Skip to main content
main-content
Top

2021 | Book

Computer Networks and Inventive Communication Technologies

Proceedings of Third ICCNCT 2020

Editors: S. Smys, Ram Palanisamy, Álvaro Rocha, Grigorios N. Beligiannis

Publisher: Springer Nature Singapore

Book Series : Lecture Notes on Data Engineering and Communications Technologies

share
SHARE
insite
SEARCH

About this book

This book is a collection of peer-reviewed best selected research papers presented at 3rd International Conference on Computer Networks and Inventive Communication Technologies (ICCNCT 2020). The book covers new results in theory, methodology, and applications of computer networks and data communications. It includes original papers on computer networks, network protocols and wireless networks, data communication technologies, and network security. The proceedings of this conference is a valuable resource, dealing with both the important core and the specialized issues in the areas of next generation wireless network design, control, and management, as well as in the areas of protection, assurance, and trust in information security practice. It is a reference for researchers, instructors, students, scientists, engineers, managers, and industry practitioners for advance work in the area.

Table of Contents

Frontmatter
A Multi-hop Energy-Efficient Cluster-Based Routing Using Multi-verse Optimizer in IoT

Wireless sensor networks (WSNs) constitute an immense number of mobile or stationary sensor nodes that rely on multi-hop and self-organization. Internet of Things (IoT) is based on interconnected objects which are capable of communicating with each other and also collecting data related to their situation. Hence, for IoT networks, rather than energy-efficient communication, the most crucial challenge for conserving the cost of network management is the reduction of Internet connection numbers. To extend the network lifetime and to avoid interference, an extensively applied method is the efficient usage of energy. The objective of the low energy adaptive clustering hierarchy (LEACH) is the selection of sensor nodes as the cluster-heads (CHs) at several cycles such that there is the acquisition of the extreme power excess’s outcome and, later, its dispersal in the whole WSN. Tabu search (TS) is a metaheuristic which controls the local heuristic search procedure such that there is the exploration of the solution space beyond its local optimality. Multi-verse optimizer (MVO) algorithm is influenced by astrophysics’ multi-verse theory. This work has proposed a multi-hop energy-efficient cluster-based routing with multi-verse optimizer.

Vimal Kumar Stephen, Sanjiv Sharma, Antonio Rutaf Manalang, Faiza Rashid Ammar Al-Harthy
A Study on Suffix Trees and Their Applications in Genome Sequences Using MUMmer

A suffix tree is a data structure that stores a string of text. In bioinformatics, the suffix trees can be used as a tool to align the entire genomes of closely related organisms, to compare sequence assemblies at different phases and to compare and analyze the results. This paper focuses on the aligning of various whole genome sequences, using suffix trees. In the present investigation, the implementations of the suffix trees are discussed, focusing on two models for comparison and alignment. The first model is multiple sequence alignment based on a suffix tree and center-star strategy (MASC), which is a leading method to find solutions for the alignment of evolutionarily related sentences, considering evolutionary events like mutations, insertions, deletions and rearrangement. MASC applies to deoxyribonucleic acid (DNA), ribonucleic acid (RNA) and protein sequences. The second model maximal unique match (MUMmer) is discussed. Multiple genome alignment tools (Clustal Omega, Clustal W, MUSCLE, MAFFT and Mauve) are compared with MUMmer and are proving the efficiency of MUMmer, because of the use of the suffix trees.

Elena Cirkoska, Chandra Kishore, Subash Chandra Bose, Azath Mubarakali, T. R. Reshmi, Ninoslav Marina
Analysis and Design Recommendations for Nepal Tourism Website Based on User Perspective

Tourism websites are the face of the tourism industry, and tourism websites must be well designed to meet user expectations. This work evaluates and analyzes the user requirements to propose design solutions and recommendations for the tourism website of Nepal. The study is drawn on the basis closed-end survey and technical testing of the official tourism website of Nepal. The samples of the study comprise of inbound tourist from 10 different countries and domestic tourist from student, businessmen, government officials, and IT experts group. The survey is exploratory based on convenience sampling. An interview with selected respondents is taken to refine requirements and validate the web model in an iterative process. The work proposes a web design model for the tourism industry of Nepal built on web design principles of content quality, design quality, organization quality, user-friendliness, and technology adaptation. The proposed model includes different rating algorithms and features like recommender systems, social networks, user customization, e-commerce, and search optimizations techniques for meeting comprehensive user needs. This work is important for the tourism officials, tourist, and businessmen to get reliable and detailed information on the tourism industry of Nepal. It serves as a knowledge base for the website developers, government officials to take care of issues, and requirements while designing tourism websites.

Deepanjal Shrestha, Tan Wenan, Neesha Rajkarnikar, Deepmala Shrestha, Seung Ryul Jeong
Smart Residence Incorporating with Internet of Things (IoT)

With the advent of technologies and along with the progress of Internet-based information society, the rapid growth of the automation system has been evolved around us. Internet of Things (IoT) is most spelt ones. In this paper, IoT is applied in the automation of home appliances at a low cost. This smart residence is based on Arduino with the necessary interface to enable Internet and the control of power through an Ethernet shield with relay shield, and the project consumed more real-life interactions along with embedded software solutions. The proposed system does not need any physical presence to control the electrical appliances of home. Smart residence can be controlled by a mobile app or any modern web browsers. The home appliances connected with the electric switch can be turned on/off remotely. The users also can check the status of the home appliances, whether turn on/off. The proposed system also able to report the current state of temperature, humidity, and light intensity of the room. In future, such a system will be enhanced further to incorporate intelligence in the system.

Mst. Eshita Khatun, Debashish Kumer Shingho, Md. Saddam Hossain, Syed Akhter Hossain
Packet Priority Scheduling for Data Delivery Based on Multipath Routing in Wireless Sensor Network

Wireless sensor network (WSN) is a collection of sensor nodes with limited power supply and limited transmission capability. Forwarding of data packets takes place in multi-hop data transmission through several possible paths. This paper presents a packet priority scheduling for data delivery in multipath routing, which utilizes the notion of service differentiation to permit urgent traffic to arrive in the sink node in a suitable delay, and decreases the end-to-end delay through the distribution of the traffic over several paths. During the construction path phase, from the sink node to the source node, the packet priority scheduling multipath routing (PPSMR) utilizes the remaining energy, node available buffer size, packet reception ratio, number of hops, and delay to select the best next hop. Furthermore, it adopts packet priority and data forwarding decision, which categorizes the packets to four classes founded on reliability and real-time necessities, and allows the source node to make data forwarding decision depending on the priority of the data packet and the path classifier to select the suitable path. Results show that PPSMR achieves lower average delay, low average energy consumption, and high packet delivery ratio than the EQSR routing.

Abdulaleem Ali Almazroi, M. A. Ngadi
Design of Treadmill Cycle-Walking on Wheels

Treadmill e-cycle is a brand new way of moving. This paper presents an e-cycle, the frame of which is altered, and the belt of the treadmill is placed amidst two wheels on which the user can walk. As a person paddles to move forward in a normal bicycle, similarly in this bicycle, the person will have to walk to move forward. When the user treads, the belt moves rear. It can be driven with the help of electricity also. For electric motor support and e-cycle mode, it is fitted with a 24 V brushless dynamo motor that can be easily controlled with a controller. The controller of this motor gives us the provision of the headlamp, rear lights and a switch to on/off the motor. To power this motor, two 12 V batteries of 14 A current rating are used. Multi-purpose treadmill e-cycle consists of a monitor attached at the handle which will display the time elapsed, speed, body mass index, heart rate, etc. It will serve the purpose of both exercises and travelling at the same time. What makes this e-cycle unique are the modes in which it can be used, like normal treadmill, a treadmill with complete manual mode, e-cycle with electric motor support and standing e-bike.

Devam Garg, Astha Yadav, Garima Garga, Akhilesh Singhal, Mandeep Singh, Sunil Kumar Singla
Survey on Edge Computing–Key Technology in Retail Industry

Throughout the past few years, there have been few industries that have seen as much change with the development of innovations as the beginning of the Fourth Industrial Revolution in the retail sector. The Internet of things (IoT), advanced signage, big data and artificial intelligence (AI) have all appeared to include much potential inside a retail situation and are presently being received by organizations and associations around the globe. The multiplication of data over the years and achievement of rich cloud administrations have driven the skyline of another computing paradigm. Edge computing calls for preparing information at the edge of the system. By bringing edge computing framework into brick-and-mortar stores, retailers are exploiting the advantages being delivered by forefront advancements and broadening the life expectancy of such stores in a period of challenges and competitions with online retailers, for example, Amazon helping them in gathering insights about customer purchase patterns and trends using data analytics. This paper describes the importance of edge analytics in various computing technologies and retail industry.

Abhiraj Biswas, Ayush Jain, Mohana
Fall Prevention System for Workers Working on High Towers

Fall prevention has always been taken under mere consideration and is often ignored, but fall from heights is the biggest concern that cannot be neglected. The methods and devices so far designed to prevent the fall from towers are not too worthy to resolve and in such cases the people working on high towers are meant to be guarded with ropes or ladders. If the ropes are not tethered properly else if the ladders are detected with internal faults, then it will eventually lead to an emergency call. The fall is predicted on the basis of the working of the heart and pulse level of that individual; further it is assumed that the other protective equipment are not working at that particular point of time, where this device will play a crucial role. An emergency message or call is immediately transmitted as soon as the proposed model detects the abrupt increment in pulses. In this way, the proposed model can protect the person from head injuries or even death.

Rama Prabha, Shivam Kumar
Optimization of Merge Hypothesis for Automatic Image Caption Generation

Image captioning is the process of automatically generating appropriate image descriptions of the given images without human intervention and using appropriate algorithms. These descriptions should be in proper sentences with the natural language; hence, the algorithms used to require the fusion of computer vision techniques with those of natural language processing. The rise of social media and cybercrimes has substantially increased the amount of research done in this area. The objective is to generate image captions that are semantically closer to those generated by humans. This paper focuses on the merge architecture used in image captioning process and briefly compares it with the inject architecture also used to find image captions. Optimization of the general algorithm is done by using the vector embeddings of the words obtained from the dataset itself and using concatenation in the merging process. Finally, the performance of the proposed method is compared using a similarity measure and BLEU score.

Abhilasha Sharma, Nitin Gupta, Nitin Kumar, Nishant Sidhu
Comparative Study of Machine Learning Techniques for Chronic Disease Prognosis

Chronic diseases are the leading cause of death and disability worldwide. Researchers have contributed to the prediction of chronic diseases. However, the general focus is to predict a single disease using dataset and methods appropriate only for that disease. A comparative study of machine learning algorithms is proposed for the effective prognosis of five major chronic diseases using behavioral risk factors through a single dataset. The logistic regression and random forest models are trained using real-world data collected from the USA in 2017 by BRFSS. The user input is collected via a chatbot, the percentage of occurrence of chronic diseases is predicted, and modeled via interactive data visualization techniques to provide suggestions to lower the risk. By selecting the appropriate model for each disease through comparative study, the prediction accuracy of 91.2% is reached for heart disease, 71.8% for arthritis, 91.7% for pulmonary disease, 96.1% for kidney disease, and 86.8% for diabetes.

