-100 and Fashion-MNIST, respectively, which indicate that the proposed method will not affect the performance of the DNN models. The proposed method is also robust to model fine-tuning and pruning attacks. The detection rates for owner verification on CIFAR-10, CIFAR-100 and Fashion-MNIST are all 100% after model pruning attack, and are 90 %, 83 % and 93 % respectively after model fine-tuning attack, on the premise that the attacker wants to preserve the accuracy of the model.

GDSII layouts of IP-confidential products are heavily controlled and access is only granted to certain privileged personnel. Failure analysts are generally excluded. Without guidance from GDSII, failure analysis, specifically physical inspection based on fault isolation findings cannot proceed. To overcome this challenge, we develop an automated approach that enables image snapshots relevant to failure analysts to be furnished without compromising the confidentiality of the GDSII content in this paper. Modules built are executed to trace the suspected nets and extract them into multiple images of different pre-defined frame specifications to facilitate failure analysis.

Files in the backup of iOS devices can be a potential source of evidentiary data. Particularly, the iOS backup (obtained through a logical acquisition technique) is widely used by many forensic tools to sift through the data. A significant challenge faced by several forensic tool developers is the changes in the data organization of the iOS backup. This is due to the fact that the iOS operating system is frequently updated by Apple Inc. Many iOS application developers release periodical updates to iOS mobile applications. Both these reasons can cause significant changes in the way user data gets stored in the iOS backup files. Moreover, approximately once every couple years, there could be a major iOS release which can cause the reorganization of files and folders in the iOS backup. Directories in the iOS backup contain SQLite databases, plist files, XML files, text files, and media files. Android/iOS devices generally use SQLite databases since it is a lightweight database. Our focus in this paper is to analyze the SQLite schema evolution specific to iOS and assist forensic tool developers in keeping their tools compatible with the latest iOS version. Our recommendations for updating the forensic data extraction tools is based on the observation of schema changes found in successive iOS versions.

Cloud federation is the evolution of modern cloud computing. It provides better resource-sharing, perfect resource-utilization, and load-balancing. However, the heterogeneity of security policies and configurations between cloud service providers makes it hard for users to totally trust them. Further, the severe impact of modern cloud attacks such as cross-side channels on federated environments is a major roadblock against such evolution. Securing users' capsules (Virtual Machines and containers) against cross-side channel attacks is considered as a big challenge to cloud service providers. Moving-target Defense (MtD) by live capsule migration was introduced as an effective mechanism to overcome such challenge. However, researchers noted that even with MtD, migrated capsules can still be tracked via routing information. In this paper, we propose a novel Blockchain-based routing mechanism to enable trace-resistant Moving-target Defence (BMtD) to enable anonymous live cross-cloud migrations of running capsules in federated cloud environments. Exploiting the Vulnerable, Exposed, Attacked, Recovered (VEAR) model, simulation results demonstrated the effectiveness of BMtD in minimizing viral attack dispersion.

The main method for long-distance underwater communication is underwater acoustic communication(UAC). The bandwidth of UAC channel is narrow and the frequency band resources are scarce. Therefore, it is important to improve the frequency band utilization of UAC system. Cognitive underwater acoustic (CUA) technology is an important method. CUA network can share spectrum resources with the primary network. Spectrum sensing (SS) technology is the premise of realizing CUA. Therefore, improving the accuracy of spectral sensing is the main purpose of this paper. However, the realization of underwater SS technology still faces many difficulties. First, underwater energy supplies are scarce, making it difficult to apply complex algorithms. Second, and more seriously, CUA network can sometimes be attacked and exploited by hostile forces, which will not only lead to data leakage, but also greatly affect the accuracy of SS. In order to improve the utilization of underwater spectrum and avoid attack, an underwater spectrum sensing model based on the two-threshold energy detection method and K of M fusion decision method is established. Then, the trust mechanism based on beta function and XOR operation are proposed to combat individual attack and multi-user joint attack (MUJA) respectively. Finally, simulation result shows the effectiveness of these methods.

