Deep convolutional neural networks (Deep CNN) have achieved hopeful performancefor single image super-resolution. In particular, the Deep CNN skip Connection andNetwork in Network (DCSCN) architecture has been successfully applied to naturalimages super-resolution. In this work we propose an approach called SDT-DCSCN thatjointly performs super-resolution and deblurring of low-resolution blurry text imagesbased on DCSCN. Our approach uses subsampled blurry images in the input and origi-nal sharp images as ground truth. The used architecture is consists of a higher numberof filters in the input CNN layer to a better analysis of the text details. The quantitativeand qualitative evaluation on different datasets prove the high performance of our modelto reconstruct high-resolution and sharp text images. In addition, in terms of computa-tional time, our proposed method gives competitive performance compared to state ofthe art methods.

This paper reviews, discusses, and highlights some of the standard cloud computing vulnerabilities primarily related to virtualization platforms and their implementations while outsourcing services and resources to different end-users and business enterprises. Furthermore, we provide Blockchain-enabled solutions for virtualized cloud computing platform involving both the end-users as well as cloud service providers (CSP) to address and solve different security and privacy-related vulnerabilities using blockchain-enabled solutions to improve their cloud services and resource provisioning facilities.

Smile recognition plays a vital role in human-human and human-computer interactions. This paper demonstrates a system to recognize the genuine and posed smiles by sensing observers’ galvanic skin response (GSR), while watching sets of images and videos. The smiles were shown either in ‘paired’ or in ‘single’ forms. Here, ‘paired’ means that the same smiler was seen in both genuine and posed smile forms, otherwise the condition is referred to as ‘single’. The GSR signals were recorded and processed, and several time-domain and frequency-domain features were extracted from the processed GSR signals. Classification accuracies were found to be as high as 93.6% and 91.4% from paired and single conditions respectively. In comparison, observers were verbally 59.8% and 56.2% correct. Our results demonstrate that human subconscious responses (i.e. GSR signals) is better than their own verbal response, where the paired condition is slightly better than the single condition.

This piece proposes a novel mechanism for peer-to-peer electricity trading whereby energy tokens can only be redeemed in the same part of the day as when they were generated. The aim of this regulatory mechanism is to reduce token hoarding by consumers to better align the physical production and consumption of electricity, which in turn could decrease electrical system losses and minimise the chance of grid imbalances. To establish the effectiveness of this dayparting mechanism a market simulation is performed. This simulation is made up of 24 consumers' and five producers' profiles over a seven-day week. An optimisation is performed to most effectively allocate energy tokens from producers to consumers, aiming to minimise the total energy imported from the larger grid i.e. to make most effective use of local generation. Consumers are permitted to perform a measure of demand response by modulating their demand at certain points while keeping their total energy consumption constant. Allocated energy tokens can be consumed immediately, or during any subsequent daypart to the same type. A series of power flow analyses are performed using the market simulation out-turns to establish the electrical system effects. Consumers are found to move some demand to weekend days when demand is lower but generation is equally abundant. Electrical results reveal a decrease in system losses, as well as less fluctuation from the larger grid supply.

Convolutional neural networks (CNNs), inspired by biological visual cortex systems, are a powerful category of artificial neural networks that can extract the hierarchical features of raw data to greatly reduce the network parametric complexity and enhance the predicting accuracy. They are of significant interest for machine learning tasks such as computer vision, speech recognition, playing board games and medical diagnosis [1-7]. Optical neural networks offer the promise of dramatically accelerating computing speed to overcome the inherent bandwidth bottleneck of electronics. Here, we demonstrate a universal optical vector convolutional accelerator operating beyond 10 Tera-FLOPS (floating point operations per second), generating convolutions of images of 250,000 pixels with 8-bit resolution for 10 kernels simultaneously — enough for facial image recognition. We then use the same hardware to sequentially form a deep optical CNN with ten output neurons, achieving successful recognition of full 10 digits with 900 pixel handwritten digit images with 88% accuracy. Our results are based on simultaneously interleaving temporal, wavelength and spatial dimensions enabled by an integrated microcomb source. This approach is scalable and trainable to much more complex networks for demanding applications such as unmanned vehicle and real-time video recognition.