Geetanjali Bhola, Aman Garg, Manisha Kumari
Biometric System: Unimodal Versus Multibiometric Fusion and Its Current Applications: Review

The biometric system has developed a lot from manual verification to artificial intelligence-based multi-biometric systems. The biometric system has a vast area of application and a potential market of billions. The types of a biometric system namely unimodal and multi-biometric fusion model are discussed. A unimodal biometric system collects only one biometric trait and runs on only one mode. A unimodal biometric system faces a lot of problem despite having high accuracy and fewer error rates. Issues like intra-class variation, common biometric features, and most importantly, spoofing attacks are the major problem with the unimodal biometric system. The best solution to these problems is a multi-biometric fusion system. Multi-biometric fusion combines two or more individual biometrics to decrease the error rates and make it secure from the spoofs. Fusion techniques play a significant role in multibiometric systems because the performance depends on the method used in the fusion process. This paper discusses the importance of the biometric system in its current applications. A comparison between unimodal and multi-biometric fusion system is also performed.

Mahesh, Rahul Kumar, Kapil Sharma
Real-Time Video Surveillance System for Detecting Malicious Actions and Weapons in Public Spaces

In today’s world, thousands of surveillance cameras have been working round the clock. These cameras are installed at railway stations, ATMs, streets and all the public spaces that one must have come across in their day-to-day life. A wide variety of malicious events take place even in these locations. Security personnel has to monitor these footages continuously round the clock. But humans are not perfect; they do not have a perfect vision nor the ability to concentrate for a longer span. Hence, it is very obvious that at some point, something abnormal might happen that the security personnel might fail to notice. The definition of the abnormal event differs from one situation to another. An automated system is being proposed using deep learning and pose detection to detect unusual events in such scenarios. Such a system has wide applications in many industries. There is an increasing demand for developing an automated surveillance technique that is fast and accurate in real-world applications. The main aim is to alert the authorities while the crime is being committed to a certain premise. Currently, very few techniques are available which lack accuracy. But using pose estimation techniques for human action recognition fast and accurate results can be observed.

Mahadevan Narayanan, Suyash Jaju, Akash Nair, Apurva Mhatre, Avantika Mahalingam, Anindita Khade
Study of New Trends in Precision Agriculture

The precision agriculture field is evolving towards the Industry 4.0 era, which will help to increase production in agriculture field with the use of limited resources available. This evolution is resulting in greater varieties and yield of crops. In recent agriculture development, the advancement in various technologies such as the Internet of things and artificial intelligence has played a major role in data-driven and automated agriculture. This article has provided a review on recent development in agriculture technologies and applications as well as research efforts in the context of smart farming. Future direction in precision agriculture is also presented.

Dipali K. Dakhole, S. P. Jeno Lovesum
AgriERP Using Salesforce for Cloud Computing—An Application Intended for Agribusiness

Agribusiness is a sector purely intended to promote, develop and perform business based on agriculture requirement and production, which emphasizes and focuses on the farmers, farming and farming-related commercial activities. This sector and business have now become more complex and complicated with the latest regulations, government involvement and enhanced demand for the supply chain to customers directly from the buyer. Bringing ERP into agriculture business will surely going to improve the business analysis, farming capabilities, technology enables farming and raising the standard and livelihood of farmers. ERP is a way or strategic mechanism adopted by many industries to handle and combine the requirement of their working in the business process. Some ERP applications are playing a very vital role in the growth and development of the industry especially in the fields of agro farms and agribusiness. This paper proposed a cloud-based ERP application that covers key processes supported by modern ERP systems. It improves operational efficiency, accuracy and cost-effectiveness of core business activities through industry best practices and is intended especially for agriculture and farmer requirements, developed in the Salesforce. The proposed application provides numerous benefits in comparison with normal agricultural techniques.

Akhilesh Kumar Singh, Manish Raj, Vivek Sharma
Facial Expression Recognition System Using Different Methods

Facial expressions are one of the nonverbal channels through which the emotional state of a human being can be known.Mishra, Sweta These convey the emotional state of an individual to others.Talashi, Sheweta There are four basic emotions—happy, sad, surprise and neutral. Facial expressions are identified using various parameters, e.g., distance measured when eyebrows are raised gives us a particular facial expression. Likewise, thickness of lower lip, upper lip, width of mouth, etc., give us different types of facial expressions. The proposed system identifies these expressions using convolutional neural network (CNN) and facial action coding system (FACS) with AdaBoost method. The proposed work has used FER2013 data set and also has created own data set consisting of 80 images each of happy and sad expressions [1].

Sweta Mishra, Sheweta Talashi
Robot Navigation Through QR and Text Detection

In the rapid advancement of today’s technological world, smarter technology is believed to solve problems. QR code detection and text recognition are one of the extensively using applications. So, retrieving information using these applications accurately is one of the important factors. In this paper, the accuracy rate of both the applications is briefly characterized in terms of time consumption for retrieving information by varying distance. The experimental set-up includes a camera fixed on the robot vehicle and navigates as per the given QR and text instructions. The accuracy rate is analysed with respect to the response time of the robot. Finally, the performance of both approaches has been briefly discussed.

Rajesh Kannan Megalingam, Sathi Lakshmi Madhuri, Tangudu Santhoshini, Varsha Juluri
Investigating the Impact of Data Analysis and Classification on Parametric and Nonparametric Machine Learning Techniques: A Proof of Concept

Supervised algorithms depend on the given data for categorizing. In present work, we used both parametric and nonparametric types of classifiers. We intend to compare the performance of four popular machine learning classification algorithms—Naïve Bayes, decision trees, logistic regression, and random forest on two popular benchmarked datasets—wine quality dataset and glass identification dataset. To get a wide angle of the performance of these algorithms, we incorporated both binary and multi-class classification which also solved the problem of imbalance in the dataset. In current work, we compare and demonstrate various supervised machine learning classification algorithms on the two well-known datasets. The performance of the algorithms was measured using accuracy, recall, precision, and F1-score. It was observed that nonparametric algorithms like random forest classifier and decision tree classifier bested the parametric algorithms like logistic regression and naïve Bayes. Moreover, as the datasets were imbalanced, we figured out which algorithm performs better under what circumstances. In particular, random forest achieved best performance in terms of all considered metrics, with accuracy of 82 and 83% in wine datasets and 79% in glass identification dataset.

Sarvesh Khire, Pushkar Ganorkar, Aseem Apastamb, Suja Panicker
Analysis of Machine Learning Algorithms for RFID Based 2D Indoor Localization

Nowadays, radio frequency identification (RFID) localization techniques have been widely used in indoor positioning systems (IPS) due to their low cost and ease of deployment. The main reason for the rise in RFID localization is because of inaccuracies faced by the global positioning system (GPS) in the indoor environment due to multi-path interferences of signals. The localization methodology based on received signal strength indication (RSSI) technology for indoor RFID is currently a hot topic. Because RSSI obtained is highly dependent on environments, classic algorithms like trilateration will lead to huge errors in the accuracy of the localization. This paper introduces a novel approach for RFID based indoor localization by making use of machine learning algorithms such as artificial neural network (ANN), support vector machines (SVM) and K-nearest neighbors (KNN). Hyperparameter tuning is incorporated for increasing the accuracy of the models. Experimental results show that the ANN algorithm remarkably improves the indoor localization accuracy and is also effective in tackling nonlinear changes in radio frequency signals. Moreover, the proposed model can be used in similar environments.

S. Aravind Raamasamy, P. Shanmuga Pradeep, C. H. Mani Madhav Goud, C. A. Viswanathan Babu, M. Jayakumar
Survey on Techniques in Improving Quality of Underwater Imaging

The quality and appearance of underwater images perform a relevant role in the underwater computer vision paradigm. Wherein, the underwater images are useful in various applications to make a detailed study on underwater life. Despite the hype, the images caprited underwater undergoes various challenges like light attenuation, color absorption, type of water, etc. To address these types of issues, many algorithms are proposed. This paper provides a comprehensive study of frequent methods used to intensify the visual nature of underwater images and different underwater image datasets that are used to perform tasks regarding underwater imaging. The different quality assessment measures are also summarized in this research work. The presented methods and their shortcomings are studied to enable the in-depth comprehension of underwater image enhancement. In the end, possible future research directions are also implied.

Nagaraj V. Dharwadkar, Anjali M. Yadav
Speech and Kinova Arm-Based Interactive System with Person Tracking

This paper presents the integrated module of speech, person detection, and an assistive robotic arm. This whole project is developed in Linux Ubuntu 16.04 using robot operating system (ROS) platform in which the kinetic version is used. Speech to text and text to speech code is developed using Google API. Person detection is performed using lidar using a Kalman filter and global nearest neighbor data association algorithms. Assistive robotic arm package of Kinova Jaco is used. An external microphone is used to record the voice of a person, and RPLiDAR is used to detect, track persons, and to predict estimated 3D coordinates of a person in real time. This is an autonomous model with no manual control by any other persons, it is also a fixed system where both lidar and Kinova arm base are fixed to a table, and it can work efficiently when a person is at a distance below the 1.2 m to both lidar and arm.

Rajesh Kannan Megalingam, GaddeSakhita Sree, Gunnam Monika Reddy, IntiRohith Sri Krishna, Sreejith S. Pai, Balla Tanmayi
Comparative Study of Cooling Solutions of a Drone Based on Raspberry Pi Deducing the Most Efficient Cooling Method

Automation and robotics are recently being included in diverse applications, as they have numerous advantages and extensive research options. IoT applied in a vast area of projects and researches is very useful. One of the most talked-about robotic equipments for a diverse field of applications is the drone. Here, our research paper examines a drone based on a Raspberry Pi 3B+ model. Cooling the heat produced in the CPU is done via several methods. Concerning the heating, four cooling methods are compared and studied. The best one out of some considered active and passive cooling devices is attempted to find. The heating of the CPU is studied because heat can damage the components of the Raspberry Pi. Heat can also cause efficiency issues. The damage to Raspberry Pi from the heat will also affect the drone. Finally, from our study, it is deduced that a 40 × 40 × 20 mm heatsink applied with a thermal solution proves to be most effective in cooling the CPU, which is also indicated by our calculations.

Rohit Beniwal, Sanjay Patidar, Rohan Tomar, Shekhar, Rohit Khatta
ROS-Based GUI Controlled Robot for Indoor Mapping and Navigation

Path planning is one of the key tasks in robotics, which is used for navigation of the robot. Any mobile robot should possess the ability to navigate freely in its environment. This is important to avoid collision with obstacles, move to a certain point as per the function, etc. This research paper reviews the design of an autonomous mobile robot controlled by a graphical user interface(GUI) which possesses the ability to autonomously go to a room in a created environment when it receives the command through GUI. To achieve this, a simple robot is designed in solid works and constructed which is used for testing this application. This function can be applied to use in complex robots such as a service robot, resto bot and firefighting robot. Mapping and autonomous navigation are performed using the linorobot stack. The robot is developed in Linux (Ubuntu 16.04) using the robotic operating system (ROS). The linorobot navigation stack and GUI are separately used till date. In this research work, the results are forwarded when both navigation and GUI are integrated. The performance of the robot is evaluated by testing the ability of the robot to avoid obstacles and reach the destination in different scenarios. The discussion of the scenarios and the performance of the robot are discussed in experimental results.

Rajesh Kannan Megalingam, Balla Tanmayi, Gunnam Monika Reddy, Inti Rohith Sri Krishna, Gadde Sakhita Sree, Sreejith S. Pai
Mechanism for Saving Base Stations Energy Using Binary Particle Swarm Optimization

Energy is a precious resource and that has to be saved for a sustainable hassle-free future. The main objective of this paper is to save the base stations energy in order to increase the nodes lifetime in a network. This paper focuses on 5G networks by considering a heterogeneous nature of cells, i.e., macro and small cells. Both low-data and high-data traffic rates are taken into account. The base station will be serving the user environments that tend to overlap the nearby base stations area. Therefore, making the other base station to remain in a sleep state, and save the base stations energy. A binary particle swarm optimization is formulated for solving this approach to save the base stations energy. The results obtained are compared with the conventional schemes, and it is inferred that the proposed approach is better than the existing approaches. The aggregate delay is less according to this proposed method.

