Deep learning models are known immensely brittle to adversarial image examples, yet their vulnerability in text classification is insufficiently explored. Existing text adversarial attack strategies can be roughly divided into three categories, i.e., character-level attack, word-level attack, and sentence-level attack. Despite the success brought by recent text attack methods, how to induce misclassification with the minimal text modifications while keeping the lexical correctness, syntactic soundness, and semantic consistency simultaneously is still a challenge. To examine the vulnerability of deep models, we devise a Bigram and Unigram based adaptive Semantic Preservation Optimization (BU-SPO) approach which attacks text documents not only at a unigram word level but also at a bigram level to avoid generating meaningless sentences. We also present a hybrid attack strategy that collects substitution words from both synonyms and sememe candidates, to enrich the potential candidate set. Besides, a Semantic Preservation Optimization (SPO) method is devised to determine the word substitution priority and reduce the perturbation cost. Furthermore, we constraint the SPO with a semantic Filter (dubbed SPOF) to improve the semantic similarity between the input text and the adversarial example. To estimate the effectiveness of our proposed methods, BU-SPO and BU-SPOF, we attack four victim deep learning models trained on three real-world text datasets. Experimental results demonstrate that our approaches accomplish the highest semantics consistency and attack success rates by making the minimal word modifications compared with competitive methods.

Machine learning (ML)-based methods are widely explored to predict the quality of transmission (QoT) of a lightpath, which is expected to reduce optical signal to noise ratio (OSNR) margin reserved for the lightpath and therefore improve the spectrum efficiency of an optical network. However, many studies conducting this prediction are often based on synthetic datasets or datasets obtained from laboratory. As such, these datasets may not be amply representative to cover the entire status space of a real optical network, which is often exposed in harsh environment. There are risks of failure when using these ML-based QoT prediction models. It is necessary to develop a mechanism that can guarantee the reliability of a lightpath service even if the prediction models fail. For this, we propose to take advantage of the conventional network protection techniques that are popularly implemented in an optical network and reuse their protection resources to also protect against such a type of failure. Based on the two representative protection techniques, i.e., 1+1 dedicated path protection and shared backup path protection (SBPP), the performance of the proposed protection mechanism is evaluated by reserving different margins for the working and protection lightpaths. For 1+1 path protection, we find that the proposed mechanism can achieve a zero design-margin (D-margin) for a working lightpath thereby significantly improving network spectrum efficiency, while not scarifying the availability of lightpath services. For SBPP, we find that an optimal D-margin should be identified to balance the spectrum efficiency and service availability, and although not significant, the proposed mechanism can save an up to 0.5-dB D-margin for a working lightpath, while guaranteeing the service availability.

Emerging 5G technologies can significantly reduce end-to-end service latency for applications requiring strict quality of service (QoS). With network function virtualization (NFV), to complete a client’s request from those applications, the client’s data can sequentially go through multiple service functions (SFs) for processing/analysis but introduce additional processing delay. To reduce the processing delay from the serially-running SFs, network function parallelism (NFP) that allows multiple SFs to run in parallel is introduced. In this work, we study how to apply NFP into the SF chaining and embedding process such that the latency, including processing and propagation delays, can be jointly minimized. We introduce a novel augmented graph to address the parallel relationship constraint among the required SFs. Considering parallel relationship constraints, we propose a novel problem called parallelism-aware service function chaining and embedding (PSFCE). For this problem, we propose a near-optimal maximum parallel block gain (MPBG) first optimization algorithm when computing resources at each physical node are enough to host the required SFs. When computing resources are limited, we propose a logarithm-approximate algorithm, called parallelism-aware SFs deployment (PSFD), to jointly optimize processing and propagation delays. We conduct extensive simulations on multiple network scenarios to evaluate the performances of our schemes. Accordingly, we find that (i) MPBG is near-optimal, (ii) the optimization of end-to-end service latency largely depends on the processing delay in small networks and is impacted more by the propagation delay in large networks, and (iii) PSFD outperforms the schemes directly extended from existing works regarding end-to-end latency.

