Physical layer security (PLS) can be adopted for efficient key sharing in secured wireless systems. The random nature of the wireless channel and channel reciprocity (CR) are the main pillars for realizing PLS techniques. However, for applications that involve air-to-air (A2A) transmission, such as unmanned aerial vehicle (UAV) applications, the channel does not generally have sufficient randomness to enable reliable key generation. Therefore, this work proposes a novel system design to mitigate the channel randomness constraint and enable high-rate secret key generation (SKG) process. The proposed system integrates physically unclonable functions (PUFs) and CR principle to securely exchange secret keys between two nodes. Moreover, an adaptive and controllable artificial fading (AF) level with interleaving is used to mitigate the impact of low randomness variations in the wireless channel. The proposed system can operate efficiently even when the channel is nearly flat or time invariant. Consequently, the time required for generating and sharing a key is significantly shorter than conventional techniques. We also propose a novel bit extraction scheme that reduces the number of overhead bits required to share the intermediate keys. The obtained Monte Carlo simulation results show that a key agreement can be reached at the first trial for moderate and high signal-to-noise ratios (SNRs), which is substantially faster than other PLS techniques. Moreover, the results show that inducing AF into static channels reduces the mismatch ratio between the generated secret sequences and degrades the eavesdropper’s capability to predict the secret keys.

The data related to the classification of micro-Doppler result of human's and held object's motion

In order to satisfy the demand for designing efficient neural networks with an appropriate tradeoff between model performance (e.g., classifification accuracy) and computational complexity, the differentiable neural architecture distillation  (DNAD) algorithm is developed based on two cores, namely, search by deleting and search by imitating. On the one hand, to derive neural architectures in a search space where cells of the same type no longer share the same topology, the super-network progressive shrinking (SNPS) algorithm is developed based on the framework of differentiable architecture search (DARTS), i.e., search by deleting. Unlike conventional DARTS-based approaches which produce neural architectures with simple structures and derive only one architecture during the search procedure, the SNPS algorithm is able to derive a Pareto-optimal set of architectures with more complex structures (more powerful representation ability) by forcing the dynamic super-network shrink from a dense structure to a sparse one progressively. On the other hand, since knowledge distillation (KD) has shown great effectiveness to train a compact student network with the help of a powerful over-parameterized teacher network, we combine the KD with SNPS to derive the DNAD algorithm, i.e., search by imitating. By minimizing the behavioral difference between the dynamic super-network and the teacher network, the over-fifitting of one-level DARTS is avoided and well-performed neural architectures are derived. Experiments on CIFAR10 and ImageNet demonstrate that both SNPS and DNAD are able to derive neural architectures in one search procedure with competitive tradeoffs between the model performance and the computational complexity.

Impairment in persons with multiple sclerosis (PwMS) can often be attributed to symptoms of motor instability and fatigue. Symptom monitoring and queued interventions often target these symptoms. Clinical metrics are currently limited to objective physician assessments or subjective patient reported measures. Recent research has turned to wearables for improving the objectivity and temporal resolution of assessment. Our group has previously observed wearable assessment of supervised and unsupervised standing transitions to be predictive of fall-risk in PwMS. Here we extend the application of standing transition quantification to longitudinal home monitoring of symptoms. Subjects (N=23) with varying degrees of MS impairment were recruited and monitored with accelerometry for a total of ~6 weeks each. These data were processed using a deep learning activity classifier to isolate periods of standing transition from which descriptive features were extracted for analysis. Participants completed daily and biweekly assessments describing their symptoms. From these data, Canonical Correlation Analysis was used to derive digital phenotypes of MS instability and fatigue. We find these phenotypes capable of distinguishing fallers from non-fallers, and further that they demonstrate a capacity to characterize symptoms at both daily and sub-daily resolutions. These results represent promising support for future applications of wearables, which may soon augment or replace current metrics in longitudinal monitoring of PwMS.

