Biblio

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.

The human face conveys a significant amount of information. Through facial expressions, the face is able to communicate numerous sentiments without the need for verbalisation. Visual emotion recognition has been extensively studied. Recently several end-to-end trained deep neural networks have been proposed for this task. However, such models often lack generalisation ability across datasets. In this paper, we propose the Deep Facial Expression Vector ExtractoR (DeepFEVER), a new deep learning-based approach that learns a visual feature extractor general enough to be applied to any other facial emotion recognition task or dataset. DeepFEVER outperforms state-of-the-art results on the AffectNet and Google Facial Expression Comparison datasets. DeepFEVER’s extracted features also generalise extremely well to other datasets – even those unseen during training – namely, the Real-World Affective Faces (RAF) dataset.

In power networks, it is important to detect a cyber attack. In this paper, we propose a detection method of false data injection (FDI) attacks. FDI attacks cannot be detected from the estimation error in power networks. The proposed method is based on the distributed state estimation, and is used the tentative estimated state. The proposed method is demonstrated by a numerical example on the IEEE 14-bus system.

Most of the studies of the existing object detection models are studies to better detect the objects to be detected. The problem of false detection of objects that should not be detected is not considered. When an object detection model that does not take this problem into account is applied to an industrial field close to humans, false detection can lead to a dangerous situation that greatly interferes with human life. To solve this false detection problem, this paper proposes a method of fine-tuning the backbone neural network model of the object detection model using the Outlier Exposure method and applying the class-specific uncertainty constant to the confidence score to detect the object.

Machine learning offers performance improvements and novel functionality, but its life cycle performance is understudied. In areas like cognitive communications, where systems are long-lived, life cycle trade-offs are key to system design. Herein, we consider the use of deep learning to classify spectrograms. We vary the label-space over which the network makes classifications, as may emerge with changes in use over a system’s life cycle, and compare heterogeneous transfer learning performance across label-spaces between model architectures. Our results offer an empirical example of life cycle challenges to using machine learning for cognitive communications. They evidence important trade-offs among performance, training time, and sensitivity to the order in which the label-space is changed. And they show that fine-tuning can be used in the heterogeneous transfer of spectrogram classifiers.

Polar codes have been shown to provide an effective mechanism for achieving physical-layer security over various wiretap channels. A majority of these schemes require channel state information (CSI) at the encoder for both intended receivers and eavesdroppers. In this paper, we consider a polar coding scheme for secrecy over a Gaussian wiretap channel when no CSI is available. We show that the availability of a shared keystream between friendly parties allows polar codes to be used for both secure and reliable communications, even when the eavesdropper knows a large fraction of the keystream. The scheme relies on a predetermined strategy for partitioning the bits to be encoded into a set of frozen bits and a set of information bits. The frozen bits are filled with bits from the keystream, and we evaluate the security gap when the cyclic redundancy check-aided successive cancellation list decoder is used at both receivers in the wiretap channel model.

Based on the structure of turbo-polar codes, a secure symmetric encryption scheme is proposed to enhance information transmission security in this paper. This scheme utilizes interleaving at information bits and puncturing at parity bits for several times in the encoder. Correspondingly, we need to do the converse interleaving and fill zeros accurately at punctured position. The way of interleaving and puncturing is controlled by the private key of symmetric encryption, making sure the security of the system. The security of Secure Turbo-Polar Codes (STPC) is analyzed at the end of this paper. Simulation results are given to shown that the performance and complexity of Turbo-Polar Codes have little change after symmetric encryption. We also investigate in depth the influence of different remaining parity bit ratios on Frame Error Rate (FER). At low Signal to Noise Rate (SNR), we find it have about 0.6dB advantage when remaining parity bit ratio is between 1/20 and 1/4.

Traditional DDoS attacks mainly send massive data packets through the attacking machine, consuming the network resources or server resources of the target server, making users unable to use server resources to achieve the purpose of denial of service. This type of attack is called a Flooding-based DDoS (FDDoS) attack. It has the characteristics of large traffic and suddenness. However, Low-rate DDoS (LDDoS) attack is a new type of DDoS attack. LDDoS utilize the TCP congestion control mechanism and sends periodic pulses to attack, which can seriously reduce the TCP flow throughput of the attacked link. It has the characteristics of small traffic and strong concealment. Each of these two DDoS attack methods has its own hard-to-handle characteristics, so that there is currently no particularly effective method to prevent such attacks. This paper uses time-frequency analysis to classify and filter DDoS traffic. The proposed filtering method is designed as a system in the actual environment. Experimental results show that the designed filtering algorithm can resist not only FDDoS attacks, but also LDDoS attacks.

