Biblio

Starting from our previous research where we in- troduced a general framework for supporting data-driven privacy-preserving big data management in distributed environments, such as emerging Cloud settings, in this paper we further and significantly extend our past research contributions, and provide several novel contributions that complement our previous work in the investigated research field. Our proposed framework can be viewed as an alternative to classical approaches where the privacy of big data is ensured via security-inspired protocols that check several (protocol) layers in order to achieve the desired privacy. Unfortunately, this injects considerable computational overheads in the overall process, thus introducing relevant challenges to be considered. Our approach instead tries to recognize the “pedigree” of suitable summary data representatives computed on top of the target big data repositories, hence avoiding computational overheads due to protocol checking. We also provide a relevant realization of the framework above, the so- called Data-dRIven aggregate-PROvenance privacy-preserving big Multidimensional data (DRIPROM) framework, which specifically considers multidimensional data as the case of interest. Extensions and discussion on main motivations and principles of our proposed research, two relevant case studies that clearly state the need-for and covered (related) properties of supporting privacy- preserving management and analytics of big data in modern distributed systems, and an experimental assessment and analysis of our proposed DRIPROM framework are the major results of this paper.

Nowadays the massive amount of data is produced from different sources and lots of applications are processing these data to discover insights. Sometimes we may get unexpected results from these applications and it is not feasible to trace back to the data origin manually to find the source of errors. To avoid this problem, data must be accompanied by the context of how they are processed and analyzed. Especially, data-intensive applications like e-Science always require transparency and therefore, we need to understand how data has been processed and transformed. In this paper, we propose mathematical formulation and implementation of query inversion techniques to trace the provenance of data in a relational database management system (RDBMS). We build mathematical formulations of inverse queries for most of the relational algebra operations and show the formula for join operations in this paper. We, then, implement these formulas of inversion techniques and the experiment shows that our proposed inverse queries can successfully trace back to original data i.e. finding data provenance.

Face detection is an application widely used in video surveillance systems and it is the first step for subsequent applications such as monitoring and recognition. For facial detection, there are a series of algorithms that allow the face to be extracted in a video image, among which are the Viola & Jones waterfall method and the method by geometric models using the Hausdorff distance. In this article, both algorithms are theoretically analyzed and the best one is determined by efficiency and resource optimization. Considering the most common problems in the detection of faces in a video surveillance system, such as the conditions of brightness and the angle of rotation of the face, tests have been carried out in 13 different scenarios with the best theoretically analyzed algorithm and its combination with another algorithm The images obtained, using a digital camera in the 13 scenarios, have been analyzed using Matlab code of the Viola & Jones and Viola & Jones algorithm combined with the Kanade-Lucas-Tomasi algorithm to add the feature of completing the tracking of a single object. This paper presents the detection percentages, false positives and false negatives for each image and for each simulation code, resulting in the scenarios with the most detection problems and the most accurate algorithm in face detection.

It is often necessary to disclose training data to the public domain, while protecting privacy of certain sensitive labels. We use information theoretic measures to develop such privacy preserving data disclosure mechanisms. Our mechanism involves perturbing the data vectors to strike a balance in the privacy-utility trade-off. We use maximal information leakage between the output data vector and the confidential label as our privacy metric. We first study the theoretical Bernoulli-Gaussian model and study the privacy-utility trade-off when only the mean of the Gaussian distributions can be perturbed. We show that the optimal solution is the same as the case when the utility is measured using probability of error at the adversary. We then consider an application of this framework to a data driven setting and provide an empirical approximation to the Sibson mutual information. By performing experiments on the MNIST and FERG data sets, we show that our proposed framework achieves equivalent or better privacy than previous methods based on mutual information.