The electricity demand is increasing day by day and that is exhausting limited conventional energy resources rapidly. Thus, renewable energy is becoming very attractive source of electricity to meet the load and prevent the global warming threat. The challenges in integration of renewable energy sources are its intermittency and old existing power system infrastructure. Renewable energy sources collocated with energy storage facilities can mitigate the effect of renewable energy uncertainty. However, high penetration of green energy is only possible by upgrading the existing power system network. Most of the time transmission lines are under loaded from its actual power handling capacity, as worst possible power handling capacity of line is considered while scheduling power to generators. Actual power capacity of line depends on many environmental factors and changes accordingly with the weather conditions. This article elucidates the impact of weather variation on economic operation of a practical Indian 75 bus power system network placed at Uttar Pradesh, India. The simulation is carried out by accounting the proposed solar power plants in that region and the effect of dynamic line rating on renewable curtailment, thermal power fuel cost and line congestion has been studied.

This article introduces a method of evaluating subsamples until any prescribed level of classification accuracy is attained, thus obtaining arbitrary accuracy. A logarithmic reduction in error rate is obtained with a linear increase in sample count. The technique is applied to specific emitter identification on a published dataset of physically recorded over-the-air signals from 16 ostensibly identical high-performance radios. The technique uses a multi-channel deep learning convolutional neural network acting on the bispectra of I/Q signal subsamples each consisting of 56 parts per million (ppm) of the original signal duration. High levels of accuracy are obtained with minimal computation time: in this application, each addition of eight samples decreases error by one order of magnitude.

A document by Amjad khan . Click on the document to view its contents.

This paper is about one of the functions of the data link layer, which controls access to shared media. The algorithm presented must be able to operate accurately and efficiently. In the approach focused on in this article, stations must specify the time of using services for access to the server. Then they will enter a system that arrives at the exit without any collisions; it will use the server based on the timing of the stations themselves. Many methods have been published on this, but in this paper .We tried a novel approach with its special features where the performance of this method is shown compared to other methods.

Multivariate time series (MTS) are captured in a great variety of real-world applications. However, analysing and modelling the data for classification and forecasting purposes can become very challenging if values are missing in the data set. The need for imputation methods, to fill the gaps in MTS, is well known. Thus, a great variaty of algorithms for solving this task has been proposed in the literature. However, research community is constantly working on the development of advanced algorithms, that fulfill the special requirements of multidimensional temporal data, since most of the existing imputation methods treat MTS as ordinary structured data and fail to model the temporal relationships within and between sequences of observations. The main emphasis of MTS imputation research is currently put on deep learning (DL) models, especially models making use of generative adversarial networks (GANs). In our survey, we present a general categorization of imputation algorithms and introduce groups of hybrid GAN-models used for the MTS imputation task, which we investigate and discuss in detail. A quantitative comparison of the hybrid GANs’ performance regarding MTS imputation is presented based on our findings in the literature.

Electroencephalography (EEG) enables online monitoring brain activity, which can be used for neurofeedback. One of the growing applications of EEG neurofeedback is to facilitate meditation practice. Specifically, EEG neurofeedback can be used to alert participants whenever they get distracted during meditation practice based on changes in their brain activity. In this study, we develop machine learning models to detect moments of distraction (due to mind wandering or drowsiness) during meditation practice using EEG signals. We use EEG data of 24 participants while performing a breath focus meditation with experience sampling and extract twelve linear and non-linear EEG features. Features are fed to ten supervised machine learning models to classify (i) Breath Focus Awake (BFA) vs Breath Focus Sleepy (BFS), and (ii) BFA vs Mind Wandering (MW). We observe that the linear features achieve a maximum accuracy of 86% for classifying awake (BFA) and sleepy (BFS), whereas non-linear features have more predictive ability for classifying between BFA and MW with a maximum accuracy of nearly 78% . In addition, visualization of unsupervised t-SNE lower embeddings supports the evidence of distinct clusters for each condition. Overall our results show that machine learning algorithms can successfully identify periods of distraction during meditation practice in novice meditators based on linear and non-linear features of the EEG signal. Consequently, our results have important implications for the development of mobile EEG neurofeedback protocols aimed at facilitating meditation practice.

