Hardware obfuscation is a design transformation technique that transforms a design to protect its confidentiality against untrusted parties. In particular, it aims at protecting proprietary hardware intellectual property (IP) blocks against reverse engineering (RE), piracy, and extraction of design secrets. A wide array of existing works on hardware obfuscation have demonstrated its security promises and theoretical robustness against diverse attacks. However, these techniques lack in (1) scalability to large commercial-scale designs, (2) ease of integration with existing electronic design automation (EDA) tool flow, (3) ability to protect against emergent attack modes, such as structural analysis-based attacks, and (4) ability to efficiently trade-off security with design overhead. The latter requires an effective metric to quantify robustness against RE attacks. In this paper, we introduce a practical state space obfuscation algorithm and associated automation tool, ProtectIP, that address the above shortcomings. The algorithmic steps have polynomial complexity and are scalable for large designs. We have developed a complete EDA tool flow that integrates ProtectIP. We show exponential resistance can be achieved against all known RE attacks while incurring modest area overhead (24% on average) with negligible impact (maximum 5% overhead) on critical-path delay. We quantify the level of achieved security using metrics and show that an intelligent attacker with partial knowledge of the obfuscation process has a low probability (maximum 0:33) of succeeding in RE attacks.

In this paper, we investigate the optimal power allocation at the MAC layer in both the asymptotic and the finite blocklength regimes in a two-user NOMA uplink network, under statistical delay constraints captured through the link-layer rate. Using the link-layer effective rate analytic expressions, we provide closed-form expressions of the optimal power coefficients in low and high signal-to-ratio (SNR) regimes. These are validated by an extensive set of simulations, showing that that the proposed power allocation policy optimizes the performance of NOMA compared to OMA at Layer 2.

An online topology estimation algorithm for nonlinear structural equation models (SEM) is proposed in this paper, addressing the nonlinearity and the non-stationarity of real-world systems. The nonlinearity is modeled using kernel formulations, and the curse of dimensionality associated with the kernels is mitigated using random feature approximation. The online learning strategy uses a group-lasso-based optimization framework with a prediction-corrections technique that accounts for the model evolution. The proposed approach has three properties of interest. First, it enjoys node-separable learning, which allows for scalability in large networks. Second, it offers privacy in SEM learning by replacing the actual data with node-specific random features. Third, its performance can be characterized theoretically via a dynamic regret analysis, showing that it is possible to obtain a linear dynamic regret bound under mild assumptions. Numerical results with synthetic and real data corroborate our findings and show competitive performance w.r.t. state-of-the-art alternatives.

The paper studies the method for online teaching using Multi-Track Modular Teaching (MT2). MT2 is an advanced teaching-learning method for better training of students with deep skills.  Covid19 pandemic and the associated lockdown periods, coupled with the requirements of social distancing, have played a major disruption in the academic processes across all the world’s universities. Due to the inability to conduct physical classes and lectures, most universities have opted for online classes using video conference tools. One way to make the computer networking classes, during lock-down periods, more effective is to use the Multi-Track Modular Teaching (MT2) technique which organizes a course into multiple tracks and modules. This paper presents the mod?ifications carried out to MT2-based computer networking course and experiences from teaching it online during the lock-down period. While the online classes offered an alter?native to the completely disrupted teaching and learning process, observations on their effectiveness reveal certain important considerations in order to make them more useful. Results from a feedback survey of the students who under?took the course is also provided in this paper, giving insights to certain difficulties and issues in the conduct of online classes in developing countries.

The increasing interest in the residential renewable energy sources leveraged the development of the Battery Energy Storage Systems (BESS) ecosystems that allow the losses mitigation caused by the lack of intelligent storage of unused energy. BESS’s have reduced user’s costs by storing the overproduced energy by the renewable sources and using it when the demand was bigger than the production. Despite that, cost minimizing can yet be improved by applying a Energy Management System (EMS) which takes in consideration several factors and draws a policy to reduce costs. Model-based approaches have been thought as a solution but the increasing complexity of the models turned the researchers into a model-free direction using Reinforcement Learning. This paper creates an innovative agent, trained by Deep Reinforcement Learning algorithms, capable of reducing the costs in a building. A literature review of other approaches, a solution formulation to solve the problem, and evaluation using synthetic data and data supplied by an energy company are contained. Furthermore, evaluation scenarios and techniques that can lead to agent’ instability are mentioned. Results indicate that the agent has better performance in terms of revenue when compared with the same BESS operation without the agent.