This letter presents a gap-waveguide phase shifter based on ridged unit cells with glide-symmetric configuration. The proposed unit cell design provides a higher stable phase shift compared to a conventional ridged  unit cell whose ridge height and waveguide width are tuned to achieve a stable phase shift. Through the insertion of glide-symmetric holes with semi-circle base in the ridged waveguide, a stable phase shift in a wide frequency range is achieved. Depending on the radii of the holes, the operating frequency range of the phase shifter can be tuned. A 90o phase shifter in millimeter-wave range is designed and manufactured to validate the analysis. The experimental results reveal a 85o +- 5o phase shift from 33 GHz to 43 GHz (26.3% bandwidth). In this bandwidth, the reflection coefficient is below -10 dB with a maximum insertion loss of 0.7 dB.

A classical computer works with ones and zeros, whereas a quantum computer uses ones, zeros, and superpositions of ones and zeros, which enables quantum computers to perform a vast number of calculations simultaneously compared to classical computers. In a cloud-supported cyber-physical system environment, running a machine learning application in quantum computers is often difficult, due to the existing limitations of the current quantum devices. However, with the combination of quantum-classical neural networks (NN), complex and high-dimensional features can be extracted by the classical NN to a reduced but more informative feature space to be processed by the existing quantum computers. In this study, we developed a hybrid quantum classical NN to detect an amplitude shift cyber-attack on an in-vehicle controller area network (CAN) dataset. We showed that using the hybrid quantum-classical NN, it is possible to achieve an attack detection accuracy of 94%, which is higher than a Long short-term memory (LSTM) NN (87%) or quantum NN alone (62%)

The optimal design of interior permanent magnet synchronous motors requires a long time because finite element analysis (FEA) is performed repeatedly. To solve this problem, many researchers have used artificial intelligence to construct a prediction model that can replace FEA. However, because the training data are generated by FEA, it takes a very long time to obtain a sufficient amount of data, making it impossible to train a large-scale prediction model. Here, we propose a method for generating a large amount of data from a small number of FEA results using machine learning. An automatic design system with a deep generative model and a convolutional neural network is then constructed. With its sufficient data, the proposed system can handle three topologies and three motor parameters in a wide range of current vector regions. The proposed system was applied to multi-objective optimization design, with the optimization completed in 13-15 seconds.

Deep Neural Networks (DDNs) have shown vulnerability to well-designed adversarial examples. Researchers in industry and academia have proposed many adversarial example defense techniques. However, none can provide complete robustness. The cutting-edge defense techniques offer partial reliability. Thus, complementing them with another layer of protection is a must, especially for mission-critical applications. This paper proposes a novel Online Selection and Relabeling Algorithm (OSRA) that opportunistically utilizes a limited number of crowdsourced workers to maximize the ML system’s robustness. OSRA strives to use crowdsourced workers effectively by selecting the most suspicious inputs and moving them to the crowdsourced workers to be validated and corrected. As a result, the impact of adversarial examples gets reduced, and accordingly, the ML system becomes more robust. We also proposed a heuristic threshold selection method that contributes to enhancing the prediction system’s reliability. We empirically validated our proposed algorithm and found that it can efficiently and optimally utilize the allocated budget for crowdsourcing. It is also effectively integrated with a state-of-the-art black-box defense technique, resulting in a more robust system. Simulation results show that OSRA can outperform a random selection algorithm by 60% and achieve comparable performance to an optimal offline selection benchmark. They also show that OSRA’s performance has a positive correlation with system robustness.

Fine-grained and wide-scale connectivity is a precondition to fully digitalize the manufacturing industry. Driven by this need, new technologies such as time-sensitive networking (TSN), fifth-generation (5G) wireless networks, and industrial Internet-of-things (IIoT) are being applied to industrial communication networks to reach the desired connectivity spectrum. With TSN emerging as a wired networking solution, converging information technology (IT) and operational technology (OT) data streams, 5G is upscaling its access and core networks to function as an independent or a transparent TSN carrier in demanding OT use-cases. In this article, we discuss the drivers for future industrial wireless systems and review the role of 5G and its industrial-centric evolution towards meeting the strict performance standards of factories. We also elaborate on the 5G deployment options, including frequency spectrum allocation and private networks, to help the factory owners discern various dimensions of solution space and concerns to integrate 5G in industrial networks.