Cloud computing is the flexible platform to outsource the data from local server to commercial cloud. However cloud provides tremendous benefits to user, data privacy and data leakage reduce the attention of cloud. For protecting data privacy and reduce data leakage various techniques has to be implemented in cloud. There are various types of cloud environment, but we concentrate on Hybrid cloud. Hybrid cloud is nothing but combination of more than two or more cloud. Where critical operations are performed in private cloud and non critical operations are performed in public cloud. So, it has numerous advantages and criticality too. In this paper, we focus on data security through encryption scheme over Hybrid Cloud. There are various encryption schemes are close to us but it also have data security issues. To overcome these issues, Attribute Based Encryption Scheme with Dynamic Attributes Supporting (ABE-DAS) has proposed. Attribute based Encryption Scheme with Dynamic Attributes Supporting technique enhance the security of the data in hybrid cloud.

With the concept of “wireless city”, 3G, WIFI and other wireless network coverages have become more extensive. Data transmission rate has achieved a qualitative leap, providing feasibility for the implementation of mobile video surveillance solutions. The mobile video monitoring system based on the bee population optimization algorithm proposed in this paper makes up for the defects of traditional network video surveillance, and according to the video surveillance system monitoring command, the optimal visual effect of the current state of the observed object can be rendered quickly and steadily through the optimization of the camera linkage model and simulation analysis.

Crime scene investigation and preservation are fundamentally the pillars of forensics. Numerous cases have been discussed in this paper where mishandling of evidence or improper investigation leads to lengthy trials and even worse incorrect verdicts. Whether the problem is lack of training of first responders or any other scenario, it is essential for police officers to properly preserve the evidence. Second problem is the criminal profiling where each district department has its own method of storing information about criminals. ARCRIME intends to digitally transform the way police combat crime. It will allow police officers to create a copy of the scene of crime so that it can be presented in courts or in forensics labs. It will be in the form of wearable glasses for officers on site whereas officers during training will be wearing a headset. The trainee officers will be provided with simulations of cases which have already been resolved. Officers on scene would be provided with intelligence about the crime and the suspect they are interviewing. They would be able to create a case file with audio recording and images which can be digitally sent to a prosecution lawyer. This paper also explores the risks involved with ARCRIME and also weighs in their impact and likelihood of happening. Certain contingency plans have been highlighted in the same section as well to respond to emergency situations.

Application of the attention model hybrid building based on the artificial intelligence in the course of the intelligent architecture history is studied in this article. A Hadoop distributed architecture using big data processing technology which combines basic building information with the building energy consumption data for the data mining research methods, and conduct a preliminary design of a Hadoop-based public building energy consumption data mining system. The principles of the proposed model were summarized. At first, the intelligent firewall processes the decision data faster, when the harmful information invades. The intelligent firewall can monitor and also intercept the harmful information in a timelier manner. Secondly, develop a problem data processing plan, delete and identify different types of problem data, and supplement the deleted problem data according to the rules obtained by data mining. The experimental results have reflected the efficiency of the proposed model.

Cloud computing has been envisioned as a next generation information technology (IT) paradigm. The risk of losing data stored with any untrustworthy service provider is the key barrier to widespread uptake of cloud computing. This paper proposes an algebraic signature based remote data possession checking (RDPC) scheme to verify the integrity of the data stored in the cloud. This scheme integrates forward error-correcting codes to enhance the data possession guarantee, which can recover the data when a small amount of file has been deleted. The scheme allows verification without the need for the auditor to compare against the original data, which reduces the communication complexity dramatically. The storage complexity of cloud user is reduced to several bytes' information. Extensive security analysis and simulation show that the proposed scheme is highly provably secure. Finally, experiment results reveal that the computation performance is effective, and bounded by disk I/O.

