Skip to main content
main-content
Top

2023 | Book

Proceedings of Third International Conference on Computing, Communications, and Cyber-Security

IC4S 2021

Editors: Dr. Pradeep Kumar Singh, Dr. Sławomir T. Wierzchoń, Dr. Sudeep Tanwar, Prof. Dr. Joel J. P. C. Rodrigues, Dr. Maria Ganzha

Publisher: Springer Nature Singapore

Book Series: Lecture Notes in Networks and Systems

share
SHARE
insite
SEARCH

About this book

This book features selected research papers presented at the Third International Conference on Computing, Communications, and Cyber-Security (IC4S 2021), organized in Krishna Engineering College (KEC), Ghaziabad, India, along with Academic Associates; Southern Federal University, Russia; IAC Educational, India; and ITS Mohan Nagar, Ghaziabad, India, during October 30–31, 2021. It includes innovative work from researchers, leading innovators, and professionals in the area of communication and network technologies, advanced computing technologies, data analytics and intelligent learning, the latest electrical and electronics trends, and security and privacy issues.

Table of Contents

Frontmatter

Communication and Network Technologies

Frontmatter
Enhancement of Energy Efficiency in Wireless Sensor Network with Mobile Sink: A Survey

The energy consumed by any activity taking place in WSN should be controlled such that limited energy in terms of battery backup remains focus throughout. In the case of dying nodes, battery discharge may cause the network to get disconnected. WSN design issues, e.g., location of sensor nodes, scheduling activities, routes of data flow, mobile sink route, should be dealt with keeping energy limitation in mind. The sensor nodes sense the data from the area of concern and communicate the same to the sink for processing. Sensor nodes deployed in various application areas have limited memory, computational power, and battery backup. There is no defined topology of such network and frequently changing environment, very less amount of battery, and limited storage capability of the nodes. It is essential that each node in the network has knowledge about the routing path to the sink which is energy efficient. Since random placement of the nodes restrains coders from presuming routing table data at the sensor nodes, numerous methods have been suggested to create a dynamic path up to sink. Numerous researches are performed for WSN using the mobile sink. Most of the research activities focused on energy conservation in the background while proposing approaches for clustering, data flow paths, trajectory design, etc. In the WSN with a mobile sink, the trajectory of the sink node plays a vital role. Designing of trajectory is an NP-hard problem. With the use of nature-inspired techniques, e.g., particle swarm optimization (PSO), genetic algorithm (GA), etc., can be used for generating a nearly optimal paths for the mobile sink. In this current article, the authors make attempt to present the summary of various strategies for energy-efficient data collection methodology and energy-efficient path planning of mobile sink in wireless sensor networks.

Akhilesh Kumar Srivastava, Suneet Kumar Gupta, Rijwan Khan
Conversion of Intermittent Water Supply to Continuous Water Supply of Chandigarh: A Case Study

In India, where water is supplied to residents on an intermittent basis due to limited sources, implementing a continuous water supply scheme in an Indian city seems quite an arduous affair. To manage limited water resources to pump for 24 h without adding new water sources makes the project more quandary to work further. Achieving the milestone of continuous water supply in India would be a challenge worth taking and worth benefiting from. The case study prepared here provides insights regarding the importance and impact of the proposed methods and models here, based on the data received from government agencies working in this field. The goal is to prepare the outcomes of converting intermittent water supply of a city into regular water supply for the whole of Chandigarh city and not just a part of it. With the help of data collected from site visits, experimental analysis, and pilot experiments done on small scales, a case study has been prepared of how achieving continuous water supply can be made possible for a pan city with a population of more than 1 million.

Sanjeev Chauhan, R. M. Belokar
A Novel Compression Method for Transmitting Multimedia Data in Wireless Multimedia Sensor Networks

This paper discusses various compression methods used in wireless sensor networks. Compressed sensing is the emerging signal processing tool that makes the transmission of data easy via low-data rate links. In the wireless sensor network applications, a group of sensors is used to sense any events and make decisions, and the collaborated information sensed by different tiny sensing devices are used to give the decisions about the occurrence of the particular events. According to the different applications and data types, the quality of service parameters and designing parameters for nodes are different. For dealing with low bandwidth in a sensor network, it is most important to reduce the transmitted data bits between sensor nodes or from nodes to sink. In the case of multimedia data such as image signals, the compression is beneficial for the reduction of these bits because fewer bits required less transmission energy. In some situations of the multimedia sensor network, some loss is accepted without affecting the too much quality of results. Data collected by nodes are spatially correlated with each other, so the image samples collected over time by the nodes are also correlated with each other. If only some samples are transmitted, then these samples are sufficient to give the knowledge about the suspected object inside the monitoring area, so the transformation-based compression technique is the good solution for the compression in the case of the multimedia sensor network. In this paper, a Hadamard transform-based compression technique is discussed for image compression with the consideration of different designing parameters of an image signal. In that manner, this work helps us to select the transform and source coding schemes for the compression of image data inside the wireless multimedia sensor network.