High cross-polarization isolation (CPI) is crucial to the accurate polarization measurement using polarimetric phased array radar (PPAR). In this paper, we propose an adaptive direction-dependent polarization state configuration (AD2PSC) method to improve the polarization isolation. Compared with conventional fixed polarization state of radiated wave whether it is linear, circular or elliptical polarization state, our AD2PSC approach configures the polarization state on basis of beam steering. To achieve the adaptive configuration of magnitude and phase of the dual-polarization antenna, an improved steepest descent algorithm is put forward. To facilitate the uniform representation for the polarization measurement application of PPAR, the universal expressions of intrinsic and measured backscatter matrices are derived for arbitrary polarization state. The dualpolarization dipole array is used to assess the priority of our proposed method. Compared with the conventional approaches, our approach could obtain higher CPI while being available for a larger scanning range. The configured CPI meets the specific polarization requirement for PPAR.

This is a point of view of Devops in embedded domain

This piece serves as an exploratory study into the combination of blockchain-aware smart meters and radical economic paradigms. Load shedding is examined through this novel lens. Load shedding is an established means of preventing system collapse under emergency conditions. An auctioned load shedding hierarchy method is proposed whereby participating parties bid on tokens (SALtoks) guaranteeing access to supply during load shedding. This method is based on the recently developed Self Assessed Licenses Sold via Auction paradigm developed. Under this paradigm, SALtoks are subject to taxes based on valuations, while also being subject to continuous auction. A case study is performed on a hypothetical island power system. Higher ranking SALtoks are held by larger consumers. The method is found to mitigate economic impacts by allowing industry to maintain supply during load shedding. Smaller consumers, while subject to more frequent interruptions, gain the public benefit of increased income from taxation.

This article represents a new method for Data Retention failure (DR) in NAND Flash memories by analyzing the charge leakage phenomena. Retention failure accrue when stored data changes its level. This usually happens at and because of a high temperature environment that accelerate leakage. The amount of electron charge in the cell is rather hard to calculate. However, we know that a verification process done at the end of the programming with respect to the amount of charge, is apparently enough to indicate the proper logical level. Moreover, when the cell is exposed to high temperature the charge leaks out but the manufacturer‘s guarantee that the memory storage will withstand the JEDS218 standard, this mean that this amount of charge should be enough for 10 years at temperature ambient range of 25÷55°C. Hence, after 10 years, in the worst-case scenario, the amount of charge might be so low that its results can adversely affect its logical level. In our previous work, we found that the programed cell array affected by temperature as shown in figure 1, begin with the average voltage is equal to Vt1 and later, when the leakage occurs, the average voltage decreases while the variation increases.

The COVID-19 pandemic has severely impacted many industries, in particular the healthcare sector exposing systemic vulnerabilities in emergency preparedness, risk mitigation, and supply chain management. A major challenge during the pandemic was related to the increased demand of Personal Protective Equipment (PPE) resulting in critical shortages for healthcare and frontline workers. The lack of information visibility combined with the inability to precisely track product movement within the supply chain requires an robust traceability solution. Blockchain technology is a distributed ledger that ensures a transparent, safe, and secure exchange of data among supply chain stakeholders. The advantages of adopting blockchain technology to manage and track PPE products in the supply chain include decentralized control, security, traceability, and auditable time-stamped transactions. In this paper, we present a blockchain-based approach using smart contracts to transform PPE supply chain operations. We propose a generic framework using Ethereum smart contracts and decentralized storage systems to automate the processes and information exchange and present detailed algorithms that capture the interactions among supply chain stakeholders. The smart contract code was developed and tested in Remix environment, and the code is made publicly available on Github. We present detailed cost and security analysis incurred by the stakeholders in the supply chain. Adopting a blockchain-based solution for PPE supply chains is economically viable and provides a streamlined, secure, trusted, and transparent mode of communication among various stakeholders.

