(SoE) School of Engineering

<p>Shiv Nadar University is a comprehensive, multidisciplinary, research-focused, and student-centric University offering a full range of academic programs at the undergraduate, postgraduate and doctoral level. With state-of-the-art infrastructure, the University comprises of academic wings, sophisticated labs, international standard sports facilities, amphitheaters, auditorium, conference rooms and smart classrooms. The University&rsquo;s goal is to become internationally recognized for the quality of its research and creative endeavors and their applicability to improving quality of life, generating new insights and expanding the boundaries of human knowledge creativity. Committed to excellence in teaching, research and service, the University aims to serve the higher education needs of India and the world beyond.</p>

Shiv Nadar University

Cyberbullying is one of the leading causes of mental health issues in the younger population, often leading to depression, stress, and suicidal tendencies. Often it is observed that timely intervention can bring awareness towards unintentional cyberbullying. This paper presents an approach to the prevention of cyberbullying via social networking. The objective of this work is real-time detection of the degree of offensiveness and alerting the user typing the message. We also propose corrective measures, by displaying an alternative word for any offensive word that is typed in the suggestion box in real-time. This enables the user to prevent unintentional cyberbullying. The proposed solution is integrated into an app, that displays the relevant statistics as well as visualization of the user typing pattern. The model to detect the offensiveness percentage can achieve 97.77\% accuracy and is the backbone of the entire approach. © 2021, Springer Nature Switzerland AG.

Computer Science and Engineering

Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)

A Real Time Analysis of Offensive Texts to Prevent Cyberbullying

Graph learning from the brain signals deals with capturing the changes in functional relationship between the brain regions during mental active and relaxed states. This paper investigates different graph learning techniques, namely geometry, signal similarity, and Graphical LASSO based methods for the classification of mental task from electroencephalogram (EEG) signals. Graph spectral energy based metric using Graph Signal Processing (GSP) technique is presented to classify mental active state from relaxed state. A binary KNN classifier is used to analyse each graph learning technique on publicly available Keirn and Aunon mental task EEG database. Performance of different graphs is then analysed and compared using classification Accuracy and F-Score. © 2021 IEEE.

World University root node

Electrical Engineering

Proceedings of the 2021 IEEE 18th India Council International Conference, INDICON 2021

Evaluating Different Graph Learning Techniques for Mental Task EEG Signal Classification

The enhanced capacity-Achieving non-orthogonal multiple access (NOMA) systems suffer from detection errors due to the interference of various users in the uplink (UL) scenario. Successive interference cancellation (SIC) errors propagation and reception of signals from multiple antennas result in inter-user interference (IUI) and inter-Antenna interference (IAI), respectively. Dirty paper coding (DPC) mitigates the IAI by diagonalizing the channel matrix, whereas SIC reduces the IUI by using the back-substitution method. This paper analyses the UL NOMA-MIMO system's performance by proposing a joint DPC-OSIC detector with known channel state information (CSI). QR-based OSIC reduces the IUI while applying all the constellations to the lowest channel gain user. The proposed joint DPC-OSIC is analyzed over a joint DPC-ZF and DPC-ML performance. Simulation results validate that the proposed detector achieves near maximum likelihood (ML) performance with relatively low complexity. © 2021 IEEE.

IEEE Region 10 Annual International Conference, Proceedings/TENCON

Performance of Uplink NOMA-MIMO System with Joint DPC-OSIC Detector

Bias Temperature Instability (BTI) is one of the most crucial reliability issues in modern CMOS technology. It leads to shift in device parameters, which eventually affect circuit performance. SRAM is a widely used circuit which occupies a considerable area in microprocessors. Hence it is important to understand the impact of BTI on performance/stability of SRAM. As device degradation due to BTI depends on gate activity, SRAM performance strongly depends on the data stored in the cell. In this paper, the gate activity is incorporated by activity factor ‘α’ which takes into account the various data patterns stored in the cell. Although many studies have reported the impact of BTI on SRAM performance, none focused on the worst degradation scenario. Our analysis with varying activity factor ‘α’ provides an opportunity to identify the data pattern for worst case degradation. Shift in threshold voltage due to BTI is modelled according to continuous and non-continuous applied gate bias, using physics-based compact model. Process variability is incorporated using Monte Carlo (MC) simulations and worst-case degradation at distribution tail is identified. In this paper, we consolidate various SRAM performance metrics from literature over the last three decades and demonstrate the impact of BTI and process variability with activity factor ‘α’ on these metrics. © 2020 Elsevier B.V.

Integration

Analysis of SRAM metrics for data dependent BTI degradation and process variability

This research deals with the design and fabrication of novel designs as a means to harness wind energy using ‘Savonius' Turbine. It is generally employed to harvest the low to extremely low wind speed potentials. They can be particularly of high use in the regions where the wind speeds are not high enough for installing conventional horizontal axis wind turbines. The paper introduces a novel concept of testing a Savonius wind turbine with dimple structures on its blades and investigating its effect on the turbine's efficiency. The study also deals with the experimental validation of the effect of using a converging ducted structure with a single stage and double stage configurations of Savonius wind turbine, apart from comparing the turbine's performance with and without endplates. The results of the study prove that power coefficient increases with the addition of dimple structures on the blades and it can be further augmented by using a ducted structure with the Savonius turbine. © 2020. All rights reserved.

