In this work, we present a new Nyquist pulse that shows high tolerance to the timing-jitter, which is one of the key factors that ceil the end-to-end bit error rate performance of terahertz communications systems. As a proof-of-concept, an experiment is conducted using a 300 GHz photonics-based terahertz communications link in order to demonstrate the performance of the proposed waveform.

Recent years have witnessed the rapid deployment of smart homes; most of them are controlled by remote servers in the cloud. Such designs raise security and privacy concerns for end users. In this paper, we describe the design of Sovereign, a home IoT system framework that provides end users complete control of their home IoT systems. Sovereign lets home IoT devices and applications communicate via application-named data and secures data directly. This enables direct, secure, one-to-one and one-to-many device-to-device communication over wireless broadcast media. Sovereign utilizes semantic names to construct usable security solutions. We implement Sovereign as a publish-subscribe-based development platform together with a prototype home IoT controller. Our preliminary evaluation shows that Sovereign provides a systematic, easy-to-use solution to user-controlled, self-contained smart homes running on existing IoT hardware without imposing noticeable overhead.

During COVID-19 and other pandemics, endotracheal intubation is an effective and common method to save patients as the virus causes lung fibrosis and thus patients are unable to breathe spontaneously. Medical staff need to insert a tube close to the patient’s mouth, thereby leading to a high risk of cross-infection. To protect medical staff, we propose an autonomous intubation robot system (AIRS). With the developed visual servoing and hybrid control method, the entire system can simulate doctors for satisfying repeatability and safety of intubation operations. This system includes a self-driving/teleoperation platform, two co-robot arms, a new multi-functional laryngoscope, force sensors, and several cameras. In the visual servoing part, we realize recognition and location of the patient’s face, medical devices, and main physiological structures to provide real-time navigation. In the hybrid control part, we establish an oral model, propose an offline planning method and PID controllers by combining force, vision, and motion, and apply Virtual Fixture to insert safely. AIRS’s validation is with a phantom model under a 2-min operation. Our proposed robot is original and promising in the area of emergent medical robots. We will further validate AIRS in clinical applications and extend the developed techniques in other general treatments.

In this paper, we propose a 3D fully automatic joint erosion algorithms for hand HR-pQCT, and its performance can reach consensus with human specialist.

With the increasing development of smart grid, multiparty cooperative computation between several entities has become a typical characteristic of modern energy systems. Traditionally, data exchange among parties is inevitable, rendering how to complete multiparty collaborative optimization without exposing any private information a critical issue. This paper proposes a fully privacy-preserving distributed optimization framework based on secure multiparty computation (SMPC) secret sharing protocols. The framework decomposes the collaborative optimization problem into a master problem and several subproblems. The process of solving the master problem is executed in the SMPC framework via the secret sharing protocols among agents. The relationships of agents are completely equal, and there is no privileged agent or any third party. The process of solving subproblems is conducted by agents individually. Compared to the traditional distributed optimization framework, the proposed SMPC-based framework can fully preserve individual private information. Exchanged data among agents are encrypted and no private information disclosure is assured. Furthermore, the framework maintains a limited and acceptable increase in computational costs while guaranteeing optimality. Case studies are conducted to demonstrate the principle of secret sharing and verify the feasibility and scalability of the proposed methodology.

Influenza, an acute viral respiratory disease that is currently causing severe financial and resource strains worldwide. With the recent COVID-19 pandemic exceeding 153 million cases worldwide, there is a need for a low-cost and contactless surveillance system to detect symptomatic individuals, more so in counties with limited healthcare resources. As with many diseases, there are bio-clinical signals relating to the physical symptoms. The main objective of this study was to develop FluNet, a novel, proof-of-concept, low-cost and contactless device for the detection of high-risk individuals. The system passively conducts face detection in the longwave infrared domain with a precision rating of 0.9798 and mean intersection over union of 0.7386 while sequentially taking the temperature trend of faces with a thermal accuracy of ± 1 K. While in parallel determining if someone in audible proximity is coughing by using a custom deep convolutional neural network with a precision rating of 0.9519. In addition to presenting FluNet, two datasets have been constructed, one for face detection in the longwave infrared domain consisting of 250 images of 20 participants’ faces at various rotations and coverings, including face masks. The other for the real-time detection of cough patterns comprised of a sizeable dataset of 40,482 cough / not cough sounds, coupled with a new lightweight artificial neural network architecture for the classification of cough spectrograms. These findings could be helpful for future low-cost edge computing applications for influenza-like monitoring.