This study presents a Distribution System State Estimation (DSSE) algorithm for unbalanced smart grids and evaluates its performance with special focus on the effect of the unavailability of measurements, which is a common problem on smart grids. The proposed DSSE algorithm is based on the Weighted Least Squares (WLS) method and the unbalanced power flow analysis and addresses several challenges such as unbalanced consumption, various configurations and types of loads, line couplings and lack of measurements. To validate the proposed algorithm’s performance, the IEEE 4-node test feeder and the IEEE European Low Voltage test feeder, comprising 907 nodes, are used as benchmarks. A detailed analysis of the DSSE state results, their deviations from the IEEE test feeders, and their sensitivity regarding measurements’ availability, demonstrates the accuracy and robustness of the proposed solution.

The log-sum regularization has been always drawing widespread attention in the field of sparse problem. However, it brings about a non-convex, non-smooth, and non-Lipschitz optimization problem that is difficult to tackle. To overcome the problem, an iterative threshold algorithm of log-sum regularization is proposed in this paper. Firstly, by deducing the derivative mathematical expression of log-sum function, a property theorem about solution for log-sum regularization is established. Secondly, based on the above theorem, the optimal setting rules of the compromising parameters are elaborated, and the iterative log-sum threshold algorithm is proposed. Thirdly, under the situation that the compromising parameters of log-sum regularization are relatively small, it can be proven that the proposed algorithm converges to a local minimizer of log-sum regularization. Finally, a series of simulations are implemented to examine performance of the algorithm, and the results exhibit that the proposed algorithm outperforms the state-of-the-art algorithms in terms of iterations and precision.

The centrality and diversity of the labeled data are very influential to the performance of semi-supervised learning (SSL). Most existing SSL models select the labeled data randomly and equally allocate the labeling quota among the classes, leading to considerable unstableness and degeneration of performance. Active learning has been proposed to address the problem of instance selection in learning with a few labels, but its iterative and progressive procedure causes heavy computing cost. Optimal leading forest (OLF) has the advantage of revealing the difference evolution along a path within a subtree. This study unsupervisedly constructs a leading forest that forms another metric space, based on which it is convenient to select most central and divergent samples with one shot. The labeling quota can be allocated flexibly according to the data distribution. A discrete optimization problem is formulated based on the new metric space to select the samples to label. With the small number of selected instances, the kernelized large margin projection can be efficiently learned to classify the remaining unlabeled samples. The multi-modal issue in SSL is effectively addressed by the multi-granular structure of leading forest that readily facilitates multiple local metrics learning. Extensive experimental results demonstrate that the proposed method achieved competitive efficiency and encouraging accuracy when compared with the state-of-the-art methods. The code is available at https://github.com/alanxuji/DeLaLA

This research paper proposes an advanced method for Power Flow Analysis (PFA) in unbalanced distribution networks considering high Renewable Energy Sources (RES) penetration with special attention to technical and environmental Key Performance Indicators (KPIs). The innovation of the proposed unbalanced power flow algorithm lies on the tree-like approach of the unbalanced network with the use of Backward/Forward Sweep (BFS) methods, which renders the algorithm significantly faster than the usual methodologies. The PFA is applied on the unbalanced IEEE-13 node test feeder including two-axis tracking Photovoltaics (PV) for three different load profiles, i.e., commercial, residential and mixed, for one representative day of the year. The aforementioned cases are compared to each other in terms of voltage deviation, distribution losses, self-consumption, emissions and reverse power flow. Finally, sensitivity analysis is carried out in order to highlight the effect of RES production on the unbalanced distribution network.