Deep convolutional neural networks (CNNs) generate intensive inter-layer data during inference, which results in substantial on-chip memory size and off-chip bandwidth. To solve the memory constraint, this paper proposes an accelerator adopted with a compression technique that can reduce the inter-layer data by removing both intra- and inter-channel redundant information. Principal component analysis (PCA) is utilized in the compression process to concentrate inter-channel information. The spatial differences, truncation, and reconfigurable bit-width coding are implemented inside every feature map to eliminate the intra-channel data redundancy. Moreover, a particular data arrangement is introduced to enhance data continuity to optimize PCA analysis and improve compression performance. A CNN accelerator with the proposed compression technique is designed to support the on-the-fly compression process by pipelining the reconstruction, CNN computation, and compression operation. The prototype accelerator is implemented using 28-nm CMOS technology. It achieves 819.2GOPS peak throughput and 3.75TOPS/W energy efficiency with 218.5mW. Experiments show that the proposed compression technique achieves a compression ratio of 21.5%~43.0% (8-bit mode) and 9.8%~19.3% (16-bit mode) on state-of-the-art CNNs with a negligible accuracy loss.

Mammography is the most widely used breast screening method for early detection and routine monitoring of breast cancer. Most mammogram classification or segmentation techniques are limited to simple binary classifications of the abnormalities in mammograms or classifying specific cropped regions. However, these methods encode merely the texture features of the X-ray-filmed lesions, despite the relationships and patterns. Thus, most of these models have inefficient learning capacities, hindered by supervised learning processes that require extensive annotated data, which is not feasible in the medical domain. To overcome these challenges, we model the deep texture and spatial features of every small segment of the mammogram in a complex multi-graph network. Moreover, we proposed a semi-supervised mammogram multi-graph attention network (MMGAT) to classify each element into all probabilistic categories. The attention and the dense relationships in the multi-graph network empower the proposed model with significant learning capacities, resulting in a competitive performance outperforming the recent state-of-the-art mammogram classification models. The proposed model can localise every abnormal segment of the mammogram and classify each as a malignant or benign region of a mass or calcification cancerous. The proposed MMGAT requires only 40\% of the data for training to achieve the highest AUC of 0.97 and an F1 score of 0.95 on the publically available DDSM.

5G and beyond networks are expected to support a wide range of services, with highly diverse requirements. Yet, the traditional “one-size-fits-all” network architecture lacks the flexibility to accommodate these services. In this respect, network slicing has been introduced as a promising paradigm for 5G and beyond networks, supporting not only traditional mobile services, but also vertical industries services, with very heterogeneous requirements. Along with its benefits, the practical implementation of network slicing brings a lot of challenges. Thanks to the recent advances on Machine Learning (ML), some of these challenges have been addressed. In particular, the application of ML approaches is enabling the autonomous management of resources, in the network slicing paradigm. Accordingly, this paper presents a comprehensive survey on contributions on ML in network slicing, identifying major categories and sub-categories in the literature. Key takeaways are also presented and open research challenges are discussed, together with potential solutions.

The South African rail industry is seeking a critical measure to monitor and save energy usage in the freight rail sector. The rail industry is experiencing increasing operational costs and high energy consuming driving records  along various coal lines. Eco-driving is a modern and efficient way of driving that emphasizes fuel efficiency, speed, and safety. This study provides an algorithm to find the optimal trajectory for a freight train hauling a load over a specific distance. Optimised speed profile is composed of optimal acceleration, coasting, and deceleration. The Freight Eco-Driving Energy Optimiser (FEDEO) solution is not yet applied to freight trains globally, especially in Africa. In this study, the eco-driving strategy of a freight train is formulated as an optimisation problem, whose objective function is the energy cost. The decision variables are the tractive and braking effort notches, and the speed, acceleration and distance limits are formulated as constraints. The formulated eco-driving problem is solved by Mixed-Integer Non-Linear Programming (MINLP) from the Opti-Solver toolbox. The FEDEO algorithm is applied to a train consisting of eight 19E locomotives with two-hundred CCR-9 wagons, over a distance of 90.64 km. The results show up to 34.76% reduction in energy costs.