Due to their flexibility and low cost deployment, unmanned aerial vehicles (UAV) will most likely act as base stations and backhaul relays in the next generation of wireless communication systems. However, these UAVs—in the untethered mode—can only operate for a finite time due to limited energy they carry in their batteries. In free-space optical communications, one solution is to transport both the data and the energy from the source to the UAV through the laser beam—a concept known as simultaneous lightwave information and power transfer (SLIPT). In this study, we have analyzed the SLIPT scheme for laser-powered decode-and-forward UAV relays in an optical wireless backhaul. The major goal of this study is to optimally allocate the received beam energy between the decoding circuit, the transmitting circuit and the rotor block of the relay in order to maximize a quality-of-service metric such as maximum achievable rate, outage or error probabilities. As expected, we note that the optimal power allocation depends heavily on the source-relay and relay-destination channel conditions. In the final part of this study, we have maximized the operational time of the UAV relay given that the maximum achievable rate stays above a certain threshold in order to meet a minimum quality-of-service requirement.

To support the explosive growth of the Internet of Things (IoT), Uplink (UL) grant-free Non-Orthogonal Multiple Access (NOMA) emerges as a promising technology. It has the potential of offering scalable and low-cost solutions for the resource-constrained Massive Machine Type Communication (mMTC) systems. In principle, the grant-free NOMA enables small signaling overhead and low access latency time by circumventing complicated grant-access based procedures which is commonly found in the legacy wireless networks. In a UL grant-free system, a complete Multi-User Detection (MUD) algorithm not only performs the Active User Detection (AUD) but also the Channel Estimation (CE) and the Data Detection (DD). By exploiting the naturally occurring sparse user activity in the mMTC systems, the MUD problem can be solved using a wide range of Compressive Sensing based algorithms (CS-MUD). However, some alternative routes have been explored in the literature as well. The utility of these algorithms, in general, revolve around some assumptions about the channel or the availability of perfect channel information at the Base Station (BS). How these assumptions are met in a practical circumstance is, however, an important concern. In this work we devise an end-to-end MUD using Deep Neural Network (DNN) where we relax these assumptions. We approximate an ensemble of trained DNN based MUD using Knowledge Distillation (KD) to enable fast AUD at the Base Station (BS). Furthermore, using the inter-resource correlation, we estimate the channels of the active users which is an ill-posed problem otherwise. We carry out elaborate numerical investigation to validate the efficacy of the proposed approach for the UL grant-free NOMA systems.

The ongoing evolution of optical networks towards autonomous systems supporting high-performance services beyond 5G requires advanced functionalities for automated security management. To cope with evolving threat landscape, security diagnostic approaches should be able to detect and identify the nature not only of existing attack techniques, but also those hitherto unknown or insufficiently represented. Machine Learning (ML)-based algorithms perform well when identifying known attack types, but cannot guarantee precise identification of unknown attacks. This makes Root Cause Analysis (RCA) crucial for enabling timely attack response when human intervention is unavoidable. We address these challenges by establishing an ML-based framework for security assessment and analyzing RCA alternatives for physical-layer attacks. We first scrutinize different Network Management System (NMS) architectures and the corresponding security assessment capabilities. We then investigate the applicability of supervised and unsupervised learning (SL and UL) approaches for RCA and propose a novel UL-based RCA algorithm called Distance-Based Root Cause Analysis (DB-RCA). The framework’s applicability and performance for autonomous optical network security management is validated on an experimental physical-layer security dataset, assessing the benefits and drawbacks of the SL- and UL-based RCA. Besides confirming that SL-based approaches can provide precise RCA output for known attack types upon training, we show that the proposed UL-based RCA approach offers meaningful insight into the anomalies caused by novel attack types, thus supporting the human security officers in advancing the physical-layer security diagnostics.

Epilepsy affect almost 1% of the worldwide population. An early diagnosis of seizure types is a crucial patient-dependent step for the treatment selection process. The selection of the proper treatment relies on the correct identification of the seizure type. As such, identifying the seizure type has the biggest immediate influence on therapy than the seizure detection, reducing the neurologist’s efforts when reading and detecting seizures in EEG recordings. Most of the existing seizure detection and classification methods are conceptualized following the patient-dependent schema thus fail to perform well with unknown cases. Our work focuses on patient-independent schema for seizure type classification and pays more attention to the explainability of the underlying attention mechanism of our method. Using a channel-wise attention mechanism, a quantification of the EEG channels contribution is enabled. Therefore, results become more interpretable and a visualization of brain lobes contribution by seizure types is allowed. We evaluate our model for seizure detection and type classification on CHB-MIT and the recently released TUH EEG Seizure, respectively. Our model is able to classify 8 seizure types with an accuracy of 98.41%, directly from raw EEG data without any preprocessing. A case study showed a high correlation between the neurological baselines and the interpretable results of our model.