Distributed Denial-of-Service (DDoS) attack is a long-lived attack that is hugely harmful to the Internet. In particular, the emergence of a new type of DDoS called Link Flooding Attack (LFA) makes the detection and defense more difficult. In LFA, the attacker cuts off a specific area by controlling large numbers of bots to send low-rate traffic to congest selected links. Since the attack flows are similar to the legitimate ones, traditional schemes like anomaly detection and intrusion detection are no longer applicable. Blockchain provides a new solution to address this issue. In this paper, we propose a blockchain-based LFA detection scheme, which is deployed on routers and servers in and around the area that we want to protect. Blockchain technology is used to record and share the traceroute information, which enables the hosts in the protected region to easily trace the flow paths. We implement our scheme in Ethereum and conduct simulation experiments to evaluate its performance. The results show that our scheme can achieve timely detection of LFA with a high detection rate and a low false positive rate, as well as a low overhead.

This paper for the first time investigate the security and reliability performance of ultra-reliable low-latency communication (URLLC) systems in the presence of randomly distributed eavesdroppers, where the impact of short blocklength codes and imperfect channel estimation are jointly considered. Based on the finite-blocklength information theory, we first derive a closed-form approximation of transmission error probability to describe the degree of reliability loss. Then, we also derive an asymptotic expression of intercept probability to characterize the security performance, where the impact of secrecy protected zone is also considered. Simulation and numerical results validate the accuracy of theoretical approximations, and illustrate the tradeoff between security and reliability. That is, the intercept probability of URLLC systems can be suppressed by loosening the reliability requirement, and vice versa. More importantly, the theoretical analysis and methodologies presented in this paper can offer some insights and design guidelines for supporting secure URLLC applications in the future 6G wireless networks.

This paper studies the secure decentralized Pliable Index CODing (PICOD) problem, where the security constraint forbids users to decode more than one message while the decentralized setting imposes that there is no central transmitter in the system, and thus transmissions occur only among users. A converse bound from the Authors' previous work showed a factor of three difference in optimal code-length between the centralized and the decentralized versions of the problem, under the constraint of linear encoding. This paper first lists all linearly infeasible cases, that is, problems where no linear code can simultaneously achieve both correctness/decodability and security. Then, it proposes linear coding schemes for the remaining cases and shows that their code-length is to within an additive constant gap from the converse bound.

This paper presents a design concept of an innovative CAPTCHA that can filter the color-vision–recognition states of different users. It can simultaneously verify the real-human-user identity, differentiate between the color-vision needs, and decide the content to be presented automatically.

Lightning Network (LN) addresses the scalability problem of Bitcoin by leveraging off-chain transactions. Nevertheless, it is not possible to run LN on resource-constrained IoT devices due to its storage, memory, and processing requirements. Therefore, in this paper, we propose an efficient and secure protocol that enables an IoT device to use LN's functions through a gateway LN node. The idea is to involve the IoT device in LN operations with its digital signature by replacing original 2-of-2 multisignature channels with 3-of-3 multisignature channels. Our protocol enforces the LN gateway to request the IoT device's cryptographic signature for all operations on the channel. We evaluated the proposed protocol by implementing it on a Raspberry Pi for a toll payment scenario and demonstrated its feasibility and security.

Blockchain started with Bitcoin, but it is higher than Bitcoin. With the deepening of applied research on blockchain technology, this new technology has brought new vitality to many industries. People admire the decentralized nature of the blockchain and hope to solve the problems caused by the operation of traditional centralized institutions in a more fair and effective way. Of course, as an emerging technology, blockchain has many areas for improvement. This article explains the blockchain technology from many aspects. Starting from the typical architecture of the blockchain, the data structure and system model of the blockchain are first introduced. Then it expounds the development of consensus algorithms and compares typical consensus algorithms. Later, the focus will be on smart contracts and their application platforms. After analyzing some of the challenges currently faced by the blockchain technology, some scenarios where the blockchain is currently developing well are listed. Finally, it summarizes and looks forward to the blockchain technology.