The exponential growth of cloud computing within the last decade has led to a severe consideration about the power requirement of the cloud data center. However, variations in the computing resource demands and fixed size of virtual machines (VMs) lead to a poor resource utilization, load imbalance, and excessive energy consumption within the data center. Dynamic VM consolidation effectively handles these problems, where VMs are placed on as few physical machines (PMs) as possible. Concurrently, VM allocation must be done strategically by considering various factors of the computing resources for minimal exploitation of them. The high frequency of VM consolidation may lead to the degradation of the system’s reliability by assigning VMs to unreliable PMs. A novel Quantum Machine learning driven Reliable Resource Allocation (QM-RRA) model is proposed to address these problems and enhance the data center’s reliability. We handle the problem by applying an Evolutionary Quantum C-Not Neural Network (QCNN) unit to forecast future resource usage at the cloud data center. In addition, Quantum Evolutionary Adaptive Algorithm (QEAA) is utilized to optimize qubit network weights. It offers energy saving by minimizing the number of active PMs, maximizing resource utilization, and enhancing the reliability of the overall data center. The QM-RRA model is evaluated by performing experiments on the Google Cluster Data set (GCD). The observed results are compared with the state-of-art methods and shows its supremacy in terms of different performance parameters like resource utilization, power consumption, and the number of active PMs. There is a significant reduction in power consumption and the number of active PMs up to 31.38%  and 44.93%, respectively. The enhancement in resource utilization is up to 25.61%, and overall system reliability is 63.35%.

Multiple longest common subsequence (MLCS) mining (a classical NP-hard problem) is an important task in many fields. Numerous applications in these fields can generate very long sequences (i.e., the length of the sequences ≥ 10^4), called big sequences. Such big sequences present a serious challenge to existing MLCS algorithms. Although significant efforts have been made to tackle the challenge, both existing exact and approximate MLCS algorithms fail to deal with big sequences as their problem-solving model MLCS-DAG (Directed Acyclic Graph) is too large to be calculated due to the memory explosion. To bridge the gap, this paper first proposes a new identification and deletion strategy of different classes of non-critical points, which are the points that do not contribute to the solution on the MLCS-DAG. It then proposes a new MLCS problem-solving graph model, called KP-MLCS-DAG (Key Point based MLCS-DAG). A novel parallel MLCS algorithm, called KP-MLCS (Key Point based MLCS), is also presented, which can mine and compress all MLCSs of big sequences effectively and efficiently. Extensive experiments on both synthetic and real-world biological sequences show that the proposed algorithm KP-MLCS drastically outperforms the existing state-of-the-art algorithms in terms of efficiency and effectiveness. The source code of KP-MLCS and related test datasets, etc., can be found online: https://github.com/kp-mlcs/KP-MLCS.

Most of the existing solutions for communication of connected vehicles (CVs) are focused on the optimization of resource allocation, however, not taking the driving routes of the CVs into account. In addition, no existing work considers the offloading of computing tasks from CVs to edge computing servers. In this paper, we introduce joint vehicular route selection and radio and computing resource allocation for CVs. The proposed algorithm minimizes the ratio of failed tasks along the entire vehicular route considering the availability of both radio and computing resources. To manage the allocation of resources among multiple CVs for each vehicular route, we develop a blockchain-based framework allowing resource reservation by means of nonfungible tokens (NFTs). Each NFT represents an exclusive right to the required amount of radio and computing resources for the given road segment and defined time interval. Simulation results show that the proposed vehicular route selection algorithm reduces the ratio of tasks not completed before deadline by up to 69% compared to the existing state-of-the-art algorithms.

Linear fractals associated with weights are investigated. Such fractals are important from a conservation law perspective that is relevant in a variety of physical systems such as materials science, sand dune fractals, barred galaxies, as well as in temporal processes like the EEG.  Weighted fractals lead to distributions that are consistent with the ubiquitous power law and the first digit phenomenon. Two linear fractal algorithms that are near optimal in the information theoretic sense are described. It is proposed that one mechanism for the emergence of these fractals is the indistinguishability amongst the particles in the evolution and transformation of physical systems.