Biblio

This paper proposes AERFAD, an anomaly detection method based on the autoencoder and the random forest, for solving the credit card fraud detection problem. The proposed AERFAD first utilizes the autoencoder to reduce the dimensionality of data and then uses the random forest to classify data as anomalous or normal. Large numbers of credit card transaction data of European cardholders are applied to AEFRAD to detect possible frauds for the sake of performance evaluation. When compared with related methods, AERFAD has relatively excellent performance in terms of the accuracy, true positive rate, true negative rate, and Matthews correlation coefficient.

We address secure resource allocation for the energy harvesting (EH) based 5G cooperative cognitive radio networks (CRNs). To guarantee that the size-limited secondary users (SUs) can simultaneously send the primary user's and their own information, we assume that SUs are equipped with orthogonally dual-polarized antennas (ODPAs). In particular, we propose, develop, and analyze an efficient resource allocation scheme under a practical non-linear EH model, which can capture the nonlinear characteristics of the end-to-end wireless power transfer (WPT) for radio frequency (RF) based EH circuits. Our obtained numerical results validate that a substantial performance gain can be obtained by employing the non-linear EH model.

Opportunistic spectrum access is one of the emerging techniques for maximizing throughput in congested bands and is enabled by predicting idle slots in spectrum. We propose a kernel-based reinforcement learning approach coupled with a novel budget-constrained sparsification technique that efficiently captures the environment to find the best channel access actions. This approach allows learning and planning over the intrinsic state-action space and extends well to large state spaces. We apply our methods to evaluate coexistence of a reinforcement learning-based radio with a multi-channel adversarial radio and a single-channel carrier-sense multiple-access with collision avoidance (CSMA-CA) radio. Numerical experiments show the performance gains over carrier-sense systems.

The decisions made by machines are increasingly comparable in predictive performance to those made by humans, but these decision making processes are often concealed as black boxes. Additional techniques are required to extract understanding, and one such category are explanation methods. This research compares the explanations of two popular forms of artificial intelligence; neural networks and random forests. Researchers in either field often have divided opinions on transparency, and comparing explanations may discover similar ground truths between models. Similarity can help to encourage trust in predictive accuracy alongside transparent structure and unite the respective research fields. This research explores a variety of simulated and real-world datasets that ensure fair applicability to both learning algorithms. A new heuristic explanation method that extends an existing technique is introduced, and our results show that this is somewhat similar to the other methods examined whilst also offering an alternative perspective towards least-important features.

The presence of robots is becoming more apparent as technology progresses and the market focus transitions from smart phones to robotic personal assistants such as those provided by Amazon and Google. The integration of robots in our societies is an inevitable tendency in which robots in many forms and with many functionalities will provide services to humans. This calls for an understanding of how humans are affected by both the presence of and the reliance on robots to perform services for them. In this paper we explore the effects that robots have on humans when a service is performed on request. We expose three groups of human participants to three levels of service completion performed by robots. We record and analyse human perceptions such as propensity to trust, competency, responsiveness, sociability, and team work ability. Our results demonstrate that humans tend to trust robots and are more willing to interact with them when they autonomously recover from failure by requesting help from other robots to fulfil their service. This supports the view that autonomy and team working capabilities must be brought into robots in an effort to strengthen trust in robots performing a service.

Influence of remanent flux on hysteresis curve of the transformer core is addressed in this paper. In addition, its significant negative effect on transformer diagnostic tests is quantified based on experimental studies. Furthermore, a novel approach is proposed to efficiently and quickly eliminate the remanent flux. This approach is evaluated based on simulation studies on a 230/63-kV power transformer. Meanwhile, experimental studies are performed on both 0.2/0.2 and 20/0.4 kV transformers. These studies reveal that the approach not only is well able to eliminate the remanent flux, but also it has various advantages over the commonly used method. In addition, this approach is equally applicable for various power, distribution, and instrument transformer types.