The burgeoning success of deep learning has raised the security and privacy concerns as more and more tasks are accompanied with sensitive data. Adversarial attacks in deep learning have emerged as one of the dominating security threats to a range of mission-critical deep learning systems and applications. This paper takes a holistic view to characterize the adversarial examples in deep learning by studying their adverse effect and presents an attack-independent countermeasure with three original contributions. First, we provide a general formulation of adversarial examples and elaborate on the basic principle for adversarial attack algorithm design. Then, we evaluate 15 adversarial attacks with a variety of evaluation metrics to study their adverse effects and costs. We further conduct three case studies to analyze the effectiveness of adversarial examples and to demonstrate their divergence across attack instances. We take advantage of the instance-level divergence of adversarial examples and propose strategic input transformation teaming defense. The proposed defense methodology is attack-independent and capable of auto-repairing and auto-verifying the prediction decision made on the adversarial input. We show that the strategic input transformation teaming defense can achieve high defense success rates and are more robust with high attack prevention success rates and low benign false-positive rates, compared to existing representative defense methods.

The paper discusses the features of the automated process control system, defines the algorithm for installing critical updates. The main problems in the administration of a critical system have been identified. The paper presents a model for recognizing anomalies in the network traffic of an industrial information system using machine learning methods. The article considers the network intrusion dataset (raw TCP / IP dump data was collected, where the network was subjected to multiple attacks). The main parameters that affect the recognition of abnormal behavior in the system are determined. The basic mathematical models of classification are analyzed, their basic parameters are reviewed and tuned. The mathematical model was trained on the considered (randomly mixed) sample using cross-validation and the response was predicted on the control (test) sample, where the model should determine the anomalous behavior of the system or normal as the output. The main criteria for choosing a mathematical model for the problem to be solved were the number of correctly recognized (accuracy) anomalies, precision and recall of the answers. Based on the study, the optimal algorithm for recognizing anomalies was selected, as well as signs by which this anomaly can be recognized.

It has been recognized that artificial intelligence (AI) will play an important role in future societies. AI has already been incorporated in many industries to improve business processes and automation. Although the aviation industry has successfully implemented flight management systems or autopilot to automate flight operations, it is expected that full embracement of AI remains a challenge. Given the rigorous validation process and the requirements for the highest level of safety standards and risk management, AI needs to prove itself being safe to operate. This paper addresses the safety issues of AI deployment in an aviation network compatible with the Future Communication Infrastructure that utilizes heterogeneous wireless access technologies for communications between the aircraft and the ground networks. It further considers the exploitation of software defined networking (SDN) technologies in the ground network while the adoption of SDN in the airborne network can be optional. Due to the nature of centralized management in SDN-based network, the SDN controller can become a single point of failure or a target for cyber attacks. To countermeasure such attacks, an intrusion detection system utilises AI techniques, more specifically deep neural network (DNN), is considered. However, an adversary can target the AI-based intrusion detection system. This paper examines the impact of AI security attacks on the performance of the DNN algorithm. Poisoning attacks targeting the DSL-KDD datasets which were used to train the DNN algorithm were launched at the intrusion detection system. Results showed that the performance of the DNN algorithm has been significantly degraded in terms of the mean square error, accuracy rate, precision rate and the recall rate.