The Kerr nonlinear optical performance of silicon nanowire waveguides integrated with 2D layered graphene oxide (GO) films is theoretically studied and optimized based on experimentally measured linear and nonlinear optical parameters of the GO films. The strong mode overlap between the silicon nanowires and highly nonlinear GO films yields a significantly enhanced Kerr nonlinearity for the hybrid waveguides. A detailed analysis for the influence of waveguide geometry and GO film thickness on the propagation loss, nonlinear parameter, and nonlinear figure of merit (FOM) is performed. The results show that the effective nonlinear parameter and nonlinear FOM can be increased by up to ≈52 and ≈79 times relative to bare silicon nanowires, respectively. Self-phase modulation (SPM)-induced spectral broadening of optical pulses is used as a benchmark to evaluate the nonlinear performance, examining the trade-off between enhancing Kerr nonlinearity and minimizing loss. By optimizing the device parameters to balance this, a high spectral broadening factor of 27.6 can be achieved ‒ more than 6 times that achieved in previous experiments. Finally, the influence of pulse chirp, material anisotropy, and the interplay between saturable absorption and SPM is also discussed. These results provide useful guidance for optimizing the Kerr nonlinear optical performance of silicon waveguides integrated with 2D layered GO films.

The problem of selecting the modulation and coding scheme (MCS) that maximizes the system throughput, known as link adaptation, has been investigated extensively, especially for IEEE 802.11 (WiFi) standards. Recently, deep learning has widely been adopted as an efficient solution to this problem. However, in failure cases, predicting a higher-rate MCS can result in a failed transmission. In this case, a retransmission is required, which largely degrades the system throughput. To address this issue, we model the adaptive modulation and coding (AMC) problem as a multi-label multi-class classification problem. The proposed modeling allows more control over what the model predicts in failure cases. We also design a simple, yet powerful, loss function to reduce the number of retransmissions due to higher-rate MCS classification errors. Since wireless channels change significantly due to the surrounding environment, a huge dataset has been generated to cover all possible propagation conditions. However, to reduce training complexity, we train the CNN model using part of the dataset. The effect of different subdataset selection criteria on the classification accuracy is studied. The proposed model adapts the IEEE 802.11ax communications standard in outdoor scenarios. The simulation results show the proposed loss function reduces up to 50% of retransmissions compared to traditional loss functions.

Optical artificial neural networks (ONNs) have significant potential for ultra-high computing speed and energy efficiency. We report a new approach to ONNs based on integrated Kerr micro-combs that is programmable, highly scalable and capable of reaching ultra-high speeds, demonstrating the building block of the ONN — a single neuron perceptron — by mapping synapses onto 49 wavelengths to achieve a single-unit throughput of 11.9 Giga-OPS at 8 bits per OP, or 95.2 Gbps. We test the perceptron on handwritten-digit recognition and cancer-cell detection — achieving over 90% and 85% accuracy, respectively. By scaling the perceptron to a deep learning network using off-the-shelf telecom technology we can achieve high throughput operation for matrix multiplication for real-time massive data processing.

In a frequency division duplexing multiple-input multiple-output (FDD-MIMO) system, the user equipment (UE) send the downlink channel state information (CSI) to the base station for performance improvement. However, with the growing complexity of MIMO systems, this feedback becomes expensive and has a negative impact on the bandwidth. Although this problem has been largely studied in the literature, the noisy nature of the feedback channel is less considered. In this paper, we introduce PRVNet, a neural architecture based on variational autoencoders (VAE). VAE gained large attention in many fields (e.g., image processing, language models, or recommendation system). However, it received less attention in the communication domain generally and in CSI feedback problem specifically. We also introduce a different regularization parameter for the learning objective, which proved to be crucial for achieving competitive performance. In addition, we provide an efficient way to tune this parameter using KL-annealing. Empirically, we show that the proposed model significantly outperforms state-of-the-art, including two neural network approaches. The proposed model is also proved to be more robust against different levels of noise.