The US Federal Communications Commission (FCC) has recently mandated the database-driven dynamic spectrum access where unlicensed secondary users search for idle bands and use them opportunistically. The database-driven dynamic spectrum access approach is regarded for minimizing any harmful interference to licensed primary users caused by RF channel sensing uncertainties. However, when several secondary users (or several malicious users) query the RF spectrum database at the same time, spectrum server could experience denial of service (DoS) attack. In this paper, we investigate the Aggregated-Query-as-a-Secure-Service (AQaaSS) for querying RF spectrum database by secondary users for opportunistic wireless communications where selected number of secondary users aka grid leaders, query the database on behalf of all other secondary users, aka grid followers and relay the idle channel information to grid followers. Furthermore, the grid leaders are selected based on their both reputation or trust level and location in the network for the integrity of the information that grid followers receive. Grid followers also use the weighted majority voting to filter out comprised information about the idle channels. The performance of the proposed approach is evaluated using numerical results. The proposed approach gives lower latency (or same latency) to the secondary users and lower load (or same load) to the RF spectrum database server when more number of secondary users (or less number of secondary users) query than that of the server capacity.

Cross site scripting (XSS) attack is one of the attacks on the web. It brings session hijack with HTTP cookies, information collection with fake HTML input form and phishing with dummy sites. As a countermeasure of XSS attack, machine learning has attracted a lot of attention. There are existing researches in which SVM, Random Forest and SCW are used for the detection of the attack. However, in the researches, there are problems that the size of data set is too small or unbalanced, and that preprocessing method for vectorization of strings causes misclassification. The highest accuracy of the classification was 98% in existing researches. Therefore, in this paper, we improved the preprocessing method for vectorization by using word2vec to find the frequency of appearance and co-occurrence of the words in XSS attack scripts. Moreover, we also used a large data set to decrease the deviation of the data. Furthermore, we evaluated the classification results with two procedures. One is an inappropriate procedure which some researchers tend to select by mistake. The other is an appropriate procedure which can be applied to an attack detection filter in the real environment.

In this paper, the physical layer (PHY)security performance of a dual-hop cooperative relaying in a cognitive-radio system in the presence of an eavesdropper is investigated. The dual-hop transmission is composed of an asymmetric radio frequency (RF)link and a free space optical (FSO)link. In the considered system, an unlicensed secondary user (SU)uses the spectrum which is shared by a licensed primary user (PU)in a controlled manner to keep the interference at PU receiver, below a predefined value. Furthermore, among M available relays, one relay with the best end-to-end signal-to-noise-ratio (SNR)is selected for transmission. It is assumed that all of the RF links follow Rayleigh fading and all of the FSO links follow Gamma-Gamma distribution. Simulations results for some important security metrics, such as the average secrecy capacity (SC), and secrecy outage probability (SOP)are presented, where some practical issues of FSO links such as atmospheric turbulence, and pointing errors are taken into consideration.

Severe class imbalance between the majority and minority classes in large datasets can prejudice Machine Learning classifiers toward the majority class. Our work uniquely consolidates two case studies, each utilizing three learners implemented within an Apache Spark framework, six sampling methods, and five sampling distribution ratios to analyze the effect of severe class imbalance on big data analytics. We use three performance metrics to evaluate this study: Area Under the Receiver Operating Characteristic Curve, Area Under the Precision-Recall Curve, and Geometric Mean. In the first case study, models were trained on one dataset (POST) and tested on another (SlowlorisBig). In the second case study, the training and testing dataset roles were switched. Our comparison of performance metrics shows that Area Under the Precision-Recall Curve and Geometric Mean are sensitive to changes in the sampling distribution ratio, whereas Area Under the Receiver Operating Characteristic Curve is relatively unaffected. In addition, we demonstrate that when comparing sampling methods, borderline-SMOTE2 outperforms the other methods in the first case study, and Random Undersampling is the top performer in the second case study.

Visible light communication (VLC) is a burgeoning field in wireless communications as it considers illumination and communication simultaneously. The broadcast nature of VLC makes it necessary to consider the security of underlying transmissions. A physical-layer security (PLS) scheme by introducing jamming LEDs is considered in this paper. The secrecy rate of an indoor VLC system with multiple LEDs, one legitimate receiver, and multiple eavesdroppers is investigated. Three distributions of input signal are assumed, i.e., truncated generalized normal distribution (TGN), uniform distribution, and exponential distribution. The results show that jamming can improve the secrecy performance efficiently. This paper also demonstrates that when the numbers of LEDs transmitting information-bearing signal and jamming signal are equal, the average secrecy rate can be maximized.