Artificial intelligence has smoothly penetrated several economic activities, especially monitoring and control applications, including the agriculture sector. However, research efforts toward low-power sensing devices with fully functional machine learning (ML) on-board are still fragmented and limited in smart farming. Biotic stress is one of the primary causes of crop yield reduction. With the development of deep learning in computer vision technology, autonomous detection of pest infestation through images has become an important research direction for timely crop disease diagnosis. This paper presents an embedded system enhanced with ML functionalities, ensuring continuous detection of pest infestation inside fruit orchards. The embedded solution is based on a low-power embedded sensing system along with a Neural Accelerator able to capture and process images inside common pheromone-based traps. Three different ML algorithms have been trained and deployed, highlighting the capabilities of the platform. Moreover, the proposed approach guarantees an extended battery life thanks to the integration of energy harvesting functionalities. Results show how it is possible to automate the task of pest infestation for unlimited time without the farmer’s intervention.

Prediction of future movement of stock prices has been a subject matter of many research work. On one hand, we have proponents of the Efficient Market Hypothesis who claim that stock prices cannot be predicted, on the other hand, there are propositions illustrating that, if appropriately modelled, stock prices can be predicted with a high level of accuracy. There is also a gamut of literature on technical analysis of stock prices where the objective is to identify patterns in stock price movements and profit from it. In this work, we propose a hybrid approach for stock price prediction using machine learning and deep learning-based methods. We select the NIFTY 50 index values of the National Stock Exchange (NSE) of India, over a period of four years: 2015 – 2018. Based on the NIFTY data during 2015 – 2018, we build various predictive models using machine learning approaches, and then use those models to predict the “Close” value of NIFTY 50 for the year 2019, with a forecast horizon of one week, i.e., five days. For predicting the NIFTY index movement patterns, we use a number of classification methods, while for forecasting the actual “Close” values of NIFTY index, various regression models are built. We, then, augment our predictive power of the models by building a deep learning-based regression model using Convolutional Neural Network (CNN) with a walk-forward validation. The CNN model is fine-tuned for its parameters so that the validation loss stabilizes with increasing number of iterations, and the training and validation accuracies converge. We exploit the power of CNN in forecasting the future NIFTY index values using three approaches which differ in number of variables used in forecasting, number of sub-models used in the overall models and, size of the input data for training the models. Extensive results are presented on various metrics for all classification and regression models. The results clearly indicate that CNN-based multivariate forecasting model is the most effective and accurate in predicting the movement of NIFTY index values with a weekly forecast horizon.

An algorithm is proposed that extracts 3D shape from shading information in a digital image. The algorithm assumes that there is only a single source of light producing the image, that the surface of the shape giving rise to the image is Lambertian (matte) and that its shape can be locally approximated by a quadratic function. Previous work shows that under these assumptions, robust shape from shading is possible, though slow for large images because a non-linear optimization method is applied in order to estimate local quadratic surface patches from image intensities. The work presented here shows that local quadratic surface patch estimates can be computed, without prior knowledge of the light source direction, via a linear least squares optimization, thus greatly improving the algebraic complexity and run-time of the existing algorithms.

If in future, the highly intelligent machines control the world, then what would be its advantages and disadvantages? Will, those artificial intelligence powered superintelligent machines become an anathema for humanity or will they ease out the human works by guiding humans in complicated tasks, thereby extending a helping hand to the human works making them comfortable. Recent studies in theoretical computer science especially artificial intelligence predicted something called ‘technological singularity’ or the ‘intelligent explosion’ and if this happens then there can be a further stage as transfused machinery intelligence and actual intelligence where the machines being immensely powerful with a cognitive capacity more than that of humans for solving ‘immensely complicated tasks’ can takeover the humans and even the machines by more intelligent machines of superhuman intelligence. Therefore, it is troublesome and worry-full to think that ‘if in case the machines turned out against humans for their optimal domination in this planet’. Can humans have any chances to avoid them by bypassing the inevitable ‘hard singularity’ through a set of ‘soft singularity’. This paper discusses all the facts in details along with significant calculations showing humanity, how to avoid the hard singularity when the progress of intelligence is inevitable.