A. D. C. Navin Dhinnesh, T. Sabapathi
Security Threats and Privacy Issues in Cloud Data

The quick advancement of Web-based applications has led to a huge amount of information being scanned and gathered for business examination or scholarly research purposes, which may disregard individual protection. Organizations, industries and individuals data are at stake. In this paper, utilizing on the Web Personal Health Record (PHR) as contextual analysis, first demonstrate the need of inquiry ability approval that lessens the security introduction coming about because of the list items, and build up a versatile structure for authorized private keyword Search (APKS) over encoded cloud information. This particular model proposes two novel answers for APKS given on-going cryptographic crude, hierarchical predicate encryption (HPE). Our answers empower efficient multi-dimensional watchword looks with a run question, permit assignment and renouncement of hunt abilities. Additionally, the proposed system improves the question protection which conceals clients’ inquiry watchwords against the server. Actualize our plan on an advanced workstation, and exploratory outcomes exhibit its appropriateness for reasonable use. Privacy has seen advancement lately as information mining of the datasets in a dispersed huge information condition has turned into a successful worldwide business which is none other than data management or data analytics which ensures the security of data.

Sanjana C. Shekar, K. Sriram, N. Jayapandian
Adaptive Frost Filtered Quantile Regressive Artificial Deep Structure Learning Framework for Image Quality Assessment

An image quality assessment (IQA) aims for predicting the quality of the images with or without any previous information of reference image. Assessing image quality is significant in image communication and processing. There are several techniques been proposed for quality assessment, but the accurate feature learning and complexity analysis are still challenging issues in various image processing applications. To improve the quality assessment with minimum complexity, an adaptive frost filtered quantile regression-based artificial deep structure learning (AFFQR-ADSL) framework is introduced. The AFFQR-ADSL framework performs the image quality assessment based on two methods, namely full reference and no reference. The AFFQR-ADSL framework with full reference (AFFQR-ADSL-FR) is carried out based on the test image with the reference image information. Initially, the numbers of input images and reference images are collected from the image database. The proposed AFFQR-ADSL-FR framework trained the input images and reference images with different layers to progressively learn the higher-level features from the raw input. The numbers of images are given to the input layer of the artificial feed-forward deep structure learning network. Then, the inputs are transferred into the hidden layers to repeatedly learn the features. In the first hidden layer, the input images are preprocessed to filter the noise present in the input images using an adaptive mean frost filtering technique. Followed by, the input images are divided into the number of patches and feature extraction is carried out in the next successive hidden layer. Then, the learned features are combined and fed into the output layer. Finally, the proposed technique uses linear quantile regression at the output layer for analyzing the extracted feature vectors and obtains the quality assessment results. Then, the AFFQR-ADSL framework with no reference (AFFQR-ADSL-NR) is carried out by using the test image without using the reference image information, extracts the feature vectors, and provides the assessment results at the output layer. Result evaluation is carried out with metrics such as mean square error, peak signal-to-noise ratio (PSNR), accuracy, and computational complexity (CC). The qualitative and quantitative results show that the proposed AFFQR-ADSL-FR and no reference achieve better results in terms of PSNR, error, and CC.

M. Dharmalingam, S. Sathyamoorthy
Classification of In-House Managed Equipment by Listing Its Parts

This paper proposes a novel research work, where the manual data are collected from various parts of the devices such as patient monitoring and syringe pump, and then the collected data are calibrated by self-organizing maps (SOM) using k-nearest neighbors (KNN), Naive Bayes, and support vector machine (SVM) classifiers. Sensitivity and specificity are used as the evaluation metrics. Comparing the accuracy and sensitivity parameters in naïve bayes and KNN classifier, the naïve bayes is performed better than KNN algorithm.

S. V. Aruna, R. Karthika
Applications of Association Rule Mining Algorithms in Deep Learning

This paper focuses on the different association rule mining algorithms and their applications in modern fields of deep learning and neural networks which form the pillar stones of new age problem solving. Association rule mining algorithms are categorized under “if–then” category as they have an antecedent (if) and a consequent (then). Deep learning is a machine learning technique which uses neural networks to pass the input to obtain outputs. The field of deep learning has become ubiquitous in all fields of problem solving due to its ability to accept raw inputs without feature extraction but this leads to overfitting. The algorithms Apriori and FPGrowth help in finding an association between features of the input which then can be reduced to a handful of features which then can be given to deep learning models. This paper attempts to explain these algorithms and their use in deep learning.

Sai Kishore, Vikram Bhushan, K. R. Suneetha
Extraction of Opinion Targets and Words from Reviews Using Collective Parallel Cluster Algorithm

Online customer reviews are considered to be a significant informative resource that is useful for both customers as well as manufacturers. The manual scanning of reviews makes the process difficult and thus the evolution of opinion targets and words arise. In the last decade, several techniques have been proposed for the extraction of opinion targets and words which have certain limitations such as long-span relations and syntactic pattern errors. In addition, the proposed methods consume more time in case of a huge volume of reviews with some erroneous review statements. To reduce the time consumption, this paper proposes a parallel collective clustering algorithm for the better extraction of ordered pairs in a shorter time despite the errors. The proposed parallel collective clustering algorithm utilizes the concept of Parallel Word Alignment Model for the generation of opinion targets and words. Extensive experimentation has been carried out with the larger reviews collected from Amazon and the proposed technique has been evaluated using familiar metrics such as Precision, Recall and F-Measure in addition to execution time. The experimental results show that the proposed approach is flexible and effective for a larger dataset with erroneous review statements.

L. Rasikannan, P. Alli, E. Ramanujam
Position Accuracy Enhancement of a Low-Cost GPS Receiver Based on Machine Learning Approach

Agriculture sector is growing rapidly in the modern era due to rise in global population and shift in trade policies. Replacing human labor with automation of agricultural equipment can alleviate labor shortages and rising costs of farm work to a greater extend. With the inclusion of GPS for navigation, IoT connectivity for remote monitoring and operation, cameras and machine vision systems, the autonomous vehicles will become more capable and self-sufficient over time. However, automation costs need to be reduced to gain a broader acceptance. Tracking a vehicle with low-precision GPS receivers tends to have errors, whereas the inclusion of a high-precision GPS receiver raises the overall cost of the vehicle. In this work, a machine learning-based post-processing techniques have been proposed to enhance position accuracy of a low-cost GPS receiver implemented on an autonomous harvester for monitoring and navigation purpose. The proposed system is integrated with cloud technology where the location coordinates from low-cost GPS receiver get updated to ThingSpeak cloud server using microcontroller. The user can track the position of the vehicle on a smart phone using the designed android app and configure the Geofence around the field for the harvester vehicle to work on. The Geofence alerts the user once the harvester vehicle moves out of the designated field.

Robin Thomas, Binoy B. Nair, S. Adarsh
Blockchain-Based Secure and Efficient Crowdsourcing Framework

Blockchain is an emerging Web-based technology that being distributed and decentralized, and public ledger provides the capability to build applications that are more secure, reliable and trustworthy. Crowdsourcing is one of the use cases of blockchain technology. It allows an organization or an individual to utilize the talent of individuals over the Internet in exchange for some rewards. Centralized crowdsourcing platform has many limitations, beginning from centralized storage to reward distribution for task completion. Blockchain-based approach for crowdsourcing solves many of the problems posed by centralized crowdsourcing. In this work, comparative analysis of various proposed blockchain-based crowdsourcing approaches has been done and Ethereum-based crowdsourcing platform has been proposed, that efficiently deals with Sybil attack, solution confidentiality breach, unbiased evaluators’ selection and task evaluation using techniques such as bit commitment and elliptic curve-based ElGamal cryptosystem.

Prerna Goel, Mohona Ghosh
Driver Drowsiness Detection System Using Conventional Machine Learning

Forewarning drowsy drivers can reduce the number of road accidents. A non-intrusive drowsiness detection system is implemented, which alerts the driver on the onset of drowsiness. A Pi camera module attached to Raspberry Pi is used to acquire and process the live video of the driver. Haar face detector in OpenCV is used for face detection followed by 68 points of facial landmark identification. Eye and Mouth Aspect Ratios, blink rate and yawning rate are the features extracted. Drowsiness detection is done using two methodologies viz. a threshold-based one and the other, employing artificial intelligence. The machine learning techniques used are LDA and SVM. Feedback is provided as an alarm if a driver is found to be drowsy. The analysis shows that machine learning-based techniques viz. LDA and SVM outperform threshold technique for the dataset considered.

Radheswarreddy Madireddy, Dulla Sai Krishna Anudeep, S. S. Poorna, K. Anuraj, M. Gokul Krishna, Ankisetty Balaji, Dammuru Jaideep Venkat
Blockchain-Enabled Microfinance Model with Decentralized Autonomous Organizations

Microfinance which promises alleviation of poverty to the less privileged, oppressed community of people and to empower womenfolk, will get enrichment if it is implemented in a blockchain with a digital currency. Microfinance promises to end the vicious cycle of debt, and poverty can be deployed in a permissioned or permissionless blockchain, and the stakeholders can form a decentralized autonomous organization which enables a transparent successful functionality. Blockchain which is a decentralized, distributed, the tamper-proof ledger can be used for reading and writing of transactions of self-help groups in a much transparent way. The decentralized autonomous organization enabled with smart contracts to trigger our intended functionality in blockchain will give the microfinance, it’s streamlined intended functionality with expert panel member’s contingency measures to meet contingency situations, angel investors to invest in the new proposals of self-help group(SHG) microfinance and initial coin offerings for funding of new projects with potential investors. The hash of the transactions is used with crypto addresses, and the elliptic curve digital signature algorithm is used for authentication. The peer pressure model of SHG leads to a 90% loan repayment with more than 50% of the line of credit for the modern-day microfinance loan applicants. The self-help group working model, know your customer of SHG, initial coin offering model is given as proof of concept in solidity code in remix IDE.

M. J. Jeyasheela Rakkini, K. Geetha
Attendance System Using Face Recognition for Academic Education

Manual attendance system in colleges is a tedious and time taking task for both the students and faculty. The teacher has to start counting from first to last student number and the last student has to pay attention from the first number to their roll number. The face-based attendance system which is very simple and easy to understand is proposed in this paper. Face recognition is playing a vital role in the field of Biomatrix. Modern cell phones are having a lock, based on face recognition, which is playing a smart role for authenticating someone Faster RCNN algorithm is used in this proposed system. Google’s TensorFlow open-source framework is used for the training data set of human faces.

Sharad R. Jadhav, Bhushan U. Joshi, Aakash K. Jadhav
A Comparative Study of Conventional Machine Learning and Deep Learning Models to Find Semantic Similarity

In today’s scenario, with many platforms trying to provide answers to every question that the users try searching on the internet, the ease of finding things is direct but constrained for a limited number of users. It becomes arduous to find the answers even though they already exist. The uncertainty is due to semantic ambiguity. With the freedom of thought and linguistic diversity, there is a need for new tools to resolve this ambiguity. A new multidisciplinary field, natural language processing (NLP), incorporated with machine learning and statistical techniques, provides powerful analysis. In this paper, the basic machine learning models and state-of-the-art model Universal Sentence Encoder (transformer and deep averaging network) are compared using NLP techniques to detect duplicate questions in the Quora dataset. This paper aims to find which models out-do the others and the performance measures that affect each model such that there is a better understanding of the requirements to get the best results in finding semantic similarity between textual questions

Akshi Saxena, Pranati Shete, Sonal Sharma, Naina Kaushik
Improved Classification of Content-Based Image Features Using Hybrid Classification Decision

Feature vector extraction is a significant aspect of content-based image classification. Researchers have proposed multiple techniques for representing the image content in the form of feature vectors using important image properties like shape, colour, texture, etc. However, a single feature vector extracted using a particular technique is mostly unable to capture important details of images. This work has attempted a decision fusion-based classification approach using two different features extracted with image binarization and image transform technique, respectively. The results of decision fusion for classification have outperformed the individual approaches.