Using a laparoscope-holding robot in minimally invasive surgery has been shown to improve surgeons’ efficiency and operation safety. However, controlling the robot can increase surgeons’ cognitive load. Therefore, a suitable interface is critical to ease the control task and reduce surgeons’ workload. A foot interface for commanding the robot is common, with two control strategies possible: Decoupled control only allows a single Cartesian axis actuation, known as decoupled commands. Hybrid control enables both a single axis and multiple axes actuation, known as coupled commands. This paper aims to find the most suitable control strategy for foot interfaces by validating two common assumptions in the literature for a laparoscope manipulation task: 1. Decoupled control is believed to have better final laparoscopic view orientation predictability, and 2. Hybrid control is assumed to be more efficient in manipulating the laparoscope. Results from a user study with 11 surgeons and surgeon trainees show that the decoupled control indeed has better predictability than the hybrid control. However, the result does not support assumption 2 with both approaches equally efficient. In addition, we analyze the interface usage patterns and safety when choosing between the coupled and decoupled commands using two surgery-like tasks in a simulator. The results show that decoupled commands are preferred for small adjustments, and coupled commands are preferred for larger motions. No obvious correlation between control safety and command selection is found. Based on the assumptions validation and usage patterns study, a guideline for foot interface control strategy selection is provided. The experiment protocol can be modified for other human-robot interaction studies in surgical and industry applications.

Machine Learning Operations (MLOps) streamline the lifecycle of machine-learning models in production. In recent years, the topic has picked the interest of practitioners, and consequently, a considerable number of tools and gray literature on architecting MLOps environments has emerged. However, this has created a new problem for organizations: selecting the most appropriate tools and design options for implementing their MLOps environments. To alleviate this problem, this paper proposes a reference architecture and requirements for MLOps by systematically reviewing 58 industrial gray literature articles. Such reference architecture drawn from the state of practice shall aid organizations in making better design and technology choices when embarking on their MLOps journey while providing a technology-independent baseline for further MLOps research.

Tools based on deep learning models have been created in recent years to aid radiologists in the diagnosis of breast cancer from mammograms. However, the datasets used to train these models may suffer from class imbalance, i.e., there are often fewer malignant samples than benign or healthy, which can bias the model towards the healthy class. In this study, we systematically evaluate several popular techniques to deal with this class imbalance, namely, class weighting, over-sampling, and under-sampling, as well as a synthetic lesion generation approach to increase the number of malignant samples. These techniques are applied when training on three diverse Full-Field Digital Mammography datasets, and tested on in-distribution and out-of-distribution samples. The experiments show that a greater imbalance is associated with a greater bias towards the majority class, which can be counteracted by any of the standard class imbalance techniques. On the other hand, these methods provide no benefit to model performance with respect to Area Under the Curve of the Recall Operating Characteristic (AUC-ROC), and indeed under-sampling leads to a reduction of 0.066 in AUC in the case of a 19:1 benign to malignant imbalance. Our synthetic lesion methodology leads to better performance in most cases, with increases of up to 0.07 in AUC on out-of-distribution test sets over the next best experiment.

We previously introduced a novel mixed reality (MR) teleguidance system, human teleoperation [1,2], in which a human (expert) leader and a human (novice) follower are tightly coupled through MR and haptics for applications such as tele-ultrasound. In this paper, a communication system suitable for human teleoperation is presented and characterized in various network conditions, over Ethernet, Wi-Fi, 4G LTE, and 5G. To study all types of latency in the system, the human response time is additionally characterized through step response tests with 11 volunteers. The step responses were obtained by tracking the position and force of the human hand in response to a change in the MR target. The round-trip communication latency is 40+/-10 ms over 5G, and down to 1+/-0.6 ms over Ethernet for typical throughputs. The human response time to a step change in position depends on the step magnitude, but is 485-535 ms, while the reaction time for forces is 150-200 ms. Both lags are decreased when tracking smooth motions. Thus, we demonstrate that the system is network agnostic and can achieve good teleoperation performance and secure, fast communication in appropriate network conditions. The presented tools and concepts are applicable to any high-performance teleoperation system, for example for remote surgery.

Due to the rapid development of satellite imaging sensors, high-resolution images are being generated for use. Various image processing algorithms, such as deep learning models, require images of reduced sizes given the computational constraints. Hence, preprocessing the images to reduce their size is crucial for any deep learning model. This paper proposes a novel approach to compress satellite images using quantum computing. A comparative study on different standard data embedding techniques used in quantum computing is undertaken. We propose four quantum compression techniques (𝑄𝐶𝑇s) by extending the unitary operations of amplitude encoding for compressing satellite images. The proposed methods provide exponential scaling as amplitude encoding is used, where 2𝑛 classical data values are encoded into 𝑛 qubits. Compression performance, visual evaluation, and quality metric evaluation were carried out to assess the proposed compression techniques. Our experimental results showed that the crucial patterns in images are retained in the compressed images without quality loss even after 75% compression. The compressed images can be used for post-processing tasks such as classification using classical or quantum computing algorithms.