In this paper, we report record DC and RF performance in β-Ga2O3 High Electron Mobility Transistor (HEMT) with field-plate T-gate using 2-D simulations. The T gate with head-length LHL of 180 nm and foot-length LFL of 120 nm is used in the highly scaled device with an aspect ratio (LG/tbarrier) of ~ 5. The proposed device takes advantage of a highly polarized Aluminum Nitride (AlN) barrier layer to achieve high Two-Dimensional Electron Gas (2DEG) density in the order of 2.3 × 1013 cm-2, due to spontaneous as well as piezoelectric polarization components. In the depletion mode operation, maximum drain current ID,MAX of 1.32 A/mm, and relatively flat transconductance characteristics with a maximum value of 0.32 S/mm are measured. The device with source-drain distance LSD of 1.9 µm exhibits record low specific-on resistance RON,sp of 0.136 mΩ-cm–2, and off-state breakdown voltage of 403 V, which correspond to the record power figure-of-merit (PFoM) of ~ 1194 MW/cm2. Additionally, current gain cut-off frequency fT and maximum oscillation frequency fMAX of 48 and 142 GHz are estimated. The obtained results show the potential of Ga2O3 HEMT for futuristic power devices.

In this paper, a joint spectrum allocation and device association problem is investigated for a federated learning aided hierarchical Industrial Internet of Things (IIoT) system for smart factory. To achieve the optimization jointly, we design a weighted learning utility maximization problem, which is a 0-1 integer linear programming problem. To solve this problem, we convert it into a weighted 3D hypergraph model by capturing the 3D mapping relation for IIoT device, subchannel, and edge server. A local search algorithm is then presented to find a 3D hypergraph matching with maximum total weights as the suboptimal solution. Simulation results demonstrate the superior performance of the proposed algorithm compared with the greedy algorithm in the system learning utility.

Microwave microfluidic sensors are typically designed with a channel in vicinity of a resonator’s fringing electric (E)-fields, to characterize the material properties of a single fluid. This paper leverages hybrid 3D and dispenser printing to realize a scalable microfluidic sensor utilizing the parallel-plate capacitance of an open-ended microstrip stub, enabling, for the first time, a tunable sensitivity. A stub-loaded microstrip line is then proposed for characterizing multiple microfluidic samples simultaneously using a simple two-port multi-band resonator. The physical constrains which limit the scalability of the proposed sensors have been analyzed analytically and numerically, prior to implementing a three-channel triple-band sensor. The microfluidic channels have been fabricated using stereolithography 3D printing with the microstrip line directly dispenser printed on a conformable polyimide substrate. To accommodate varying channel thicknesses, a tapered microstrip line is proposed to maintain the impedance matching. The fabricated sensor is characterized using binary water-IPA mixtures to evaluate its sensitivity, comparing favorably with reported 3D-printed sensors. The proposed sensor achieves over 90% accuracy in determining the real permittivity following a simple water-based calibration across the different channels, for samples with 16 oC temperature sensitivity across all channels.

Alternating current optimal power flow (ACOPF) problem is a non-convex and a nonlinear optimization problem. Similar to most nonlinear optimization problems, ACOPF is an NP-hard problem. On the other hand, Utilities and independent service operators (ISO) require the problem to be solved in almost real-time. The real-world networks are often large in size and developing an efficient and tractable algorithm is critical to many decision-making processes in electricity markets. Interior-point methods (IPMs) for nonlinear programming are considered one of the most powerful algorithms for solving large-scale nonlinear optimization problems. However, the performance of these algorithms is significantly impacted by the optimization structure of the problem. Thus, the choice of the formulation is as important as choosing the algorithm for solving an ACOPF problem. Different ACOPF formulations are evaluated in this paper for computational viability and best performance using the interior-point line search (IPLS) algorithm. Different optimization structures are used in these formulations to model the ACOPF problem representing a range of varying sparsity. The numerical experiments suggest that the least sparse ACOPF formulation with polar voltages yields the best computational results. A wide range of test cases, ranging from 500-bus systems to 9591-bus systems, are used to verify the test results.