Recently, it has been observed that $\{0,\pm1\}$-ternary codes which are simply generated from deep features by hard thresholding, tend to outperform $\{-1, 1\}$-binary codes in image retrieval. To obtain better ternary codes, we for the first time propose to jointly learn the features with the codes by appending a smoothed function to the networks. During training, the function could evolve into a non-smoothed ternary function by a continuation method, and then generate ternary codes. The method circumvents the difficulty of directly training discrete functions and reduces the quantization errors of ternary codes. Experiments show that the proposed joint learning indeed could produce better ternary codes.

Event-related potentials (ERPs) are weak activities in electroencephalography (EEG) in response to different stimuli. Detecting these potentials is of great value for many diagnostic and practical purposes. However, this is rather hard because these signals have a low signal-to-noise ratio. In this work, we present a novel approach toward ERP detection based on the fact that this template has a smooth structure compared to background EEG. A generalized likelihood ratio test is adopted to realize our algorithm. The proposed method is investigated for ERP detection from real datasets, and it demonstrates to be very effective compared to conventional state-of-the-art methods.

The effective support of 5G-Internet of Things (IoT) requires cellular service in deep coverage areas while providing long battery life for IoT devices which perform infrequent small data transmission towards the base station. Relaying is a promising solution to extend the coverage while at the same time meeting the battery life requirements of the IoT devices. Considering this, we analyze the suitability of layer-3 relaying over the 3GPP Release 16 NR-PC5 interface to support massive IoT applications. More precisely, we study the unicast connection establishment mechanism over the NR PC5 interface in a partial coverage scenario. Further, a set of optimizations on the Release 16 NR-PC5 procedure to effectively support massive IoT applications are proposed and analyzed. The obtained performance evaluation results which are presented in terms of data success probability, device power consumption, and signaling overhead, quantify how effectively the Release 16 NR-PC5 interface can support the requirement of IoT in the 5G and beyond era. The proposed sidelink small data transmission and frame-level access provides the largest gain overall and can reduce the device power consumption by an average of 68%, and signaling overhead by 15% while maintaining a data success probability of more than 90% in an IMT-2020 defined IoT traffic scenario.

In order to cope with the challenges of improving energy efficiency, increasing the integration of renewable energy, and achieving carbon emission reduction, multi-energy systems have received more and more attention in recent years and have been developing rapidly. Traditionally, different energy infrastructures are usually scheduled and operated independently, which leads to inefficient use of energy and waste of resources. By integrating into a multi-energy system, different energy infrastructures can be coupled and optimized into one unit. In this article, from a low-carbon point of view, the optimal scheduling of a real multi-energy system with hydrogen-based vehicle applications is proposed. The simulation results show that the proposed optimal scheduling can help quantify the daily operation cost and carbon emissions and achieve considerably operation cost saving and carbon reduction by reasonably arranging and utilizing all the devices in the system.

In this paper, we present a mixed-signal integrate and fire neuron designed in a 22-nm FDSOI technology. In this novel design, we deploy the back-gate terminal of FDSOI technology for a tunable design. For the first time, we show analytically and with pre- and post-layout simulations a neuron with tunable spiking frequency using the back-gate voltage of FDSOI technology. The neuron circuit is designed in the sub-threshold region and dissipates an ultra-low energy per spike of the order of Femto Joules per spike. With the layout area of only 30um^2, this is the smallest neuron circuit reported to date.

Successive linear programming (SLP) is a practical approach for solving large-scale nonlinear optimization problems. Alternating current optimal power flow (ACOPF) is no exception, particularly the large size of real-world networks. However, in order to achieve tractability, it is essential to tune the SLP algorithm presented in the literature. This paper presents a modified SLP algorithm to solve the ACOPF problem, specified by the U.S. Department of Energy’s (DOE) Grid Optimization (GO) Competition Challenge 1, within strict time limits. The algorithm first finds a near-optimal solution for the relaxed problem (i.e., Stage 1). Then, it finds a feasible solution in the proximity of the near-optimal solution (i.e., Stage 2 and Stage 3). The numerical experiments on test cases ranging from 500-bus to 30,000-bus systems show that the algorithm is tractable. The results show that our proposed algorithm is tractable and can solve more than 80\% of test cases faster than the well-known Interior Point Method while significantly reduce the number of iterations required to solve ACOPF. The number of iterations is considered an important factor in the examination of tractability which can drastically reduce the computational time required within each iteration.