This paper presents a time-domain formulation of a meshless method based on the generalized sheet transition conditions (GSTCs). This formulation is applicable for modeling a kind of dispersive metasurface and a time-varying metasurface. This is the first time a meshless method is developed for analysis of the metasurfaces using GSTCs. The significance of time-domain meshless methods originates from their capabilities for multi-scale and conformal modeling in addition to analysis of problems in a broad frequency-range. Here, for simplicity, the meshless method formulation is presented for analysis of 1-D problems consisting of metasurfaces which are monoisotropic. The extension of this method to 2-D and 3-D problems and bianisotropic metasurfaces is straightforward. The efficiency of the proposed formulations is verified by numerical examples.

While research in passive flexible circuits for Wireless Power Transfer (WPT) such as coils and resonators continues to advance, limitations in their power handling and low efficiency have hindered the realization of efficient all-printed high-power wearable WPT receivers. Here, we propose a screen-printed textile-based 6.78 MHz resonant inductive WPT system using planar inductors with concealed metal-insulator-metal (MIM) tuning capacitors. A printed voltage doubler rectifier based on Silicon Carbide (SiC) diodes is designed and integrated with the coils, showing a power conversion efficiency of 80-90% for inputs between 2 and 40 W. Compared to prior wearable WPT receivers, it offers an order of magnitude improvement in power handling along with higher efficiency (approaching 60%), while using all-printed passives and a compact rectifier. The coils exhibit a simulated Specific Absorption Rate (SAR) under 0.4 W/kg for 25 W received power, and under 21◦C increase in the coils’ temperature for a 15 W DC output. We finally demonstrate a wirelessly-powered textilebased carbon-silver Joule heater, capable of reaching up to 60◦C at 2 cm separation from the transmitter, as a potential wearable application which can only be wireless-powered using the proposed system.

The industrial sector has been making use of machines and sensors having the capability to communicate with each other. This forms a network of Industrial devices which is called Industrial Internet of Things (IIoT). IIoT is an emerging trend that can generate a huge amount of data that is vulnerable to cyber-attacks. In IIoT network, data and control instructions flow through Supervisory Control and Data Acquisition (SCADA) system. A Network Intrusion Detection System (NIDS) which can monitor realtime network traffic could be deployed between SCADA and the IIoT devices to detect cyber-attacks. NIDS with Deep Learning (DL) algorithms require large dataset for which CSE-CIC-IDS2018 and UNSW-NB15 dataset is used. The paper compares a Multi-Layer Perceptron(MLP), a Fully Connected Deep Neural Network (FCNN), and Convolutional Neural Network (CNN) on CSE-CIC-IDS2018 and UNSW-NB15 with XGBoost for feature selection.

Long-term wearable ECG monitoring has increased the influence of ambiguous factors on QRS-complex detection. Reliable ECG information must be extracted from abundant noises and confusing artifacts. To address this issue, we suggest a positive-negative QRS-complex detector (PN-QRS), a QRS-complex detection model that accounts for uncertainty. Specifically, we relate non-ECG screening and artifact recognition to epistemic and aleatoric uncertainty, respectively. To disentangle the two uncertainties, we introduce the class-biased transformation based on the uniqueness of this problem. The performance of artifact and non-ECG detection and multi-lead QRS-complex localization is then validated. The results indicate that PN-QRS reduces ECG error detection rate by 22.24% and improves non-ECG screening accuracy by 5.9% compared to conventional signal quality assessment methods. And for identification of ambiguous beats, PN-QRS achieves a F1 score of 82.41% in a manually annotated ambiguous ECG dataset, while maintaining high precision for QRS-complex detection (99.38% F1 score). Above all, for multi-lead QRS-complex location, PN-QRS is approaching the performance upper bound through integrating the frame-level results on the lead with the minimum uncertainty. Our work suggests that the proposed PN-QRS has the potential to be employed as a QRS-complex detector in wearable ECG monitoring, with the capacity to remove invalid non-ECG episodes and identify QRS-like artifacts simultaneously.