Product quantization (PQ) and its variations are popular and attractive in approximate nearest neighbor search (ANN) due to their lower memory usage and faster retrieval speed. PQ decomposes the high-dimensional vector space into several low-dimensional subspaces, and quantizes each sub-vector in their subspaces, separately. Thus, PQ can generate a codebook containing an exponential number of codewords or indices by a Cartesian product of the sub-codebooks from different subspaces. However, when there is large variance in the average amplitude of the components of the data points, directly utilizing the PQ on the data points would result in poor performance. In this paper, we propose a new approach, namely, mean-removed product quantization (MRPQ) to address this issue. In fact, the average amplitude of a data point or the mean of a date point can be regarded as statistically independent of the variation of the vector, that is, of the way the components vary about this average. Then we can learn a separate scalar quantizer of the means of the data points and apply the PQ to their residual vectors. As shown in our comprehensive experiments on four large-scale public datasets, our approach can achieve substantial improvements in terms of Recall and MAP over some known methods. Moreover, our approach is general which can be combined with PQ and its variations.

Recent cyber attacks on the power grid have been of increasing complexity and sophistication. In order to understand the impact of cyber-attacks on the power system resiliency, it is important to consider an holistic cyber-physical system specially with increasing industrial automation. In this study, device-level resilience properties of the various controllers and their impact on the microgrid resiliency is studied. In addition, a cyber-physical resiliency metric considering vulnerabilities, system model, and device-level properties is proposed. Resiliency is defined as the system ability to provide energy to critical loads even in extreme contingencies and depends on system ability to withstand, predict, and recover. A use case is presented inspired by the recent Ukraine cyber-attack. A use case has been presented to demonstrate application of the developed cyber-physical resiliency metric to enhance situational awareness of the operator, and enable better proactive or remedial control actions to improve resiliency.

In recent years, the research community has increasingly focused on understanding the security and privacy challenges posed by deep learning models. However, the security domain and the privacy domain have typically been considered separately. It is thus unclear whether the defense methods in one domain will have any unexpected impact on the other domain. In this paper, we take a step towards enhancing our understanding of deep learning models when the two domains are combined together. We do this by measuring the success of membership inference attacks against two state-of-the-art adversarial defense methods that mitigate evasion attacks: adversarial training and provable defense. On the one hand, membership inference attacks aim to infer an individual's participation in the target model's training dataset and are known to be correlated with target model's overfitting. On the other hand, adversarial defense methods aim to enhance the robustness of target models by ensuring that model predictions are unchanged for a small area around each sample in the training dataset. Intuitively, adversarial defenses may rely more on the training dataset and be more vulnerable to membership inference attacks. By performing empirical membership inference attacks on both adversarially robust models and corresponding undefended models, we find that the adversarial training method is indeed more susceptible to membership inference attacks, and the privacy leakage is directly correlated with model robustness. We also find that the provable defense approach does not lead to enhanced success of membership inference attacks. However, this is achieved by significantly sacrificing the accuracy of the model on benign data points, indicating that privacy, security, and prediction accuracy are not jointly achieved in these two approaches.

SQL injection is among the most dangerous vulnerabilities in web applications that allow attackers to bypass the authentication and access the application database. Security testing is one of the techniques required to detect the existence of SQL injection vulnerability in a web application. However, inadequate test data during testing can affect the effectiveness of security testing. Therefore, in this paper, the new algorithm is designed and developed by applying the Cartesian Product technique in order to generate a set of invalid test data automatically. A total of 624 invalid test data were generated in order to increase the detection rate of SQL injection vulnerability. Finally, the ideas obtained from our method is able to detect the vulnerability of SQL injection in web application.

On-Load Tap Changing transformers are a widely used voltage regulation device. In the context of modern or smart grids, the control signals, i.e., the tap change commands are sent through SCADA channels. It is well known that the power system SCADA networks are prone to attacks involving injection of false data or commands. While false data injection is well explored in existing literature, attacks involving malicious control signals/commands are relatively unexplored. In this paper, an algorithm is developed to detect a stealthily introduced malicious tap change command through a compromised SCADA channel. This algorithm is based on the observation that a stealthily introduced false data or command masks the true estimation of only a few state variables. This leaves the rest of the state variables to show signs of a change in system state brought about by the attack. Using this observation, an index is formulated based on the ratios of injection or branch currents to voltages of the terminal nodes of the tap changers. This index shows a significant increase when there is a false tap command injection, resulting in easy classification from normal scenarios where there is no attack. The algorithm is computationally light, easy to implement and reliable when tested extensively on several tap changers placed in an IEEE 118-bus system.