Rik Das, Khushbu Kumari, Sudeep Thepade, Pankaj Kumar Manjhi
Parallel CLARANS Algorithm for Recommendation System in Multi-cloud Environment

Multi-cloud computing is the computing platform in the cloud environment, and it differs from hybrid computing that is being implemented in most of the IT enterprises. The primary concern in the usage of the multi-cloud platform is for the handling and processing of data for the effective prediction of the online users in the recommender system. In the existing system, there is no effective clustering approach available for large data set processing, and deployment platform is not with good accuracy and execution time. This paper proposes a novel methodology using the parallel CLARANS, an algorithm for clustering that was embedded with the temporal fuzzy formal concept analysis (FCA). This method employs the homomorphic MapReduce framework (HMRF)and the time-based slope one approach to enhance the multi-cloud computation for recommender system application. The usage of these methods aids the performance to a greater extent. The input of the user is preprocessed to reduce the data inadequacy and mapper-reduce concept implemented in the HMRF before being fed into the recommender algorithm. The output of the algorithm was evaluated based on the mean average error (MAE) that was employed to predict the QoS value. The recommended approach was compared with the existing methodologies and showed enhanced results.

K. Indira, S. Karthiga, C. V. Nisha Angeline, C. Santhiya
Assessment of Emotional State of the Speaker from Speech Signals

Nowadays, the interaction between humans and computer is growing significantly. Whereas, the human emotion recognition from a dataset consisting of audio speech and signals have gained significant research interest. The task of recognizing the emotional state present in a speech signal is a typical one. Numerous algorithms and models exist to classify human emotions. In this paper, a method has been proposed to extract features from audio speech signals from humans and then classify the emotions in the test dataset. Ryerson Audio-Visual Database of Emotional Speech and Song (RAVDESS) dataset has been used for the emotion classification. Features like Mel-frequency cepstral coefficients (MFCCs) coefficients, Chroma, etc., are used to classify human emotions. Multi-layer perceptron (MLP) classifier is used as the classification algorithm to classify emotions. 41.38% of accuracy has been achieved using this method.

Parul Rawat, Arun Sharma
Integration of Speech and Vision for Perception in Assistive Robots Using Robot Operating System

Perception is a boon which is one of its kinds for every human being around the world. In this modern world, where technology has been tremendously developing, there has also been an escalation in the number of people being born with vision impairment. This paper provides a solution with a system that serves as an ability to hear and see for the assistive technologies such as service robots and rescue robots which could help for the old and crippled. This research deals with the integration of an object recognition technique with speech recognition using the robot operating system (ROS) to come up with a speech-based visual perceiving system that can aid the visually impaired people. The proposed object recognition uses Kinect v2 camera and YOLO algorithm for efficient object recognition. Here, mutual communication between the visually challenged and robot is established. This paper also discusses the system architecture implemented and tested in real and noisy environments.

Rajesh Kannan Megalingam, Motheram Manaswini, Jahnavi Yannam, Tammana Akhil, Akhil Masetti, R. V. Rohith Raj
Speech Emotion Recognition Using CNN, k-NN, MLP and Random Forest

Emotion recognition from speech has become a hot topic among researchers. This paper describes several methods to recognize emotions from speech signals using machine learning algorithms such as a k-nearest neighbour, multi-layer perceptron, convolutional neural network and random forest. Short-term Fourier transform spectrograms and mel frequency cepstral coefficients were extracted from Berlin database of emotional speech. Spectrograms were used as input for CNN. While MFCC features were input to k-NN, MLP and random forest. Each classifier demonstrated satisfactory results in the classification of seven emotions (happy, sad, angry, neutral, disgust, boredom and fear) but MLP classifier was the most prominent with an overall accuracy of 90.36%. A comparison between the performances of these classification algorithms is also presented.

Jasmeet Kaur, Anil Kumar
Plant Leaf Disease Detection Using Machine Learning Techniques

It is generally known that the plants are a great source of energy, and it is a key-enabler to resolve the significant environmental issues like global warming. However, due to the sudden climatic changes and pollution, the plant ailments are becoming more aggressive within the sustenance of this necessary source by causing more environment losses. Hence, a significant research attention is required to analyze and reduce the plant ailments appropriately. In this perspective, convolutional neural network (CNN) has disclosed an extensive performance in the detection of various plant diseases by analyzing their leaves. This paper proposes a convolutional neural network technique, where the leaf ailment can be analyzed exactly when compared to the traditional disease detection techniques. If the plant ailment is already known, and in such case, the disease severity can be accurately analyzed by using the convolution layer and max-pooling layer proposed in this research work.

K. Sudha Rani, B. Priya Madhuri
Development of Wake Detection and Analysis by Using Image Processıng

Many areas in the ocean are prohibited and the ships, submarines pass through these areas with no continuous surveillance. The timely recognition of the trespasser’s detection is given an utmost importance. Equally important is the urgent action to seek the assistance. This project aims to develop wake detection from the SAR images by using image processing. The algorithm is of evaluating an L + S decomposition (low rank + sparse) supported by radon transform (RT) for noise reduction, sparse pixel variation detection. The L + S algorithm splits the sparse object matrix from the low-rank matrix. The system will extract the feature of the wake by texture analysis using automatic detection algorithm (ADA). A microcontroller is employed to analyze the signals. The ranges for ships and submarines are set. When signals are in a specific range, an alert will be sent to the respected officers. This system permits fast detection and feature extraction of maritime targets.

Franklin Jisno Jose, M. Rajesh, Vishnu Rajan
Enhancement of Dehazing Methods Using DNN and Filtering

Murkiness in pictures is because of common ecological wonders, which makes the picture in a white shade commotion. Fog evacuation is one of the most significant research subjects nowadays to the due prevalence of utilizations progressively reconnaissance from rambles or any region under security. Both indoor and outside pictures are significant for testing cloudiness and its expulsion. Many picture-handling methods are made by analysts to evacuate cloudiness in a solitary picture. Murkiness force can be determined by a parameter known as perceptual fog density measure (PFD). It is critical to investigate this parameter for all the strategies in order to get a thought of progress. In this postulation, another methodology is made by applying an all-inclusive guided sifting procedure with a profound neural system. This proposed calculation is executed on MATLAB programming, and results are gotten by figuring the PFD in the current and proposed method. The four methods are compared with one another. The methods are global image filtering (GIF), weighted global image filtering (WGIF), globally guided image filtering (GGIF), and a proposed strategy, for example, comprehensively guided sifting with deep neural network (DNN). In GIF, the fine structure of the picture is commonly not safeguarded and an unreasonable picture is acquired. In WGIF, the PFD got is most noteworthy. In GGIF, PFD is lower and structure is not safeguarded, yet in the proposed calculation, the PDF is least with fine structure, and shading power of the image is of the best quality.

Baby Naz, Nafisur Rahman, Md. Tabrez Nafis
Biometric Enabled Patient-Centric Automated Medication Dispenser Using IoT

Patients rely on their medications to keep them healthy, but complex medication schedules can lead to accidental errors like missing doses, incorrect dosage or wrong timing. This paper proposes an idea to provide the information automatically to patients to take the right medication dosages at the prescribed time. This would prevent unexpected hospital visits related to incorrect medication usage. This system uses LCD, microcontroller, real-time clock (RTC) module, an alarm system, a biometric fingerprint scanner and the Blynk software tool used to intimate the patients to take proper dosage according to the prescribed time. Compared to the conventional pillbox that requires users or caretakers to load the box every day or every week, this automated medication dispenser would significantly reduce caretakers or users’ burden on frequently preloading pills for patients.

Nishita Anand, P. Prathibha, Payal Purohit, Reeth Nalamitha, Chennagiri Rajarao Padma
Optimization of TailorGAN Using Log Hyperbolic Cosine Loss Function

Designing of the fashion clothes has always been a crucial task for the fashion designers as it needs a lot of thinking and creativity.Siddiqui, Atique This paper will simplify this task by taking two reference images into consideration,Mane, Abhishek whereas one image would be of garment, and the other image would be of target attribute (collar), further by combining these two images, the photorealistic images will be generated with the help of GAN.In this task, it will work on a new dataset called garment set dataset, which consists of collar/sleeve and garment images. The proposed work aims to train this dataset and refine the quality of the generated images.Bawankar, Meet

Atique Siddiqui, Abhishek Mane, Meet Bawankar
An Organized Study on Data Divulge Elimination and Discernment

Exchange of information is done in every organization. Information can be confidential and non-confidential. There are several channels through which exchange of data is done. During this process of exchange, data gets leaked. Enterprises are facing a lot of issues with data leakage and also towards recognizing the leakers. It is a very important need for the organization to exchange confidential information securely and easily. The main goal is to create the data leakage prevention and detection model which detects when the allocator’s sensitive information has been disclosed by operators and also to identify the leaker that leaked the data. Leakage of confidential information leads to a great loss to the organization. In addition to financial loss, an organization’s reputation is ruined and facing its identity stolen. These losses take a lot of time to recover. In this paper, it will focus on research papers on various techniques, which consists of classification strategies to identify sensitive and non-sensitive data, distribution strategy to monitor data, and encryption technique used.

Khushbu, Poonam Nishad, Vipin Kashyap, Ishu Gupta, Ashutosh Kumar Singh
Study of Simulation Environment of Cloud Broker and Classification of Cloud Service Broker

Cloud broker could be generally defined as one of the intermediaries between the users and cloud providers to offer a service of integration, customization and aggregation of cloud services in CC. The first known cloud broker was introduced in 2008 with the name CloudSwitch which provides federated services only for Amazon EC2. The cloud in this broker is secure and seamless. Secondly, the cloud broker is introduced in 2011 with the name of the RightScale which offers a cloud management application and platform across multiple clouds in CC. Recently, CloudSim and Cloud Analyst are taking a great place in the CC simulator. The responsibilities of the cloud broker carry the important task first. Due to this, cloud consumers can easily select the cloud service provider to meet their requirement. Paper is categorized into two parts; first is related to complete cloud brokering architecture and second elaborate the simulation environment, CloudSim and Cloud Analyst.

Amrita Jyoti, Rashmi Mishra, Harivans Pratap Singh
Optimizıng Naive Bayes Probability Estimation in Customer Analysis Using Hybrid Variable Selection

Customer study is considered as an important business plan to improve the enterprise's goal. The purpose of customer analysis is to understand the potential customer within the enterprises and their organizational needs and how well the customers are pleased with the company service. To perform better customer analysis, the need for CRM is studied. But the customer data generated are in large dimensional which possibly holds correlated and uncertainties variables in the dataset. To perform better analyzes with these customer data, NB an ML model is applied. But the violation of NB assumption proposed toward variables causes NB to work shoddily. To improve customer analysis using the NB, the variable selection mechanism is proposed. The proposed hybrid mechanism is based upon the filter and the wrapper mechanism. The hybrid mechanism comprises of two phases—first using the ReliefF filter approach, the customer data are processed and ranked attribute subset is generated. Then using threshold value, best attribute set is obtained from the scored attribute subset. Then the preselected variable set is processed using SFS and genetic wrapper approaches individually to get the best optimal variable subset. Further, the variable set acquired using the proposed technique is analyzed with the NB model and performance is computed. The performance hybrid-NB is compared using the filter-NB, wrapper-NB and NB without using any variable selection mechanism. The results present proposed hybrid work better to get the best variable subset and also increase the performance of the NB classifier. Compare to the wrapper approach, the proposed hybrid approach exits less computational time.