Hardening is the process of configuring IT systems to ensure the security of the systems' components and data they process or store. The complexity of contemporary IT infrastructures, however, renders manual security hardening and maintenance a daunting task. In many organizations, security-configuration guides expressed in the SCAP (Security Content Automation Protocol) are used as a basis for hardening, but these guides by themselves provide no means for automatically implementing the required configurations. In this paper, we propose an approach to automatically extract the relevant information from publicly available security-configuration guides for Windows operating systems using natural language processing. In a second step, the extracted information is verified using the information of available settings stored in the Windows Administrative Template files, in which the majority of Windows configuration settings is defined. We show that our implementation of this approach can extract and implement 83% of the rules without any manual effort and 96% with minimal manual effort. Furthermore, we conduct a study with 12 state-of-the-art guides consisting of 2014 rules with automatic checks and show that our tooling can implement at least 97% of them correctly. We have thus significantly reduced the effort of securing systems based on existing security-configuration guides. In many organizations, security-configuration guides expressed in the SCAP (Security Content Automation Protocol) are used as a basis for hardening, but these guides by themselves provide no means for automatically implementing the required configurations. In this paper, we propose an approach to automatically extract the relevant information from publicly available security-configuration guides for Windows operating systems using natural language processing. In a second step, the extracted information is verified using the information of available settings stored in the Windows Administrative Template files, in which the majority of Windows configuration settings is defined. We show that our implementation of this approach can extract and implement 83% of the rules without any manual effort and 96% with minimal manual effort. Furthermore, we conduct a study with 12 state-of-the-art guides consisting of 2014 rules with automatic checks and show that our tooling can implement at least 97% of them correctly. We have thus significantly reduced the effort of securing systems based on existing security-configuration guides. In this paper, we propose an approach to automatically extract the relevant information from publicly available security-configuration guides for Windows operating systems using natural language processing. In a second step, the extracted information is verified using the information of available settings stored in the Windows Administrative Template files, in which the majority of Windows configuration settings is defined. We show that our implementation of this approach can extract and implement 83% of the rules without any manual effort and 96% with minimal manual effort. Furthermore, we conduct a study with 12 state-of-the-art guides consisting of 2014 rules with automatic checks and show that our tooling can implement at least 97% of them correctly. We have thus significantly reduced the effort of securing systems based on existing security-configuration guides. We show that our implementation of this approach can extract and implement 83% of the rules without any manual effort and 96% with minimal manual effort. Furthermore, we conduct a study with 12 state-of-the-art guides consisting of 2014 rules with automatic checks and show that our tooling can implement at least 97% of them correctly. We have thus significantly reduced the effort of securing systems based on existing security-configuration guides.

With the growing numbers of security attacks causing more and more serious damages in software systems, security cannot be added as an afterthought in software development. It has to be built in from the early development phases such as requirement and design. The role responsible for designing a software system is termed an “architect”, knowledgeable about the system architecture design, but not always well-trained in security. Moreover, involving other security experts into the system design is not always possible due to time-to-market and budget constraints. To address these challenges, we propose to define an asset-based security assistance in this paper, to help architects design secure systems even if these architects have limited knowledge in security. This assistance helps alert threats, and integrate the security controls over vulnerable parts of system into the architecture model. The central concept enabling this assistance is that of asset. We apply our proposal on a telemonitoring case study to show that automating such an assistance is feasible.

As the hottest frontier technology in the field of artificial intelligence, machine learning is subverting various industries step by step. In the future, it will penetrate all aspects of our lives and become an indispensable technology around us. Among them, network security is an area where machine learning can show off its strengths. Among many network security problems, privacy protection is a more difficult problem, so it needs more introduction of new technologies, new methods and new ideas such as machine learning to help solve some problems. The research contents for this include four parts: an overview of machine learning, the significance of machine learning in network security, the application process of machine learning in network security research, and the application of machine learning in privacy protection. It focuses on the issues related to privacy protection and proposes to combine the most advanced matching algorithm in deep learning methods with information theory data protection technology, so as to introduce it into biometric authentication. While ensuring that the loss of matching accuracy is minimal, a high-standard privacy protection algorithm is concluded, which enables businesses, government entities, and end users to more widely accept privacy protection technology.

Controller Area Network is the prominent communication protocol in automotive systems. Its salient features of arbitration, message filtering, error detection, data consistency and fault confinement provide robust and reliable architecture. Despite of this, it lacks security features and is vulnerable to many attacks. One of the common attacks over the CAN communication is the replay attack. It can happen even after the implementation of encryption or authentication. This paper proposes a methodology of supressing the replay attacks by implementing authenticated encryption embedded with timestamp and pre-shared initialisation vector as a primary key. The major advantage of this system is its flexibility and configurability nature where in each layer can be chosen with the help of cryptographic algorithms to up to the entire size of the keys.