Internet of Things (IoT) is rapidly emerging as the manifestation of the networked society vision. But its centralized architecture will lead to a single point of failure. On the other hand, it will be difficult to handle communications in the near future considering the rapid growth of IoT devices. Along with its popularity, IoT suffers from a lot of vulnerabilities, which IoT developers are constantly working to mitigate. This paper proposes a new protocol called Mirage which can be used for secure and decentralized communication of IoT devices. This protocol is built based on security principles. Out of which Mirage mainly focuses on authentication, integrity, and non-repudiation. In this protocol, devices are authenticated via secret keys known only to the parties involved in the communication. These secret keys are not static and will be constantly changing for every communication. For ensuring integrity, an intermediary is asked to exchange the hash of the messages. As the intermediary nodes are lending their computing and networking powers, they should be rewarded. To ensure non-repudiation, instead of going for trusted third parties, blockchain technology is used. Every node in the network needs to spend a mirage token for sending a message. Mirage tokens will be provided only to those nodes, who help in exchanging the hashes as a reward. In the end, a decentralized network of IoT devices is formed where every node contribute to the security of the network.

With the rapid proliferation of Internet of Thing (IoT) devices, many security challenges could be introduced at low-end routers. Misbehaving routers affect the availability of the networks by dropping packets selectively and rejecting data forwarding services. Although existing Reputation Management (RM) systems are useful in identifying misbehaving routers, the centralized nature of the RM center has the risk of one-point failure. The emerging blockchain techniques, with the inherent decentralized consensus mechanism, provide a promising method to reduce this one-point failure risk. By adopting the distributed consensus mechanism, we propose a blockchain-based reputation management system in IoT networks to overcome the limitation of centralized router RM systems. The proposed solution utilizes the blockchain technique as a decentralized database to store router reports for calculating reputation of each router. With the proposed reputation calculation mechanism, the reliability of each router would be evaluated, and the malicious misbehaving routers with low reputations will be blacklisted and get isolated. More importantly, we develop an optimized group mining process for blockchain technique in order to improve the efficiency of block generation and reduce the resource consumption. The simulation results validate the distributed blockchain-based RM system in terms of attacks detection and system convergence performance, and the comparison result of the proposed group mining process with existing blockchain models illustrates the applicability and feasibility of the proposed works.

This paper presents a general model to assess the mixed-degradation profiles of critical components in a Cyber-Physical System (CPS) based on the reliability of its critical physical and software components. In the proposed assessment, the cyber aspect of a CPS was approached from a software reliability perspective. Although extensive research has been done on physical components degradation and software reliability separately, research for the combined physical-software systems is still scarce. The non-homogeneous Poisson Processes (NHPP) software reliability models are deemed to fit well with the real data and have descriptive and predictive abilities, which could make them appropriate to estimate software components reliability. To show the feasibility of the proposed approach, a case study for mixed-degradation profiles assessment is presented with n physical components and one major software component forming a critical subsystem in CPS. Two physical components were assumed to have different degradation paths with the dependency between them. Series and parallel structures were investigated for physical components. The software component failure data was taken from a wireless network switching center and fitted into a Weibull software reliability model. The case study results revealed that mix-degradation profiles of physical components, combined with software component profile, produced a different CPS reliability profile.

Taking care of security at the cloud is a major issue that needs to be carefully considered and solved for both individuals as well as organizations. Organizations usually expect more trust from employees as well as customers in one hand. On the other hand, cloud users expect their private data is maintained and secured. Although this must be case, however, some malicious outsiders of the cloud as well as malicious insiders who are cloud internal users tend to disclose private data for their malicious uses. Although outsiders of the cloud should be a concern, however, the more serious problems come from Insiders whose malicious actions are more serious and sever. Hence, insiders' threats in the cloud should be the top most problem that needs to be tackled and resolved. This paper aims to find a proper solution for the insider threat problem in the cloud. The paper presents a Mobile Edge Computing (MEC) mitigation model as a solution that suits the specialized nature of this problem where the solution needs to be very close to the place where insiders reside. This in fact gives real-time responses to attack, and hence, reduces the overhead in the cloud.