Triaxial HTS AC power cables, that incorporates three HTS phases wound around a single core within a single cable is considering the optimal solution for low and medium voltages. Such a design permits to save expensive HTS conductor and to increase the power density transmitted. It is difficult to increase the transmission capacity of a triaxial cable by increasing its operating voltage and therefore, insulation thickness between phases. The other way is to increase the operation current by using a multilayer structure in each phase of the triaxial cable. This task demands a complicated optimization analysis and a precise manufacturing technology. Following our previous studies, we developed and tested the triaxial cable prototype with two layers per phase made of ReBCO wire. The optimal parameters of the cable (twist pitch, diameters of phases, etc.) have been determined using the numerical simulation methods developed. DC and AC tests of the cable has been performed. In the result, the feasibility to provide a uniform current distribution in a triaxial cable, each phase of which consists of two layers, has been demonstrated. With an outer diameter of ~24 mm only, this is the most compact triaxial HTS power cable manufactured and tested up to date with currents up to 4 kA per phase. The anticipated power transfer with three phases of this cable can be as large as ~34 MW.

DevOps is the combination of cultural mindset, practices, and tools that increases a team’s ability to release applications and services at high velocity. Transitioning to DevOps requires a change in culture and mindset first. DevOps is about removing the barriers between two traditionally delimited teams, development and operations. With DevOps, these two teams work together to optimize both the productivity of developers and the reliability of operations. They strive to communicate frequently, increase efficiencies, and improve the quality of services they provide. They take full ownership for their services, often beyond where their stated roles or titles have traditionally been scoped. This paper aims to bring different phases of software development lifecycle into DevOps implementation strategy.This paper highlights on DevOps practices which span across different phases of the Software Development Lifecycle (SDLC).

Keypoints of objects reflect their concise abstractions, while the corresponding connection links (CL) build the skeleton by detecting the intrinsic relations between keypoints. Existing approaches are typically computationally-intensive, inapplicable for instances belonging to multiple classes, and/or infeasible to simultaneously encode connection information. To address the aforementioned issues, we propose an end-to-end category-implicit Keypoint and Link Prediction Network (KLPNet), which is the first approach for simultaneous semantic keypoint detection (for multi-class instances) and CL rejuvenation. In our KLPNet, a novel Conditional Link Prediction Graph is proposed for link prediction among keypoints that are contingent on a predefined category. Furthermore, a Cross-stage Keypoint Localization Module (CKLM) is introduced to explore feature aggregation for coarse-to-fine keypoint localization. Comprehensive experiments conducted on three publicly available benchmarks demonstrate that our KLPNet consistently outperforms all other state-of-the-art approaches. Furthermore, the experimental results of CL prediction also show the effectiveness of our KLPNet with respect to occlusion problems.

Powered prostheses can enable individuals with above-knee amputations to ascend stairs step-over-step. To accomplish this task, available stair ascent controllers impose a pre-defined joint impedance behavior or follow a pre-programmed position trajectory. These control approaches have proved successful in the laboratory. However, they are not robust to changes in stair height or cadence, which is essential for real-world ambulation. Here we present an adaptive stair ascent controller that enables individuals with above-knee amputations to climb stairs of varying stair heights at their preferred cadence and with their preferred gait pattern. We found that modulating the prosthesis knee and ankle position as a function of the user's thigh in swing provides toe clearance for varying stair heights. In stance, modulating the torque-angle relationship as a function of the prosthesis knee position at foot contact provides sufficient torque assistance for climbing stairs of different heights. Furthermore, the proposed controller enables individuals to climb stairs at their preferred cadence and gait pattern, such as step-by-step, step-over-step, and two-steps. The proposed adaptive stair controller may improve the robustness of powered prostheses to environmental and human variance, enabling powered prostheses to more easily move from the lab to the real-world.