R. Siva Subramanian, D. Prabha
An Open Model Approach to Predict Sleep Efficiency Using Nonlinear Regression

Lifestyle diseases are caused by a lack of physical activity, unhealthy eating and sleep deprivation. Chronic sleep deficiency increases the risk of heart disease, high blood pressure, diabetes and obesity. A system was developed using various sensors to monitor various lifestyle factors like steps taken, calories burnt as well as the heart rate. These actigraphy data are used to predict the sleep efficiency of the person by training a degree 4 polynomial regression model with LASSO regularization and fivefold cross-validation, the RMSE is 329, and the score is 0.115. This can be improved further by using deep learning techniques. Sensors used in such devices must be improved to hope for more efficient prediction.

Amitha Deep, Arya Sudarsan, K. M. Keerthana, Ashwin Nambiar, G. Shekar
Adaptability Against Eavesdropping Attack in WSN: Co-Operative Beamforming and Relays Approach

Security is the major concern in wireless sensor networks. To reduce security threats in WSNs physical layer, each node could be designed with an antenna which forms a MIMO communication system. This research work proposed a novel and effective solution to enhance the security of WSN. The sensor node at the transmitter side are used to transmit the beamforming signal to the receiver while few nodes are used to create artificial noise to jam the eavesdropper. Beamforming is used in WSNs for effective communication which enhances the received signal to noise ratio (SNR) at the receiver side. Based on beamforming, smart antennas are modified to switch the direction and adjusts the transmitted power and phase through intermediate relay nodes and constructively it is used in the destination side to combine the signal. This new beamforming technique needs the information about the channel for security concern in these relay system of destination as well as an eavesdropper. For physical layer security beamforming area along with schemes such as amplify and forward relays, decode and forward relays and also co-operative jamming relays where adopted, to confuse the eavesdropper. Two channels are used in the experimental model such as main/legitimate channel, which is used between the relay node and destination node, and the other one is the wiretap channel which is used between the intermediate relay node and the eavesdropper.

Somu P. Parande, Jayashree D. Mallapur
A Survey on Different Scheduling Algorithms in Operating System

Allotting resources for process execution is one of the important procedures to be done in any operating system. Mostly in battery operated systems, where the resources are scarce, it becomes more challenging to have the resource management done, having promised all the processes equal opportunity of resource accesses. There are several algorithms for resource allotment developed and each one of the algorithms thrives to do resource allotment to all the processes dynamically. In this paper, a detailed review of the existing scheduling techniques for resource management has been done. Also, state-of-the-art scheduling mechanisms have been discussed and a comparison has been done among the existing state-of-the-art mechanisms. Finally, the challenges faced by scheduling techniques and how to overcome is also covered.

G. Chitralekha
Contactless Transaction Using Wearable Ring with Biometric Fingerprint Security Feature

Card transaction is one of the most widely used transaction methods in the payment industry. There are more than 50.6 million credit cards and 840.6 million debit cards in India. These cards are so confidential that special care has to be taken while carrying and handling these cards. Carrying these cards everywhere and every time is fewer burdens on the user. Latest near field communication (NFC) enabled smart cards are not asking for personal identification number (PIN) while doing transaction so this is considered as another problem. If such NFC enabled card is stolen then there will be a huge loss. The intention behind this paper is to showcase the idea to solve the above problems in a more secured way. This research problem motivates to come with an appropriate solution with the idea of the wearable ring which has a smart card chip installed in it and has biometric fingerprint scanner support. Anyone can wear this ring so no burden to carry bank cards. It has biometric fingerprint sensor so without user’s authentication transaction will not be carried out, this is the great advantage over existing card payments and rings. This is the proposed idea with properly tested on paper considering parameters like reading range of NFC communication frequency, power source needs to run biometric fingerprint sensor and NFC tag, personalize the MasterCard applet in-ring etc. Future scope of this idea is the actual implementation of hardware that brings the new product in the market.

Amit Magdum, E. Sivaraman, Prasad B. Honnavalli
Ayurvedic Plant Recognition Using Multi-learners Model

Plants are an important source of natural medicine as they synthesize a large number of chemical compounds to sustain their own life, against the attacks of insects, fungus, animals etc. Herbal medicines are traditionally used in many societies worldwide. The project aims for automated plant detection. The datasets used include feature dataset from Kaggle leaf Classification and feature dataset extracted from manually created leaf image dataset of Kerala plants using Histogram of Oriented Gradients(HOG) method. The model was developed after studying the performance of 7 classifiers and choosing the best 3 based on their performance metrics. The majority and weighted voting technique been used for the type prediction of plant leaves. Dimensionality reduction using Principal Component Analysis(PCA) was done without compromising accuracy and a comparative study was performed. Test results illustrate that the multi-learners approach provides better performance than the single learner's approach. Accuracy of multi-learners approximates 97–100% for Kaggle's dataset.

Annie Sonia, K. K. Sherly, Dominic Mathew
Top Threats to Cloud: A Three-Dimensional Model of Cloud Security Assurance

The incredible growth in the cloud applications and services reflects a positive swing in the thought processes of the business decision makers for cloud adoption.Kumar, Rakesh However, ever-evolving security and privacy issues continue to influence the decision makers to delay the cloud adoption.Goyal, Rinkaj In this integrationist exposition, the previous publications are enriched and enhanced to holistically analyze different threats to cloud computing to conceptualize a three-dimensional model of cloud security assurance. These three dimensions, namely Security Solution, Security Operation, and Security Compliance, are interwoven to address the top threats to cloud computing, which are identified and reported by the cloud security alliance (CSA) research group in their latest and previous reports. The model will help practitioners to design and implement a security assurance system for a cloud ecosystem to strengthen trust in the cloud and accelerate its adoption to bring agility and velocity in cloud applications and services delivery in a cost-effective way.

Rakesh Kumar, Rinkaj Goyal
Anomaly Motion Detection and Tracking for Real-Time Security System

Almost every corner of the world is under surveillance. However, crimes (anomalies) can occur anywhere and anytime. Duration of anomalous activities is for a very short time as compared to normal activities. Advancements in computer vision have proved Convolutional Neural Networks (CNN) to be a compelling solution for anomaly detection in videos. This paper proposes to bring together different methods for effective anomaly detection. The dependency of Machine Learning on massive sets of the hand-labeled training dataset is time-consuming. To avoid this, a predictive model is constructed that learns with weak supervision using Multiple Instance Learning (MIL) approaches. The combination of motion information (optical flow) along with spatial information (RGB) obtained from 3-dimensional Convolutional Networks (C3D), using 2-stream architecture, offers a stronger visual representation of videos, thus enhancing anomaly detection performance. TVL1 is used to obtain the optical flow representation of videos. After detection, suspicious behaviors in videos are tracked on the UCF Crime dataset videos.

Pratik V. Kolaskar, Amulya R. Maitre, Prachi R. Khopkar, Shreyas S. Gaikwad, Deepa Abin
Analysis of E-learning Effectiveness in Higher Education

This paper explains the importance and effectiveness of e-learning in higher education. This is based on the student’s usage pattern observation for different e-resources. The study was conducted for a distance learning institution, offering courses in higher education. The observation recorded the patterns for a span of six months. The paper discusses the e-learning framework in broader terms and also discusses the advantages of online learning. It tries to get an insight into the degree of effectiveness of each resource. The paper also tries to explain the role of the learner in e-learning. It throws light on the student behavior in the e-learning framework, i.e., what all resources are there, how they are being used by the students and what are their benefits. It gives a thought about the preliminary requirements for the success of e-learning. This paper justifies the factors responsible for the quality of the e-learning.

Ramya S. Gowda, V. Suma
AHB Protocol Verification Using Reusable UVM Framework and System Verilog

Electronic circuits with reduced size can sustain high-performance operating at higher frequencies. This is possible due to System on Chip design. Multiple functional blocks are implemented in a single chip and the data transmission between multiple blocks might result in complications. Protocols are extensively used to connect IP blocks on Structured System On-Chips and are considered as the backbone for the SoCs. Advanced high-performance bus (AHB) protocol is a third-generation AMBA that is Advanced Microcontroller Bus Architecture specification designed by ARM which supports high-frequency data transmission. The AMBA AHB protocol standard is widely used for on-chip communication. This paper focuses on implementing working verification environments in System Verilog (SV) and in reusable Universal Verification Methodology (UVM) methodology to verify the AHB design specification along with their corresponding functional coverage to judge the quality of the environments. The read/write operations along with 16 beat wrapping and incrementing burst transfers are implemented and verified in verification environments. The functional coverage achieved in System Verilog is improved while working in UVM environment. There is a need for an efficient reusable framework to realize Coverage Driven Verification (CDC) which is provided by the universal verification methodology that results in a reduction of time expense to verify a design system from the very start. The simulation results for the implementations are obtained using QuestaSim 10.4e simulation tool.

Manoj Harshavardhan, Ganapathi Hegde
A Research Insights of Big Data Analytics: Tools and Techniques, Issues and Challenges

Big Data Analytics (BDA) supports organizations to recognize new opportunities by analyzing with their existing data which in turn, transforms the business into smarter business efficiently and effectively, with higher profits. Big Data Analytics plays a vital role in perspectives of (i) Data Science, (ii) Business, (iii) Real-time Usability, and (iv) Job Market. Big Data Analytics inspects large amounts of data; and scrutinizes them to obtain certain significant values from the uncovered hidden patterns, correlations and other perceptions. This paper highlights the significances of Big Data, Big Data Analytics, tools and techniques used Data management in Big Data Analytics. This paper also focuses to kindle research interests among the research scholars to march toward the blooming fields in Big Data.

D. Bhuvaneswari, J. Gnana Jayanthi
Attendance Automation Using Computer Vision and Biometrics-Based Authentication-A Review

Face detection and recognition methods have enhanced a lot in the last decade in terms of accuracy, speed, and overall performance. This development has completely changed many systems. Attendance management is one example of this change. Earlier, the pen-paper-based method was used for marking and storing attendance. The improvement in biometric detection and recognition methods has resulted in the possibility to build a complete automated attendance management system. Face recognition based methods for attendance management has many advantages over the traditional attendance methods. The main challenge with the face recognition based method for attendance is to ensure that it recognizes every detected face. In this paper, previous works of several authors on face recognition based attendance management systems have been discussed, and a comparative study has been performed between the works. This paper shows that the need for cost-effective and network independent face recognition based attendance system having high accuracy still exists.

Suraj Raj, Saikat Basu
Hybridization of TrellisNet with CNN

This paper is an introduction to how hybridization is performed on a sequential model namely TrellisNet that affects its accuracy.Jaiswal, Akshat TrellisNet is a temporal convolution Network with a special architecture.Duvvada, Prashanth An attempt is made to combine this network with the Convolution Layer to improve Trellis Net performance.Nair, Lekha S. For this task, here EMNIST, which is an extension of MNIST that contains English handwritten letters and digits is used. The goal is to demonstrate the efficiency and performance of the current combined model. This also gives us an insight of performance this model as compared to other popular models used for Handwritten character recognition. TrellisNet with generalized weight matrices is a truncated Recurrent Network. Thus, a Convolution Layer followed by the defined TrellisNet is designed for achieving the targeted results aimed in this paper. This stacked architecture has delivered encouraging results. The results establish the characteristics of the model and show the problems this model can address.