In the article the problem of survivability evaluation of control systems is considered. Control system is presented as a graph with edges that formalize minimal control systems consist of receiver, transmitter and a communication line connecting them. Based on the assumption that the survivability of minimal control systems is known, the mathematical model of survivability evaluation of not minimal control systems based on fuzzy logic is offered.

An advanced persistent threat (APT) attack is a powerful cyber-weapon aimed at the specific targets in cyberspace. The sophisticated attack techniques, long dwell time and specific objectives make the traditional defense mechanism ineffective. However, most existing studies fail to consider the theoretical modeling of the whole APT attack. In this paper, we mainly establish a theoretical framework to characterize an information-based APT attack on the internal network. In particular, our mathematical framework includes the initial entry model for selecting the entry points and the targeted attack model for studying the intelligence gathering, strategy decision-making, weaponization and lateral movement. Through a series of simulations, we find the optimal candidate nodes in the initial entry model, observe the dynamic change of the targeted attack model and verify the characteristics of the APT attack.

In recent years, technology use has assumed an important role in the support of human activities. Intellectual work has become the main preferred human activity, while structured activities are going to become ever more automatized for increasing their efficiency. For this reason, we assist to the diffusion of ever more innovative devices able to face new emergent problems. These devices can interact with the environment and each other autonomously, taking decisions even without human control. This is the Internet of Things (IoT) phenomenon, favored by low cost, high mobility, high interaction and low power devices. This spread of devices has become uncontrolled, but security in this context continues to increase slowly. The purpose of this work is to model and evaluate a new IoT security system. The context is based on a generic IoT system in the presence of lightweight actuator and sensor nodes exchanging messages through Message Queue Telemetry Transport (MQTT) protocol. This work aims to increase the security of this protocol at application level, particularly mitigating Denial of Service (DoS) attacks. The system is based on the use of a host Intrusion Detection System (IDS) which applies a threshold based packet discarding policy to the different topics defined through MQTT.

The article considers some aspects of modeling information security circuits for information and communication systems used in Smart City. As a basic research paradigm, the postulates of game theory and mathematical dependencies based on Markov processes were used. Thus, it is possible to sufficiently substantively describe the procedure for selecting rational variants of cyber security systems used to protect information technologies in Smart City. At the same time, using the model proposed by us, we can calculate the probability of cyber threats for the Smart City systems, as well as the cybernetic risks of diverse threats. Further, on the basis of the described indicators, rational contour options are chosen to protect the information systems used in Smart City.

Cyber-Physical Systems are distributed, heterogeneous systems that communicate and exchange data over networks. This creates semantic dependencies between the individual components. In the event of an error, it is difficult to identify the source of an occurring error that is spread due to those underlying dependencies. Tools such as the Information Flow Monitor solve this problem, but require compliance with a protocol. Nodes that do not adhere to this protocol prevent errors from being tracked. In this paper, we present a way to bridge these black-box nodes with a dependency model and to still be able to use them in monitoring tools.

In the context of Advanced Persistent Threat (APT) attacks, this paper introduces a model, called Nuke, which tries to provide a more operational reading of the attackers' lifecycle in a compromised network. It allows to consider the notions of regression; and repetitiveness of final objectives achievement. By confronting this model with examples of recent attacks (Equifax data breach and TV5Monde sabotage), we emphasize the importance of the attack chronology in the Cyber Threat Intelligence (CTI) reports, as well as the Tactics, Techniques and Procedures (TTP) used by the attacker during his progression.

Oscillating-water-column (OWC) devices are a very important type of wave energy converters which have been extensively studied over the years. Although most designs of OWC are based on floating or fixed structures exposed above the surface level, little is known from completely submerged systems which can benefit from reduced environmental loads and a simplified structural design. The submerged type of resonant duct consists of two OWCs separated by a weir and air chamber instead of the commonly used single column. Under conditions close to resonance, water flows from the first column into the second one, resulting in a positive flow through the system from which energy can be extracted by a hydro turbine. While existing work has looked at the study of the behaviour of one OWC, this paper addresses the dynamic interaction between the two water columns including the mass transfer mechanism as well as the associated change of momentum. A numerical time-domain model is used to obtain some initial results on the performance and response of the system for different design parameters. The model is derived from 1D conservation of mass and momentum equations, including hydrodynamic effects, adiabatic air compressibility and turbine induced damping. Preliminary results indicate that the mass transfer has an important effect both on the resonance amplification and on the phase between the motion of the two columns. Simulation results are presented for the system performance over several weir heights and regular wave conditions. Further work will continue in design optimization and experimental validation of the proposed model.