This article presents a practical approach for secure key exchange exploiting reciprocity in wireless transmission. The method relies on the reciprocal channel phase to mask points of a Phase Shift Keying (PSK) constellation. Masking is achieved by adding (modulo 2π) the measured reciprocal channel phase to the PSK constellation points carrying some of the key bits. As the channel phase is uniformly distributed in [0, 2π], knowing the sum of the two phases does not disclose any information about any of its two components. To enlarge the key size over a static or slow fading channel, the Radio Frequency (RF) propagation path is perturbed to create independent realizations of multi-path fading. Prior techniques have relied on quantizing the reciprocal channel state measured at the two ends and thereby suffer from information leakage in the process of key consolidation (ensuring the two ends have access to the same key). The proposed method does not suffer from such shortcomings as raw key bits can be equipped with Forward Error Correction (FEC) without affecting the masking (zero information leakage) property. To eavesdrop a phase value shared in this manner, the Eavesdropper (Eve) would require to solve a system of linear equations defined over angles, each equation corresponding to a possible measurement by the Eve. Channel perturbation is performed such that each new channel state creates an independent channel realization for the legitimate nodes, as well as for each of Eves antennas. As a result, regardless of the Eves Signal-to-Noise Ratio (SNR) and number of antennas, Eve will always face an under-determined system of equations. On the other hand, trying to solve any such under-determined system of linear equations in terms of an unknown phase will not reveal any useful information about the actual answer, meaning that the distribution of the answer remains uniform in [0, 2π].

Multiple-input multiple-output (MIMO) techniques are currently the de facto approach for increasing the capacity and reliability of communication systems. Spatial modulation (SM) is presently one of the most eminent MIMO techniques. As, it combines the advantages of having higher spectral efficiency than repetition coding (RC) while overcoming the inter-channel interference (ICI) faced by spatial multiplexing (SMP). Moreover, SM reduces system complexity. In this paper, for the first time in literature, the use of MIMO techniques is explored in Internet-of-Things(IoT) deployments by introducing a novel technique called security aware spatial modulation (SA-SM).SA-SM provides a low complexity, secure and spectrally efficient technique that harvests the advantages of SM, while facing the arising security concerns of IoT systems. Using an undemanding modification at the receiver, SA-SM gives an extra degree of technology independent physical layer security. Our results show that SA-SM forces the bit-error-rate (BER) of an eavesdropper to not exceed the range of 10-2, which is below the forward-error-correction (FEC) threshold. Hence, it eradicates the ability of an eavesdropper to properly decode the transmitted signal. Additionally, the efficiency of SA-SM is verified in both the radio and visible light ranges. Furthermore, SA-SM is capable of reducing the peak-to-average-power-ratio (PAPR) by 26.2%.

Extended interaction oscillators (EIOs) are high-frequency vacuum-electronic sources, capable to generate millimeter-wave to terahertz (THz) radiations. They are considered to be potential sources of high-power submillimeter wavelengths. Different slow-wave structures and beam geometries are used for EIOs. This paper presents a quantitative figure of merit, the critical unloaded oscillating frequency (fcr) for any specific geometry of EIO. This figure is calculated and tested for 2π standing-wave modes (a common mode for EIOs) of two different slowwave structures (SWSs), one double-ridge SWS driven by a sheet electron beam and one ring-loaded waveguide driven by a cylindrical beam. The calculated fcrs are compared with particle-in-cell (PIC) results, showing an acceptable agreement. The derived fcr is calculated three to four orders of magnitude faster than the PIC solver. Generality of the method, its clear physical interpretation and computational rapidity, makes it a convenient approach to evaluate the high-frequency behavior of any specified EIO geometry. This allows to investigate the changes in geometry to attain higher frequencies at THz spectrum.

This paper investigates the secrecy outage performance of a dual-hop decode-and-forward (DF) mixed radio-frequency/underwater optical wireless communication (RF/UOWC) system. We consider a one-antenna source node ( S), communicating with one legitimate destination node (D) via a multi-antenna DF relay (R) node. In this context, the relay node receives the incoming signal from S via an RF link, which is subject to Rayleigh fading, then performes selection-combining (SC) followed by decoding and then re-encoding for transmission to the destination over a UOWC link, subject to mixture Exponential-Gamma fading. Under the assumption of eavesdroppers attempting to intercept the S-R (RF side), a closed-form expression for the secrecy outage probability is derived. Our analytical results are corroborated through computer simulations, which verifies their validity.