Akshat Jaiswal, Prashanth Duvvada, Lekha S. Nair
Vehicle Identification Using Automotive LIDAR VLP-16 and Deep Learning

In this research, a deep learning-based technique is presented for vehicle identification from the 3D point cloud data obtained using the automotive-gradeVelodyne LIDAR VLP_16. Last few years LIDAR based systems have been receiving a lot of attention in the automotive industry because of LIDAR’s properties like independence from brightness conditions and environmental conditions like night, fog situations. The way Normal LIDAR works, nearer objects are dense in the point cloud and far away objects are not dense. Because of this property, it makes the vehicle detections on the road highly difficult. From the surveyed literature, it is seen that most of the LIDAR-based frameworks are designed to work with a multi-LIDAR setup for vehicle detection, making this process very costly and computationally very intensive. Hence, a low-cost single LIDAR vehicle identification framework is proposed for Indian road conditions. Vehicle identification is performed using a Convolutional Neural Network-based deep learning model(YOLO-V3). The developed framework was successfully validated by deploying it on an actual road within Amrita - Coimbatore campus.

M. P. Keertheeswaran, Binoy B. Nair
Improved Scheduling Algorithm for Load Balancing in Cloud Computing

Distributed computing becomes a substitution to the global pattern of registering. It's a cutting-edge kind of exploitation of the intensity of web and wide space organize (WAN) to give assets remotely. It's a substitution goal and technique to see high accessibility, adaptability, cost diminished, and on request calculable. In any case, distributed computing has a few difficulties like poor asset use that includes profound effects inside the exhibition of distributed computing. These issues emerged in light of the enormous measures of information. Along these lines, the need for efficient and ground-breaking distributed computing load-leveling calculations is one of the premier fundamental issues during this space to fortify the presence of distributed computing. A few specialists arranged shifted load levels and employment arranging calculations in distributed computing, anyway, there’s still some powerlessness inside the framework execution freight lopsidedness. Along these lines, during this examination, in general, propose a heap leveling equation to fortify the exhibition and intensity in heterogeneous distributed computing setting. Also, propose a crossbreed equation bolstered association and ravenous recipe; it takes endowments of each arbitrary and avaricious calculations. The equation thinks about this asset information and the centralized server capacity issue to comprehend the targets.

Midde Ranjit Reddy, D. Raghava Raju, T. Venkata Naga Jayudu
Smart Twitter Analysis on Location Using Kali Linux

Twitter is a rich source for mining views and opinions that are often connected with habits or show a pattern.Sikri, Nishant Social media platforms add geo-location as a parameter in tweets or posts, making it accessible for marketing tactics and focus on a user base. A similar scenario is when a local business owner aims to open a new store or a cafe in a neighborhood. To address this problem, location-based tweets help to keep track of post timings, interests, number of active users, and opinions. The extraction of analyzed data can help achieve targeted advertisements, open restaurant chains, skin clinics, and set up shops. This results in overcoming the trial and error method that utilizes a lot of the investment and often goes unnoticed by the audience.

Nishant Sikri
Video Steganography: Storing Data in the Transition of Frames

In today’s fast-paced world, security has attained an extremely important place in the technological sphere.Paranjape, Tejas In the last couple of decades, a huge boom was seen in the number of techniques of steganography that were developed.Paradhi, Vineet Steganography is the science of concealing data in an image, an audio, or a video file, such that the data is not at all discernible to the human senses.Shah, Kewal This paper intends to delve into the relatively new technique of video steganography.Hole, Varsha New algorithms are explored to hide information not in the individual frames of the video, but in the transition of the frames. The least significant bit technique is used in our algorithm. This method will provide a more secure method as compared to the existing techniques, as it is a novel approach to hiding data in plain sight.

Tejas Paranjape, Vineet Paradhi, Kewal Shah, Varsha Hole
Evaluation of Execution Time by Various Multicore Processors on the Parallel Sorting Algorithm

As the technology grows, the multicore processors set the new benchmark for this era of technology by achieving a standard performance of completing a task. The performance of the processor can be achieved by multiple numbers of cores on a single die, on-chip, and off-chip cache size, power utilization, and many more. But this is necessary to ensure that as the load is increased on a particular number of the core processor the time taken by the processor to complete the task is also increased or not. This paper contributes to ensuring that as the load is increased the execution time is also increased.

Pranjal Joshi, Pankaj Acharya
Automation and Integration of SSI Test Cases for Abis and A-Interface in GSM Using Robot Framework

Mobile wireless communication has undergone vast advancements in the last few decades due to the growth of subscribers. There is a rapid growth in the number of features required by the users and the companies need to deliver products in a very short time. As the demands of users increasing the time required to release a product into the market also increases. Global System for Mobile Communication (GSM) is a second-generation mobile cellular system in cellular networks. Exchange Terminal for Packet Transport (ETP) is a plug-in-unit placed between Base Transceiver System (BTS)—Base Station Controller (BSC) and BSC—Media Gateway (MGW) in GSM to improve the bandwidth and is an important component to handle GSM and GPRS (General Packet Radio Services) calls. It is used in BSC to support the packet-based transmission of signaling and payload between the BTS and the BSC for Abis and A over IP (AoIP) interfaces. In this paper, to carry out automation testing, an automation tool called Robot Framework is used that helps to test ETP functionality. This framework allows execution of Sub-System Integration (SSI) Test cases for Continuous Integration and executes the test cases developed for ETP feature at one stretch without manual intervention.

A. T. Sudhan, G. Paramesh, G. Ranjani
A Study on Human Face Recognition Techniques

The facial recognition system is a technology that is efficient in recognizing or proving a human by using the image or video. It is a way of identifying human by using technology. The facial recognition system uses biometrics to capture the facial features from the photograph or video. It compares the information from the known database to find a match. There are many facial recognition techniques used nowadays. Among the biometrics, facial recognition plays a major role. The facial recognition uses machine learning to discover, match and recognize the person, and it is widely used in a variety of ways. The facial recognition system is used to differentiate among the users, and it produces an accurate result. In this work, facial recognition and its techniques will be briefly described.

S. Ranjani
Ensemble of Multiple Classifiers for Accelerometer Based Human Fall Detection

Fall is a major health threat for the elderly. Automatic detection of human fall using machine learning algorithms is necessary to reduce the consequence of human fall. In the proposed method, nine different features resultant, variance, standard deviation, root mean square, Euclidean norm, skewness, kurtosis, and geometric mean have been calculated from accelerometer data. Stacking based ensemble of Naive Bayes, KNN, J48, and Random forest classifier has been used to discriminate fall activities for ADL. The performance of each classifier has also been calculated for comparing the performance of each classifier with a stacking-based ensemble method. It was found that even some individual classifier gives better result in sensitivity or specificity but stacking based ensemble of these classifier gives better results in a combination of sensitivity and specificity. With stacking-based ensemble classifiers, 89% sensitivity and 95% specificity have been achieved which is higher than the performance of individual classifiers.

Rashmi Shrivastava, Manju Pandey
A Data Privacy Approach Using Shamir’s Secret Scheme in Permissioned Blockchain

Information sharing is the basic necessity for advancement in research works.Sunil Kumar, K. Information sharing in the existing technology is relied on trusted third parties (TTP).Jain, Kurunandan There is a lack of trust for transparency, security, and immutability in these systems because of the participation of TTP.Subramanian, Narayanan For these issues, the blockchain technology can be used to secure the information sharing in the decentralized network by maximizing the advantages of the decentralized file system (DFS). Blockchain faces serious issues with data privacy, which is an obstacle for the application of the blockchain. This paper introduces the usage of the Shamir’s secret sharing (SSS) system to the blockchain. The proposal resists passive and disguise attacks. The proposed systems with SSS have results with lesser computational time compared to the advanced encryption standard (AES) 128. In non-transactional scenarios, the proposed system is functional and effective.

K. Sunil Kumar, Kurunandan Jain, Narayanan Subramanian
Impact of Cyberattacks on Electronic Patient Health/Medical Records

Digital health offers several benefits from different perspectives. Therefore, the transition from paper to meaningful cost-effective electronic health records (E-HR) is the focus of several countries. In the literature, E-HR functionalities are discussed extensively. It may be noted that security measures to curb potential cyberthreats in E-HR need further attention. The working of health monitoring needs a transformation from a reactive mode to a proactive and predictive mode using contemporary technologies. While implementing the E-HR security/privacy perspective is essential since the safety of the medical records is key to the safe life of the patient. This paper highlights security/privacy issues, cyberthreats to E-HR, the pros, and cons of contemporary technologies such as IoT networks and discusses meaningful E-HR so that patients and society significantly benefit as believed by experts and policymakers.

B. R. Arun Kumar
An Overview of Blockchain-Based Smart Contract

This work is motivated by the recent flare-up in the rapid technological developments of the evolving blockchain technology. Blockchain technology provides a mechanism to do transactional communication in a trustless manner, without the need of a trusted third party. An appealing feature or application of blockchain technology is a smart contract. The smart contract is a piece of code which executes on the blockchain technology to enforce the negotiation between two parties in the absence of a trusted third party. This paper aims to provide an ample overview of blockchain-based smart contract, starting with the introduction, architectural overview, working mechanism, application areas and research gaps which can be addressed in future research.

Satpal Singh Kushwaha, Sandeep Joshi
Shape and Texture Features Extraction Using Segmented Histopathological Images

For women, breast cancer occupies the second position in causing the occurrence as well as mortality. Optimum segmentation and feature extraction play a crucial role while categorizing medical images. The proposed paper integrates marker-based watershed approach with K-means clustering data for optimum segmentation. It deals with detail component protection. The work focus on feature extraction from the segmented histopathological images. Feature selection is necessary for minimizing the redundant parameters. Optimum features necessary for image categorization were evaluated. The proposed work provides high accuracy during image classification.

U. Rajyalakshmi, K. Satya Prasad, S. Koteswara Rao
Automated Continuous Integration (ACI) Scheme Based on Jenkins

Continuous integration (CI) essentially manages the automation of improvement procedures or development processes and build/code integration/test automation. CI turning out to be such a very much development practice has likewise demonstrated the amount progressively profitable the advancement procedure can be with automation. It has become the catalyst for continuous delivery (CD) practices—expanding the automation through the whole software pipeline, through organizing and production. The most ideal way to make CI quicker and progressively productive is to automate the build and testing process. Jenkins is a CI tool that helps in automating the complete process, decreasing the work of a developer and check the advancement at each progression of software development. As an extensible automation server, Jenkins is utilized as a basic server or transformed into the continuous delivery center for any undertaking project.

Syeda Gazala Rizvi, G. S. Mamatha
A Hybrid Learning System for Telecom Churn Prediction Using Ensemble Learning

The information-based prediction models have gained huge quality throughout the past few decades with machine learning techniques. Combining multiple classifiers to create hybrid classifiers is a very big challenge. Designing the best ensemble learning method to perform prediction is imprecise in this sector. Identification of attribute which likely impacts the churn is missing in many existing works. Because of the vast monetary value of the client churn in the telecom industry, the businesses have analyzed numerous factors from the whole world (like decision value, decision quality, client service reaction time, etc.) with the help of many machine learning techniques. Telecom churn prediction using ensemble learning uses various ML algorithms. The comparison of the best algorithm is done to predict the churn which is derived with more accuracy than the existing system.