Mechanical Engineering

International Journal of Renewable Energy Research

Numerical and Experimental Investigations on a Dimpled Savonius Vertical Axis Wind Turbine

Diabetic Macular Edema (DME) is an advanced stage of Diabetic Retinopathy (DR) and can lead to permanent vision loss. Currently, it affects 26.7 million people globally and on account of such a huge number of DME cases and the limited number of ophthalmologists, it is desirable to automate the diagnosis process. Computer-assisted, deep learning based diagnosis could help in early detection, following which precision medication can help to mitigate the vision loss. Method: In order to automate the screening of DME, we propose a novel DMENet Algorithm which is built on the pillars of Convolutional Neural Networks (CNNs). DMENet analyses the preprocessed color fundus images and passes it through a two-stage pipeline. The first stage detects the presence or absence of DME whereas the second stage takes only the positive cases and grades the images based on severity. In both the stages, we use a novel Hierarchical Ensemble of CNNs (HE-CNN). This paper uses two of the popular publicly available datasets IDRiD and MESSIDOR for classification. Preprocessing on the images is performed using morphological opening and gaussian kernel. The dataset is augmented to solve the class imbalance problem for better performance of the proposed model. Results: The proposed methodology achieved an average Accuracy of 96.12\%, Sensitivity of 96.32\%, Specificity of 95.84\%, and F−1 score of 0.9609 on MESSIDOR and IDRiD datasets. Conclusion: These excellent results establish the validity of the proposed methodology for use in DME screening and solidifies the applicability of the HE-CNN classification technique in the domain of biomedical imaging. © 2020 Singh, Gorantla. This is an open access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.

Fulltext

PLoS ONE

DMenet: Diabetic Macular Edema diagnosis using Hierarchical Ensemble of CNNs

Concentrated solar power tower (CSPT) technology coupled with molten salt thermal storage is a very promising approach to harness solar energy as it offers good efficiency along with cheap storage and it can thus meet the fluctuating demand for electricity in a grid. This helps to avoid additional investment in fossil fuel–powered plants used for base load and peaking demand. However, their large-scale adoption is constrained by their relatively high capital cost. One of the most important cost elements in a CSPT plant is the cost of heliostats themselves. Normally heliostats use square-shaped reflector, whereas it is reported that heliostats with reflectors of aspect ratio (width/height ratio) of approximately 2.0 are cheaper due to reduced wind loads on them. Further, simulation of heliostat fields populated with rectangular heliostats of an aspect ratio of 2.0 has shown that such heliostat fields are not only cheaper to build but also optically more efficient. This paper discusses the results of a small experimental setup created to validate the results of these simulations. © 2020, © 2020 Taylor & Francis Group, LLC.

Energy Sources, Part A: Recovery, Utilization and Environmental Effects

Effect of aspect ratio of heliostats on the optical efficiency of solar tower power plant–an experimental analysis

Silica Nanoparticles (SiNPs), due to its tunable properties and composition, meets the needs in specified applications. Rice Husk (RH), an agricultural waste, abundant in amorphous silica has its extensive application in various industrial products. However, the presence of impurities in RHs makes the extraction process difficult and optimum experimental conditions are needed to achieve maximum yield with specific properties for its specified applications. In this work, Rice Husk Ash (RHA) is prepared from RH by acid leaching followed by silica gel extraction from RHA via controlled pyrolysis at three temperatures (700°C, 750°C and 800°C). Pyrolyzed RHA is then dried and pulverized to get SiNPs. Characterization of SiNPs is done using UV-Vis Spectroscopy, X-Ray Diffraction (XRD) and Scanning Electron Microscopy (SEM). UV-Vis Spectroscopy and XRD analysis shows that the SiNPs extracted from RHA at 750°C exhibited the highest peak at 260 nm and 2θ=26° respectively and XRD and SEM analysis show that crystallite and particle size is in close agreement with each other. The optimum temperature being 750°C, silica gel formed at a pH 6 is more stable, transparent and thicker in comparison to the silica gel formed at varied pH. Moreover, the sodium silicate sample exhibits a bright green fluorescence and the intensity of fluorescence is higher for the sodium silicate prepared at 750°C range than at any other temperature. © 2020 Scientific Publishers. All rights reserved.

Chemical Engineering

Journal of Scientific and Industrial Research

Extraction of silica nanoparticles from Rice Husk Ash and its characterization

Indoor occupancy estimation has become an important area of research in the recent past. This work investigates the feasibility of an Internet of Things (IoT) based university classroom occupancy estimation system. As IoT devices generate voluminous data at high rates, the centralized cloud computing approach is found to generate high latencies. The client server based cloud architecture has been compared with the decentralized edge computing architecture for building the occupancy estimation system. The performance of these architectures has been compared using two performance metrics: latency and throughput. The occupancy estimation models using carbon dioxide and relative humidity as inputs, have been developed using multiple linear regression and quantile regression. Root Mean Square Error (RMSE) and Mean Absolute Percentage Error (MAPE) have been used to compare the performance of our estimation models. © 2020 The Authors. Published by Elsevier B.V.

Procedia Computer Science

IoT-based Indoor Occupancy Estimation Using Edge Computing

Good air quality is critical to healthy and productive life in the indoor environment. Quality of air inside buildings depends on several factors. Till now, it was not possible to measure and monitor indoor air quality (IAQ) through mobile phones and wireless sensors in real time, as standardized miniature sensors were not commercially available. With the introduction of wireless air quality sensors that can pair with most contemporary smartphones, it is now possible to monitor indoor environment pervasively in real time, take precautionary measures, and maintain a healthy life. In this work, we have created an IoT sensing system to monitor and analyse the time variation of carbon dioxide in a university classroom. Using the sensed data, forecast models have been developed to predict the build-up of carbon dioxide, an indicator of IAQ. © Springer Nature Singapore Pte Ltd. 2020.