This paper investigates the secrecy outage performance of an energy harvesting-based dual-hop amplify-and-forward (AF) mixed radio-frequency/underwater optical wireless communication (RF/UOWC) system. A single-antenna source node (S) is considered, communicating with one legitimate destination node (D) with the aid of a multi-antenna AF relay (R) device. In this setup, the relay node receives the incoming signal from S via an RF link, which is subject to Nakagami-m fading, then performs maximal-ratio-combining (MRC) followed by a fixed-gain amplification, before transmitting it to the destination via a UOWC link, subject to mixture Exponential-Gamma fading. Assuming the presence of a malicious eavesdropper attempting to intercept the S- R hop, a tight approximate expression for the secrecy outage probability is retrieved. The derived results provide useful insights into the influence of key system parameters on the secrecy outage performance. Our analytical results are corroborated through computer simulations, which verifies their validity.

Internet of Things (IoT) devices industry is rapidly growing, with an accelerated increase in the list of manufacturers offering a wide range of smart devices selected to enhance end-users' standard of living. Security remains an after-thought in these devices resulting in vulnerabilities. While there exists a cryptographic protocol designed to solve such authentication problem, the computational complexity of cryptographic protocols and scalability problems make almost all cryptography-based authentication protocols impractical for IoT. Wireless RFF (Radio Frequency Fingerprinting) comes as a physical layer-based security authentication method that improves wireless security authentication, which is especially useful for the power and computing limited devices. As a proof-of-concept, this paper proposes a universal SDR (software defined Radio)-based inexpensive implementation intended to sense emitted wireless signals from IoT devices. Our approach is validated by extracting mobile phone signal bursts under different user-dedicated modes. The proposed setup is well adapted to accurately capture signals from different telecommunication standards. To ensure a unique identification of IoT devices, this paper also provides an optimum set of features useful to generate the device identity fingerprint.

Security attacks against Internet of Things (IoT) are on the rise and they lead to drastic consequences. Data confidentiality is typically based on a strong symmetric-key algorithm to guard against confidentiality attacks. However, there is a need to design an efficient lightweight cipher scheme for a number of applications for IoT systems. Recently, a set of lightweight cryptographic algorithms have been presented and they are based on the dynamic key approach, requiring a small number of rounds to minimize the computation and resource overhead, without degrading the security level. This paper follows this logic and provides a new flexible lightweight cipher, with or without chaining operation mode, with a simple round function and a dynamic key for each input message. Consequently, the proposed cipher scheme can be utilized for real-time applications and/or devices with limited resources such as Multimedia Internet of Things (MIoT) systems. The importance of the proposed solution is that it produces dynamic cryptographic primitives and it performs the mixing of selected blocks in a dynamic pseudo-random manner. Accordingly, different plaintext messages are encrypted differently, and the avalanche effect is also preserved. Finally, security and performance analysis are presented to validate the efficiency and robustness of the proposed cipher variants.

Intrusion detection system is described as a data monitoring, network activity study and data on possible vulnerabilities and attacks in advance. One of the main limitations of the present intrusion detection technology is the need to take out fake alarms so that the user can confound with the data. This paper deals with the different types of IDS their behaviour, response time and other important factors. This paper also demonstrates and brings out the advantages and disadvantages of six latest intrusion detection techniques and gives a clear picture of the recent advancements available in the field of IDS based on the factors detection rate, accuracy, average running time and false alarm rate.

We introduce a new defense mechanism for stochastic control systems with control objectives, to enhance their resilience before the detection of any attacks. To this end, we cautiously design the outputs of the sensors that monitor the state of the system since the attackers need the sensor outputs for their malicious objectives in stochastic control scenarios. Different from the defense mechanisms that seek to detect infiltration or to improve detectability of the attacks, the proposed approach seeks to minimize the damage of possible attacks before they actually have even been detected. We, specifically, consider a controlled Gauss-Markov process, where the controller could have been infiltrated into at any time within the system's operation. Within the framework of game-theoretic hierarchical equilibrium, we provide a semi-definite programming based algorithm to compute the optimal linear secure sensor outputs that enhance the resiliency of control systems prior to attack detection.