G. Sandhya, S. Samarpana, R. Sangeetha Vani
IoT: A Novel Method for White Coat Effect (WCE) Detection from Cloud for Improving Patient’s Treatments

White coat effect (WCE) is a major issue in medical research due to the variation of multiple occasions of blood pressure (BP). BP is the most dynamic factor in a clinical problem which varies from time, place, season and movements of the body functions. BP reading is an important basic feature and initial step prediction of hypertension, diabetes and obesity. The World Health Organization (WHO) estimate of hypertension becomes the most significant premature death in the global world and 2025 hypertension increased nearly 1.56 billion people. The treatment of hypertension is mainly based on occasion level of BP. Medical diagnosis is a difficult outcome result of unstable measures from formal methods. Many researchers find the BP variations and provide improvement suggestions for treating a patient from clinical health data and home-based reading measurements. However, existing methods have no reliability, efficiency and accuracy in treatments. In this paper, a narrative technique is proposed as cloud computing reference model for justification of patient treatments which follows daily update in routine manner order to evade white coat syndrome (WCS). Additionally, the heartbeat rate, glucose monitor, pedometer measurement performance of patient health data from a cloud are examined and also compared the results of various scenarios. Moreover, our experimental analysis in Hadoop sandbox reveals 85% accuracy and safety clinical prediction in healthcare treatments. Thus, patient can prevent from an unnecessary excess of treatments and avoid tablets intake which leads to the cause of side effects or organ damage.

E. S. Madhan, K. Padmanaban
Adaptive Sub-Optimal Bit Flipping Decoding Algorithm for LDPC

This paper proposes a new iterative algorithm for low-density parity check (LDPC) codes. The flipping function is altered much more effectively to increase the performance of the decoding process. Weights are derived based on the reliability of the bits. The algorithm is validated through simulation results that compare various decoding algorithms. A significant gain (>0.3 dB) is observed over the standard weighted bit flipping (WBF) algorithm while the increase in complexity of computation is not so noticeable. LDPC codes are increasingly used in applications that require efficient and highly reliable transfer of information over limited bandwidth or return channel-constrained links of noise-affected channels.

S. Ashwin Kumar, R. Aravinthan, Ch. Sriram Reddy, A. Aswant Ramana, K. Pargunarajan
Analysis and Design of Smart Office Automation System

The technology of automation is playing a very vital role in human’s life. It makes the work trouble-free and effortless. This paper deals with the conception and practice of an automated office system using node MCU along with a mobile application. This system is based on different but crucial functionalities such as raising alarms if noise levels pass the required thresholds, checking the quality of air, watching the heat and wetness of the atmosphere at work, and detection of really bright sources of light like fire which may hinder the progress. To observe the workplace remotely, an android mobile application is developed with the help of sensors that provide real-time information over wireless networks. This application helps in preventing the hazardous event to happen, thereby creating a safer environment for the workforce.

Sahil Gupta, Swati Gupta
Driver Drowsiness Detection

Driver’s drowsiness is considered as a major reason behind accidents on road, around the globe. Driving nonstop for a long period of time will cause accidents. The consequences of drowsy state are the same as alcohol, and it will create a driver’s driving inputs poorer, destroy the driver’s reaction times, and blur driver’s thought processes. To prevent such disastrous situations, a real-time driver monitoring system is implemented using OpenCV, where the aspect ratios of extracted contour features of eye and mouth are measured, and an alarm is generated. With EAR 0.25 and MAR 0.75, the results show that the alarm is generated for the blinks. The robustness of the implementation has been verified by changing the EAR and MAR, values, and best results are given for the EAR 0.25 and MAR 0.75.

V. T. Sai Sandeep Raju, Meena Belwal
Detection and Classification of Distributed DoS Attacks Using Machine Learning

Distributed denial of service (DDoS) attacks target the websites and online services in which the attacker floods them with more traffic than the server or network that can hold. With the sophistication of technology and the rise of the Internet, such attacks are becoming more common and easier to perform. According to a study by the Cisco Visual Networking Index (VNI) in 2017, global projections of the overall number of DDoS attacks would significantly increase by 2022 and might cross 14.5 million marks, thereby doubling in number. A system is proposed that aims to provide an efficient way for the detection of DDoS attacks in a network. As previously, machine learning has been widely used for intrusion detection and classification of the type of attack compared to other techniques and intrusion detection system (IDS). This system makes use of different machine learning algorithms (extreme gradient boosting, K-nearest neighbour, stochastic gradient descent, and Naive Bayes) and a deep learning architecture (convoluted neural network) to identify attacks and classify them. The result shows that XGBoost achieves the highest accuracy, while CNN and KNN also give comparable figures. Our code is available at https://github.com/mohak1/Detection-and-Classification-of-Distributed-DoS-Attacks-using-Machine-Learning .

G. Usha, Mohak Narang, Akash Kumar
Host Platform Security and Mobile Agent Classification: A Systematic Study

Mobile agents are a successful use-worthy model for applications that are distributed and for partially associated computations. Mobile agents are assuming critical jobs in the research and business world as they can go starting with one hub then onto the next with their current state and convey and gather data to or from different hosts. Albeit mobile agents have a lot of focal points in different areas of application, and there are numerous safety issues. Mobile agents can transport with them obscure malign codes causing harm to other agents or leading to mobile agent host platform’s disruption. So, some security issues have been discussed and judge the existing solution to. This paper talks about different methods to secure the mobile agent and the agent platform, along with certain advantages and disadvantages of the examined innovations and takes cognizance on brief examination among classifying procedures to discover malign mobile agents.

Ayushi Acharya, Hari Prasad, Vinod Kumar, Ishu Gupta, Ashutosh Kumar Singh
Narrow Band Spectrum Sensing of Cognitive Radio for Wireless Services

For longer communication, wireless operations are used nowadays. Wireless communications lead the life of every human by offering several varieties of services. Hence, the growth of wireless networks is increasing day-to-day to meet the applications. But today’s spectrum policy is an insufficient usage of the radio spectrum with the advancement of technology. Cognitive radio (CR) came into existence as an effective solution for spectrum congestion. Of all the operations of CR, spectrum sensing plays a vital role. In the different narrowband spectrum sensing techniques, the energy detection method is the basic approach to sense the spectrum, because of its low complexity and speed. Another technique called matched filtering method gives high detection probability in short detection time. This CR technology provides certain applications such as appropriate monitoring for patients in rural areas, along with timely hospital visits which save lives and hospital services automation, emergency alerts, etc. The simulation results proved the performance of matched filtering even at low SNRs of −6 dB.

T. V. N. L. Aswini, K. Padmaraju, B. Leela Kumari
Optimal Phasor Measurement Units Placement in Radial Distribution Networks Using Integer Linear Programming

The traditional passive distribution networks are evolving into active distribution networks with the integration of dispersed generation (DG) to distribution networks. The conventional network monitoring systems do not monitor and provide the information accurately at a faster rate due to the intermittency nature of sources such as wind and solar. Hence, a real-time accurate and faster monitoring equipment like phasor measurement unit (PMU) is needed. The focal objective of this work is to bring the power distribution network more closely aligned with the smart grid communication technology for better system monitoring conditions. This paper presents the deployment of PMU optimal allocation in the radial distribution network by adopting an integer linear programming (ILP) technique for the system’s full observability at a standard operating condition. Standard radial test feeders such as 12-bus, 15-bus, 28-bus, IEEE 33-bus, IEEE 69-bus, and 119-bus are chosen to study the effect of PMU placement problem. MATLAB programming is used as a simulation software to check the observability of the above test systems.

Swathi Tangi, D. N. Gaonkar
Investigation of the Static and Dynamic Path Planning of Mobile and Aerial Robots

Autonomous mobile robots are getting more important because they can make decisions on their own in the given environment. The robot path planning is an important function in the autonomous robot. In this review, several path planning techniques are investigated to analyze the performance. The path planning techniques are categorized as static and dynamic. A Q-learning method with flower pollination algorithm and the genetic algorithm has an effective performance on the path planning of mobile robot (MR) in a dynamic environment. The ant colony optimization (ACO) method shows considerable performance in the static environment. The aerial vehicle path planning method also reviewed and the method based on geometric random tree shows the effective performance in path planning. The comparison table shows the latest research in the MR path planning with the advantages, limitations, and performance analysis.

A. Chandrashekhar, Shaik Himam Saheb, M. L. Pavan Kishore
DDoS Defense Mechanisms for SDN Control Plane

Software-defined networking (SDN) is the revolutionary technology that provides flexibility, vendor neutrality, and centralized management by separating the network operating system from the forwarding hardware.Kaur, Sukhveer However, SDN is susceptible to various security attacks due to its centralized and decoupled architecture.Kumar, Krishan This paper discusses the SDN architecture from two aspects.Aggarwal, Naveen On the one hand, SDN framework is used to secure conventional networks. On the other hand, SDN itself becomes a victim of a denial of service attack. DDoS attack becomes a significant threat to SDN architecture because it can easily overload the controller and flood the flow table of OpenFlow switch by sending an excessive amount of traffic to OpenFlow switch that results in network performance degradation. This paper presents the various DDoS attacks in the SDN framework. Finally, detailed analysis of multiple DDoS defense mechanisms to protect the control plane is performed.

Sukhveer Kaur, Krishan Kumar, Naveen Aggarwal
Wearable On-Body Antenna for WBAN

This paper focuses on a bandage design wearable an on-body adhesive antenna designed to operate at a frequency of 2.4 GHz and 3.1 GHz defined wide body area network (WBAN) frequency. The material used in designing the microstrip antenna is Rogers RT Duroid 5880 having a permittivity of 2.2 and loss tangent of 0.0009 with a height of 1.5 mm and it incorporates a probe feed. The antenna has a return loss of −27.4665 at 2.4 GHz and -27.0123 at 3.1 GHz and SAR of 0.752 W/Kg. The antenna is designed using software tool high frequency structural simulator (HFSS) (Version 15).

Anushka N. Mohanty, Divyanshi Nath, Jessica Sadavarte, Tazeen Shaikh
A Study of Android Malware Detection Using Static Analysis

Smartphones have experienced blazing popularity and become more and more sophisticated over recent years; as a result, they have become an appealing target for malware authors. Malware keeps evolving and becomes more dangerous and harmful each day. Therefore, it is the need of the hour to detect and stop the spread of this previously unknown malware. It is also vital to detect and classify this malware, and machine learning has proven to be helpful in this field. For machine learning algorithms to achieve better performance, it is necessary to collect essential features from the application by reverse-engineering the APK file. However, this gives malware authors the upper hand as they started developing special anti-analysis techniques to mislead the machine learning-based analysis by obfuscation techniques to hide the application’s malicious behaviour. This study summarizes the static malware detection approach using different machine learning techniques, with emphasis on present-day research, challenges, and future directions.

Kapil Sharma, Anish Singh, Prateek Arora
Cloud-Based Real-Time Wearable Health Monitoring Device Using IoT

Internet of things (IoT) is the collection of smart objects that collect and exchange data. Using communication technologies, IoT further transmits the data for storage and processing in local databases or unlimited cloud infrastructure. IoT offers a wide spectrum of applications such as home automation, smart farming, smart retail and business applications and health care. This paper proposes cloud-based real-time monitoring of health information using wearable devices. It monitors ECG, pulse rate, temperature and skin sweat rate using sensors and stores the data in the IBM cloud. This enables the doctors to view the data and diagnose any abnormal condition of the individual. The wearable device has been tested with 10 different patients.