With the emergence of cloud, content-based image retrieval (CBIR) on encrypted domain gain enormous importance due to the ever increasing need for ensuring confidentiality, authentication, integrity and privacy of data. CBIR on outsourced encrypted images can be done by extracting features from unencrypted images and generating searchable encrypted index based on it. Visual descriptors like color descriptors, shape and texture descriptors, etc. are employed for similarity search. Since visual descriptors used to represent an image have crucial role in retrieving most similar results, an attempt to combine them has been made in this paper. The effect of combining different visual descriptors on retrieval precision in secure CBIR scheme proposed by Xia et al. is analyzed. Experimental results show that combining visual descriptors can significantly enhance retrieval precision of the secure CBIR scheme.

With the rise of blockchain technology, peer-to-peer network system has once again caught people's attention. Peer-to-peer (P2P) is currently being implemented on various kind of decentralized systems such as InterPlanetary File System (IPFS). However, P2P file sharing network systems is not without its flaws. Data stored in the other nodes cannot be deleted by the owner and can only be deleted by other nodes themselves. Ensuring that personal data can be completely removed is an important issue to comply with the European Union's General Data Protection Regulation (GDPR) criteria. To improve P2Ps privacy and security, we propose a monitorable peer-to-peer file sharing mechanism that synchronizes with other nodes to perform file deletion and to generate the File Authentication Code (FAC) of each IPFS nodes in order to make sure the system synchronized correctly. The proposed mechanism can integrate with a consortium Blockchain to comply with GDPR.

The objective of this paper is to propose a distributed intrusion detection model based on a multi agent system. Mutli Agent Systems (MAS) are very suitable for intrusion detection systems as they meet the characteristics required by the networks and Big Data issues. The MAS agents cooperate and communicate with each other to ensure the effective detection of network intrusions without the intervention of an expert as used to be in the classical intrusion detection systems relying on signature matching to detect known attacks. The proposed model helped to detect known and unknown attacks within big computer infrastructure by responding to the network requirements in terms of distribution, autonomy, responsiveness and communication. The proposed model is capable of achieving a good and a real time intrusion detection using multi-agents paradigm and Hadoop Distributed File System (HDFS).

With the advent of fifth generation (5G) network and increasingly powerful mobile devices, people can conveniently obtain network resources wherever they are and whenever they want. However, the problem of mobility support in current network has not been adequately solved yet, especially in inter-domain mobile scenario, which leads to poor experience for mobile consumers. MobilityFirst is a clean slate future Internet architecture which adopts a clean separation between identity and network location. It provides new mechanisms to address the challenge of wireless access and mobility at scale. However, MobilityFirst lacks effective ways to deal with multicast service over mobile networks. In this paper, we design an efficient multicast mechanism based on MobilityFirst architecture and present the deployment in current network at scale. Furthermore, we propose a hierarchical multicast packet header with additional destinations to achieve low-cost dynamic multicast routing and provide solutions for both the multicast source and the multicast group members moving in intra- or inter-domain. Finally, we deploy a multicast prototype system to evaluate the performance of the proposed multicast mechanism.

In order to accelerate the image processing speed, in this paper, a multi-core heterogeneous computing technology based on the Xilinx Zynq platform is proposed. The proposed technique could accelerate the real-time video image processing system through hardware acceleration. In order to verify the proposed technique, an Otsu binarized hardware-accelerated IP is designed in FPGA and interacts with ARM through the AXI bus. Compared with the existing homogeneous architecture processor computing, the image processing speed of the proposed technique with multi-core heterogeneous acceleration processing is significantly accelerated.