Behavioral malware detection aims to improve on the performance of static signature-based techniques used by anti-virus systems, which are less effective against modern polymorphic and metamorphic malware. Behavioral malware classification aims to go beyond the detection of malware by also identifying a malware's family according to a naming scheme such as the ones used by anti-virus vendors. Behavioral malware classification techniques use run-time features, such as file system or network activities, to capture the behavioral characteristic of running processes. The increasing volume of malware samples, diversity of malware families, and the variety of naming schemes given to malware samples by anti-virus vendors present challenges to behavioral malware classifiers. We describe a behavioral classifier that uses a Convolutional Recurrent Neural Network and data from Microsoft Windows Prefetch files. We demonstrate the model's improvement on the state-of-the-art using a large dataset of malware families and four major anti-virus vendor naming schemes. The model is effective in classifying malware samples that belong to common and rare malware families and can incrementally accommodate the introduction of new malware samples and families.

The paper dwells on the peculiarities of stylometry technologies usage to determine the style of the author publications. Statistical linguistic analysis of the author's text allows taking advantage of text content monitoring based on Porter stemmer and NLP methods to determine the set of stop words. The latter is used in the methods of stylometry to determine the ownership of the analyzed text to a specific author in percentage points. There is proposed a formal approach to the definition of the author's style of the Ukrainian text in the article. The experimental results of the proposed method for determining the ownership of the analyzed text to a particular author upon the availability of the reference text fragment are obtained. The study was conducted on the basis of the Ukrainian scientific texts of a technical area.

Remote attestation (RA) is a popular means of detecting malware in embedded and IoT devices. RA is usually realized as a protocol via which a trusted verifier measures software integrity of an untrusted remote device called prover. All prior RA techniques require on-demand operation. We identify two drawbacks of this approach in the context of unattended devices: First, it fails to detect mobile malware that enters and leaves the prover between successive RA instances. Second, it requires the prover to engage in a potentially expensive computation, which can negatively impact safety-critical or real-time devices. To this end, we introduce the concept of self-measurement whereby a prover periodically (and securely) measures and records its own software state. A verifier then collects and verifies these measurements. We demonstrate a concrete technique called ERASMUS, justify its features, and evaluate its performance. We show that ERASMUS is well-suited for safety-critical applications. We also define a new metric - Quality of Attestation (QoA).

A new class of multi-band acoustic-wave-Iumped-ele-ment-resonator (AWLR)-based bandstop filters (BSFs) is reported. It is based on\$N\$multi-resonant A WLRs-shaped by\$K\$AWLRs and 2K inverters-that are connected to an all-pass network and result in\$\textbackslashtextbackslashpmbK\textbackslashtextbackslash Nˆth\$order rejection bands. The proposed concept allows the realization of multiple rejection bands with the following characteristics: i) fractional bandwidths (FBWs) larger than the electromechanical coupling coefficient\$\textbackslashtextbackslashpmbk\_tˆ\textbackslashtextbackslash 2\$of its constituent acoustic-wave resonators, ii) continuously variable and inde-pendently-controlled FBWs, iii) intrinsically-switched stopbands, and iv) an all pass state. For proof-of-concept validation purposes a dual-band prototype was designed, built, and tested. It exhibits two stopbands centered at 418 and 433 MHz that can be continu-ously-tuned in FBW (up to 7.7:1 tuning range) and in number.

Conducted emission of motors is a domain of interest for EMC as it may introduce disturbances in the system in which they are integrated. Nevertheless few publications deal with the susceptibility of motors, and especially, servomotors despite this devices are more and more used in automated production lines as well as for robotics. Recent papers have been released devoted to the possibility of compromising such systems by cyber-attacks. One could imagine the use of smart intentional electromagnetic interference to modify their behavior or damage them leading in the modification of the industrial process. This paper aims to identify the disturbances that may affect the behavior of a Commercial Off-The-Shelf servomotor when exposed to an electromagnetic field and the criticality of the effects with regards to its application. Experiments have shown that a train of radio frequency pulses may induce an erroneous reading of the position value of the servomotor and modify in an unpredictable way the movement of the motor's axis.