Bitcoin is a representative cryptocurrency system using a permissionless peer-to-peer (P2P) network as its communication infrastructure. A number of attacks against Bitcoin have been discovered over the past years, including the Eclipse and EREBUS Attacks. In this paper, we present a new attack against Bitcoin’s P2P networking, dubbed ConMan because it leverages connection manipulation. ConMan achieves the same effect as the Eclipse and EREBUS Attacks in isolating a target (i.e., victim) node from the rest of the Bitcoin network. However, ConMan is different from these attacks because it is an active and deterministic attack, and is more effective and efficient. We validate ConMan through proof-of-concept exploitation in an environment that is coupled with real-world Bitcoin node functions. Experimental results show that ConMan only needs a few minutes to fully control the peer connections of a target node, which is in sharp contrast to the tens of days that are needed by the Eclipse and EREBUS Attacks. Further, we propose several countermeasures against ConMan. Some of them would be effective but incompatible with the design principles of Bitcoin, while the anomaly detection approach is positively achievable. We disclosed ConMan to the Bitcoin Core team and received their feedback, which confirms ConMan and the proposed countermeasures.

Bitcoin is a virtual encrypted digital currency based on a peer-to-peer network. In recent years, for higher anonymity, more and more Bitcoin users try to use Tor hidden services for identity and location hiding. However, previous studies have shown that Tor are vulnerable to traffic fingerprinting attack, which can identify different websites by identifying traffic patterns using statistical features of traffic. Our work shows that traffic fingerprinting attack is also effective for the Bitcoin hidden nodes detection. In this paper, we proposed a novel lightweight Bitcoin hidden service traffic fingerprinting, using a random decision forest classifier with features from TLS packet size and direction. We test our attack on a novel dataset, including a foreground set of Bitcoin hidden node traffic and a background set of different hidden service websites and various Tor applications traffic. We can detect Bitcoin hidden node from different Tor clients and website hidden services with a precision of 0.989 and a recall of 0.987, which is higher than the previous model.

Because Bitcoin P2P networking is permissionless by the application requirement, it is vulnerable against networking threats based on identity/credential manipulations such as Sybil and spoofing attacks. The current Bitcoin implementation keeps track of its peer's networking misbehaviors through ban score. In this paper, we investigate the security problems of the ban-score mechanism and discover that the ban score is not only ineffective against the Bitcoin Message-based DoS attacks but also vulnerable to a Defamation attack. In the Defamation attack, the network adversary can exploit the ban-score mechanism to defame innocent peers.

When biometric databases grow larger, a security breach or leak can affect millions. In order to protect against such a threat, the use of encryption is a natural choice. However, a biometric identification attempt then requires the decryption of a potential huge database, making a traditional approach potentially unfeasible. The use of selective JPEG2000 encryption can reduce the encryption’s computational load and enable a secure storage of biometric sample data. In this paper we will show that selective encryption of face biometric samples is secure. We analyze various encoding settings of JPEG2000, selective encryption parameters on the "Labeled Faces in the Wild" database and apply several traditional and deep learning based face recognition methods.

We introduce the novel concept of feature popularity with three different web attacks and big data from the CSE-CIC-IDS2018 dataset: Brute Force, SQL Injection, and XSS web attacks. Feature popularity is based upon ensemble Feature Selection Techniques (FSTs) and allows us to more easily understand common important features between different cyberattacks, for two main reasons. First, feature popularity lists can be generated to provide an easy comprehension of important features across different attacks. Second, the Jaccard similarity metric can provide a quantitative score for how similar feature subsets are between different attacks. Both of these approaches not only provide more explainable and easier-to-understand models, but they can also reduce the complexity of implementing models in real-world systems. Four supervised learning-based FSTs are used to generate feature subsets for each of our three different web attack datasets, and then our feature popularity frameworks are applied. For these three web attacks, the XSS and SQL Injection feature subsets are the most similar per the Jaccard similarity. The most popular features across all three web attacks are: Flow\_Bytes\_s, FlowİAT\_Max, and Flow\_Packets\_s. While this introductory study is only a simple example using only three web attacks, this feature popularity concept can be easily extended, allowing an automated framework to more easily determine the most popular features across a very large number of attacks and features.