This paper describes the design and the realization of a low-frequency ac magnetic-field-based indoor positioning system (PS). The system operation is based on the principle of inductive coupling between wire loop antennas. Specifically, due to the characteristics of the ac artificially generated magnetic fields, the relation between the induced voltage and the distance is modeled with a linear behavior in a bilogarithmic scale when a configuration with coplanar, thus equally oriented, antennas is used. In this case, the distance between a transmitting antenna and a receiving one is estimated using measurements of the induced voltage in the latter. For a high operational range, the system makes use of resonant antennas tuned at the same nominal resonant frequency. The quality factors act as antenna gain increasing the amplitude of the induced voltage. The low-operating frequency is the key factor for improving robustness against nonline-of-sight (NLOS) conditions and environment influences with respect to other existing solutions. The realized prototype, which is implemented using off-the-shelf components, exhibits an average and maximum positioning error, respectively, lower than 0.3 and 0.9 m in an indoor environment over a large area of 15 m × 12 m in NLOS conditions. Similar performance is obtained in an outdoor environment over an area of 30 m × 14 m. Furthermore, the system does not require any type of synchronization between the nodes and can accommodate an arbitrary number of users without additional infrastructure.

From the most known Gaussian mixture to the cutting-edge multi-Bernoulli mixture of various forms, mixture offers a fundamental means to deal with uncertainties, which has led to a variety of appealing applications in the state estimation realm based on a single sensor or a sensor network. Like noise is often used to model unknown system input, one may use various hypotheses to deal with the uncertain state space model or data association. Meanwhile, consensus may be sought over the cross-correlated sensors. These all drive a need for representing the probability distribution by a mixture of properly weighted component distributions, which fuse the information gained from different models/hypotheses or from different sensors. This technical note presents information-theoretical results which answer how the averaging/mixture approach makes sense and how the fusing weights should be designed.

Navigation in an unknown environment without any preexisting positioning infrastructure has always been hard for mobile robots. This paper presents a self-deployable ultra wideband UWB infrastructure by mobile agents, that permits a dynamic placement and runtime extension of UWB anchors infrastructure while the robot explores the new environment. We provide a detailed analysis of the uncertainty of the positioning system while the UWB infrastructure grows. Moreover, we developed a genetic algorithm that minimizes the deployment of new anchors, saving energy and resources on the mobile robot and maximizing the time of the mission. Although the presented approach is general for any class of mobile system, we run simulations and experiments with indoor drones. Results demonstrate that maximum positioning uncertainty is always controlled under the user’s threshold, using the Geometric Dilution of Precision (GDoP).

The popularity of the Android Operating System in the smartphone market has given rise to lots of Android malware. To accurately detect these malware, many of the existing works use machine learning and deep learning-based methods, in which feature extraction methods were used to extract fixed-size feature vectors using the files present inside the Android Application Package (APK). Recently, Graph Convolutional Network (GCN) based methods applied on the Function Call Graph (FCG) extracted from the APK are gaining momentum in Android malware detection, as GCNs are effective at learning tasks on variable-sized graphs such as FCG, and FCG sufficiently captures the structure and behaviour of an APK. However, the FCG lacks information about callback methods as the Android Application Programming Interface (API) is event-driven. This paper proposes enhancing the FCG to eFCG (enhanced-FCG) using the callback information extracted using Android Framework Space Analysis to overcome this limitation. Further, we add permission - API method relationships to the eFCG. The eFCG is reduced using node contraction based on the classes to get R-eFCG (Reduced eFCG) to improve the generalisation ability of the Android malware detection model. The eFCG and R-eFCG are then given as the inputs to the Heterogeneous GCN models to determine whether the APK file from which they are extracted is malicious or not. To test the effectiveness of eFCG and R-eFCG, we conducted an ablation study by removing their various components. To determine the optimal neighbourhood size for GCN, we experimented with a varying number of GCN layers and found that the Android malware detection model using R-eFCG with all its components with four convolution layers achieved maximum accuracy of 96.28%.