Getzi Jeba Leelipushpam Paulraj, Immanuel JohnRaja Jebadurai, Jebaveerasingh Jebadurai, Nancy Emymal Samuel
Performance Investigation of 4 × 20 Gbit/S-40 GHz MDM-OFDM-FSO System Under Weather Conditions

This work emphasizes on the transmission of four independent signals, each having 20 Gbit/s–40 GHz, by means of mode division multiplexing with different Hermite–Gaussian (HG) modes. The system has been incorporated using OFDM-based RoFSO link. The analysis of the system is being executed by making an allowance for the effect of changing weather conditions like light, medium and heavy rain. Square root module technique has been applied at the receiver side, and system performance has been observed by this improved detection technique. The results reveal a flourishing broadcast of 80 Gbit/s–160 GHz data over a distance of 120 km with clear weather circumstances using the planned link.

Amit Grover, Anu Sheetal
ICT Capabilities Among Micro, Small and Medium Enterprises in Relation to Demographic Characteristics with Special Reference to Kochi–Kerala

The growth of Information and Communication Technology plays a pivotal position for the development of Micro, Small and Medium Enterprises in India. Changes are imperative for the development of MSMEs. The entrepreneurs should keep a constant eye on the business environment as Information and Communication technology helps in emanating such changes that are vital for improving performance for Micro, Small and Medium Enterprises. Information and Communication technology through efficient manufacturing practices, improved supply chain management, improved human resource practices, developed customer relations etc. enhance a firm’s performance. The present study focuses on 90 entrepreneurs to know whether MSMEs are innovative enough to employ Information and Communication Technology, which improves productivity and growth of the enterprises as a whole. The study pertains to whether the entrepreneurs have developed ICT capabilities with respect to their demographic characteristics, which enhances the firm’s growth and development.

Vidhya Vinayachandran, A. S. Ambily
Design and Analysis of Rectangular Microstrip Patch Antenna for 2.4 GHz Wireless Communication Applications Using CST Microwave Studio

In this paper, the operating frequency or resonance frequency of 2.4 GHz is selected, which is best for industrial, scientific and medical (ISM) band applications. CST MWS is used as the software environment to design and simulate the performance of the antennas. The dimensions of a rectangular microstrip patch antenna have 38 mm × 29 mm × 0.13 mm. This antenna is designed using lossy FR-4 substrate material used as a dielectric with reduced size of 76 mm × 58 mm. Its dielectric constant is 4.4 and 1.6 mm thickness. This paper presents CST Microwave Studio. Here, VSWR plot, 2D Radiation pattern plot, S-parameter magnitude, bandwidth plot, directivity polar plot, E-Field, H-Field, Surface current and gain of the antenna are estimated using CST microwave studio software.

V. Prakasam, P. Sandeep, K. R. Anudeep LaxmiKanth
Internet of Things based Smart Ring System for Providing Human Safety Services during Emergency Situations

Internet of Things has a set of real-world objects that allow for connecting and monitoring the real-world objects or things through the Internet. Ensuring personal safety and detecting physical assaults are critically important issues in the real-world environment. During an emergency, the person has no time to call for help through his/her mobile phones. In this work, the proposed system provides the real-time emergency services to human beings in various emergencies like road accidents, kidnap cases, health issues, attention toward the aged individuals, person in trouble, women harassment, and explains new or unknown places in the real-world environment. It is an integration of two sub-systems like wearable finger ring device system and emergency service systems. Here each service system will communicate with one another for exchanging the information regularly. Finally, the proposed system is implemented in terms of efficiency.

Ramesh M. Kagalkar, Lokesh B. Bhajantri
Reliable Mechanism to Detect Traditional Cyber Attack Using Artificial Neural Networks

DDoS has evolved as the most common and devastating attack that has been confronted from previous years. As many networks reply simultaneously, mostly RREP will work together to accomplish a DDoS attack. Thus, no information system can tolerate and survive once they confront this ruthless attack. There are many existing intrusion detection systems to prevent and protect the system as well as network from DDoS, but still DDoS is complex to perform detection and perplexing. In this research article, an IDS has been developed based on the basics of latency and delays in neural networks. To form a multi-layer architecture, every node is kept on surveillance once the detectors are deployed in the network topology, and the activities of every single node are tracked by their close hop nodes mutually to ensure their status of survival. Only after all of the information is collected in a table, it is forwarded for integrated analysis by their selected expert module. The nodes covered in the first and second layer of firewall experience some suspected packets or streams as that of DDoS pattern and the core expert module that started right after the second firewall will take some effective action and invoke the defense module to ensure the safety of the information system. And the nodes which did not stand against defense module will be isolated first and rebooted later to ensure the normal functionality of the network.

Tariq Ahamed Ahanger, Abdullah Aljumah
IoT-Based Water Quality Monitoring System Using LoRaWAN

An efficient and practical system is a necessity for monitoring the quality of water to solve the issues faced during the utilization of water from reservoir systems. Conservative methods used for the transmission of data from each system have proven to be a tedious procedure, and the vastness of the reservoirs hinders the integrated transmission for the simultaneous comparative assessment. The introduction of an IoT-based sensor with LoRa spectrum modulation transmission technology allows the real-time integrated data collection and transmission from different systems spread across a long-range reservoir. The water quality monitoring system is embedded with a pH, turbidity, temperature, and conductivity sensors, connected to an Arduino UNO board and converted to an IoT device with a wireless data transmission module to send the data to the cloud storage facility. The transmission of data from an embedded system using a LoRa module provides a low-power and low-cost technology over a long range compared to the conventional short-ranged Wi-Fi transmission technology, which inhibits the real-time assessment and estimation of the data simultaneously from different systems.

Liloja, M. Sreelekha, Ganesh Gopakumar, K. Shahil
Enhanced Image Security Using Stenography and Cryptography

In this age of information technology, the most vital part of information and communication exchange is the Internet. With advances in information technology and the Internet, digital media has become one of the best-known tools for data transfer, but this is still facing many challenges, including issues of authentication, modification, ad copyright protection. Many techniques such as encryption, authentication, and steganography can be used to protect these digital data. To secure information, encryption and information hiding are used widely. In the proposed work, a combination of encryption and information hiding is used to provide additional security, which means that it provides two-layer protection. The security level of the hidden image is measured in the term of PSNR and MSE parameters. In stenography, low MSE value and high PSNR value are better. The proposing work improves the result of previous work in the term of PSNR and MSE. To make the image security more robust and more secure, also add the cryptographic process with steganography.

Sushmita Matted, Gori Shankar, Bharat Bhushan Jain
A Novel Approach for AES Encryption–Decryption Using AngularJS

The goal of this paper is to implement a simple encryption–decryption web application that utilizes the facility provided by AngularJS. The advanced encryption standard (AES) algorithm is implemented using the methods from the CryptoJS library. This algorithm was chosen, as it is a U.S. Federal Information Processing Standard, which was selected after a thorough process, trumping 15 competing solutions. The advantages associated with using AngularJS include the fact that it is quick to learn a language. Furthermore, two-way data-binding capability means that developer intervention is not required for data binding. Developing powerful Web applications is made easier since AngularJS architecture separates data from design. Additionally, the developers must be familiar with the model-view-controller architecture to work efficiently with AngularJS. Plaintext or instances of CryptoJS.lib.word array library may be accepted as inputs. When a string is passed for a certain key, it is considered as the passphrase, which is subsequently used to derive the key. When it comes to the CipherText, strings, or CryptoJS.lib.CipherParams, instances are accepted. All strings passed are eventually converted to a CipherParams object. Coming to the cipher output, the plain text derived after the decryption process is a word array object. The CipherText returned is a CipherParams object, which gives access to the parameters used during encryption. Ample flexibility is provided, allowing various formats to be used. Essentially, a format is nothing but two-method objects—stringify and parse. This help converts CipherParams objects and CipherText strings. The practical implementation of the method involves downloading and installing NodeJS, AngularCLI, and Visual Studio code, the detailed explanation of which is elaborated under the proposed method section. AES algorithm performs well irrespective of the size of data to be encrypted as it has a big-O time complexity which is constant—O(1).

B. N. Arunakumari, Aman Rai
AndRev: Reverse Engineering Tool to Extract Permissions of Android Mobile Apps for Analysis

One of the leading and the most popular operating system for smartphones and tablets is an Android. Being an open-source platform has also become a prime target for the attackers as growing users. This paper focuses on the work done on the Android platform by performing static analysis on the permission-based framework and permission extraction tool—AndRev, which is designed. Extracted many permission-based features by reverse engineering of the Android application (apk) files using the batch-scripted tool. AndRev tool is used to decompile apks in batch mode. Features have been stored in feature vectors. Firstly, analysis is done using feature vectors to study the pattern of permissions in applications as per the category. Two categories of apks, namely general and entertainment apps, are studied with an initial dataset of 50 applications each. Secondly, do an experimental study of applications permission removal by using a reverse engineering method. Updated apks are recompiled apps, which execute on a mobile phone as the way it executes like the original app. The study consists of ten apps from Google Play with various categories. The study concludes that it is not easy to remove permission as per the type of permission and the relation of apps permission with app’s relevant functionality. Finally, performed security analysis on the vulnerabilities within the source code, and those are used for accessing resources or unauthorized permission authorization of Android apk. For the study, many vulnerabilities based features were extracted by vulnerability assessment tool Quixxi for the Android application (apk) files. The study depicts that medium-risk vulnerabilities are higher than high- and low-risk vulnerabilities. In security analysis point of view, observations concluded would be useful to future Android app developers

Manisha Patil, Dhanya Pramod
Analysis of YouTube Communities for Indian Political News

The term community defines and explains the fact that there exists a group where the members of that group happen to have more commonality amongst each other in terms of features, comparing to the feature of that two members of different groups which would have fewer things in common. In a world that is ruled by online social networking, it is important and difficult to understand that based on what opinions or environment leads to the formation of a certain community amongst the masses. These views tends to become the foundation of society and hence country or even the world. In this paper, we attempt to understand this by experimenting on a popular social network for video sharing, i.e., YouTube. Our work consists of community detection on a real-time dataset of related videos about Indian political news. We try to get the best possible structure by choosing a suitable algorithm depending upon the scenario of our platform. As a result, the post-detection of communities, we extracted as many features possible within each group. The final step is content analysis to obtain the knowledge that we infer from each community.

Aishwarya Gambhir, Mohona Ghosh
Perlustration on Mobile Forensics Tools

Nowadays many people store more information in cell phones rather than do on their computer which leads to the increase of crimes taking place in mobile. This is also because of the extensive use of mobile devices. People refer to store more information in mobile than on a computer. Therefore, the scope of forensics in mobile has increased in the last few years as compared to computers. In this paper, a survey on tools and techniques which are in practice for mobile forensics has been carried out. The survey is the amalgamation of the in practice technical methods and discussion of some of the tools for specific operating systems. Some challenges faced by practitioners while performing forensics have also been enlisted.

Utkarsha Shukla, Bishwas Mandal, K. V. D. Kiran
Backmatter
Metadata
Title
Computer Networks and Inventive Communication Technologies
Editors
S. Smys
Ram Palanisamy
Álvaro Rocha
Grigorios N. Beligiannis
Copyright Year
2021
Publisher
Springer Nature Singapore
Electronic ISBN
978-981-15-9647-6
Print ISBN
978-981-15-9646-9
DOI
https://doi.org/10.1007/978-981-15-9647-6