Adversarial machine learning, a technique which seeks to deceive machine learning (ML) models, threatens the utility and reliability of ML systems. This is particularly relevant in critical ML implementations such as those found in Network Intrusion Detection Systems (NIDS). This paper considers the impact of adversarial influence on NIDS and proposes ways to improve ML based systems. Specifically, we consider five feature robustness metrics to determine which features in a model are most vulnerable, and four defense methods. These methods are tested on six ML models with four adversarial sample generation techniques. Our results show that across different models and adversarial generation techniques, there is limited consistency in vulnerable features or in effectiveness of defense method.

Cloud computing has become ubiquitous in the modern world and has offered a number of promising and transformative technological opportunities. However, organizations that use cloud platforms are also concerned about cloud security and new threats that arise due to cloud adoption. Digital forensic investigations (DFI) are undertaken when a security incident (i.e., successful attack) has been identified. Forensics data collection is an integral part of DFIs. This paper presents results from a survey of existing literature on challenges related to forensics data collection in cloud. A taxonomy of major challenges was developed to help organizations understand and thus better prepare for forensics data collection.

The rapid development of blockchain application technology promotes continuous exploration in the field of computer application science. Although it is still in the initial stage of development, the technical features of blockchain technology such as decentralization, identity verification, tamper resistance, data integrity, and security are regarded as excellent solutions to today's computer security technical problems. In this paper, we will analyze and compare blockchain data storage and cloud data processing technologies, focusing on the concept and technology of blockchain distributed data storage technology, and analyze and summarize the key issues. The results of this paper will provide a useful reference for the application and research of blockchain technology in cloud storage security.

In view of the continuous progress of current science and technology, cloud computing has been widely used in various fields. This paper proposes a secure data storage architecture based on cloud computing. The architecture studies the security issues of cloud computing from two aspects: data storage and data security, and proposes a data storage mode based on Cache and a data security mode based on third-party authentication, thereby improving the availability of data, from data storage to transmission. Corresponding protection measures have been established to realize effective protection of cloud data.

The subgraph matching problem arises in a number of modern machine learning applications including segmented images and meshes of 3D objects for pattern recognition, bio-chemical reactions and security applications. This graph-based problem can have a very large and complex solution space especially when the world graph has many more nodes and edges than the template. In a real use-case scenario, analysts may need to query additional information about template nodes or world nodes to reduce the problem size and the solution space. Currently, this query process is done by hand, based on the personal experience of analysts. By analogy to the well-known active learning problem in machine learning classification problems, we present a machine-based active learning problem for the subgraph match problem in which the machine suggests optimal template target nodes that would be most likely to reduce the solution space when it is otherwise overly large and complex. The humans in the loop can then include additional information about those target nodes. We present some case studies for both synthetic and real world datasets for multichannel subgraph matching.

Human-automation shared control is proposed to reduce the risk of driver disengagement in Level-3 autonomous vehicles. Although previous studies have approved shared control strategy is effective to keep a driver in the loop and improve the driver's performance, over- and under-trust may affect the cooperation between the driver and the automation system. This study conducted a human-in-the-loop driving simulation experiment to assess the effects of trust on driver's behavior of shared control. An expert shared control strategy with longitudinal and lateral driving assistance was proposed and implemented in the experiment platform. Based on the experiment (N=24), trust in shared control was evaluated, followed by a correlation analysis of trust and behaviors. Moderating effects of trust on the relationship between gaze focalization and minimum of time to collision were then explored. Results showed that self-reported trust in shared control could be evaluated by three subscales respectively: safety, efficiency and ease of control, which all show stronger correlations with gaze focalization than other behaviors. Besides, with more trust in ease of control, there is a gentle decrease in the human-machine conflicts of mean brake inputs. The moderating effects show trust could enhance the decrease of minimum of time to collision as eyes-off-road time increases. These results indicate over-trust in automation will lead to unsafe behaviors, particularly monitoring behavior. This study contributes to revealing the link between trust and behavior in the context of human-automation shared control. It can be applied in improving the design of shared control and reducing risky behaviors of drivers by further trust calibration.