Richa Tiwari, Rajesh Kumar
Live Temperature Monitoring: IoT-Based Automatic Sanitizer Dispenser and Temperature Detection Machine

This study presents a live monitoring non-contact temperature detection and sanitizer dispensing system. The method is designed for preventing infection of COVID19 viruses. It maintains and improves community health and reduces the infection’s adverse economic and social effects. The temperature detection (TD) and sanitizer dispenser (SD) subsystems of the LTM are controlled by a single microcontroller. In this study, the TD was created to operate similarly to existing and commercially available handheld infrared thermometers in terms of accuracy, show the temperature read to the user, and provide visual and audible alarms when the sensed temperature exceeds the average body temperature. Furthermore, the SD is designed to efficiently distribute sanitizer by dispensing only once and at the required amount. The experimentation study suggests that the final test findings are satisfactory, demonstrating that the LTM contributes to temperature monitoring and hand disinfection. In the end, we have discussed the limitations and future directions.

Rudresh V. Kurhe, Anirban Sur, Sharnil Pandiya
A Comparative Study of Security Issues and Attacks on Underwater Sensor Network

UWSNs are susceptible due to the unprotected acoustic path, extreme underwater atmosphere, and unique characteristics. UWSNs are subject to a broad range of security risks and malicious assaults due to their open auditory channel, hostile underwater atmosphere, and inherent characteristics. So, we outline several possible assaults at several stages of a typical UWSN communication protocol stack and discuss viable defenses. This article presents an overview of UWSN attacks, difficulties, and security and privacy issues. Also shown and addressed are contemporary security research and techniques.

Samiksha Kumari, Karan Kumar Singh, Parma Nand, Gouri Sankar Mishra, Rani Astya
Discrete Event Driven Routing in SHIP Network using CupCarbon Simulation Tool

In the world of data communication and data sharing, reliable routing and subsequent delivery of the message with minimum latency is a challenge. Various academicians and researchers around the world have devised various techniques, algorithms, and methodologies to ensure safe, secure and timely delivery of the message between source and destination nodes in the network. In the proposed work, effort has been made to demonstrate the broadcasting of messages in smart, hybrid, intermittent, and partitioned (SHIP) network to the sink node with the help of multiple intermediate neighboring node(s) with minimum delay using CupCarbon simulation tool.

Himanshu Duseja, Ashok Kumar, Rahul Johari, Deo Prakash Vidyarthi
Multiband Dual-Layer Microstrip Patch Antenna for 5G Wireless Applications

A microstrip patch antenna (MPA) is compatible with 5G wireless applications due to its lightweight, compact and conformal shape, small volume, and minimal susceptibility to manufacturing tolerances. In the past few years, communication systems have frequently required multiband antennas to prevent the use of several antennas. Here, a stacked microstrip antenna with coaxial probe feed is proposed, and its multiband characteristics are studied for different 5G applications. For 5G applications, the proposed antenna works effectively in the frequency range of 0–10GHz. The variations in the shapes of the stacked microstrip patch antenna, such as circle, pentagon, hexagon, and octagon, are investigated, and it is observed that there is multi-resonance with decreasing lower resonance frequency as the shape varies. Here, two-layer geometry is used with one driving patch and another parasitic patch. This design covers a frequency range from 2.7 to 9.6 GHz with a gain of 9 dB, directivity of 13 dBi, and radiation efficiency of up to 75%. The proposed structure is simulated on IE3D Zealand software and return loss, directivity, gain, and radiation efficiency have been analyzed and measured.

Vineet Vishnoi, Pramod Singh, Ishan Budhiraja, Praveen Kumar Malik
Distance-based Energy-Efficient Clustering Approach for Wireless Sensor Networks

Wireless sensor network (WSN) is a vast field for the research and development in many types of applications. In WSN, multiples nodes deployed into the environment; each node has energy level. Optimized energy consumption is main concern for any kind of applications like as military/battlefield, smart farming, medical science, vehicular ad hoc networks (VANET). Thousands of nodes deployment in sensor area become a typical task and later maintain the energy consumption as well. The level of consumption of energy consumption of network needs to be focused to prolong the life time. This research work increases the energy level of sensor network using distance-based technique for WSN. It elects the cluster head on the bases of distances between sensor nodes and from base station also considered. The implementation shows the graphical representation of sensor nodes and calculates the energy consumption of each node along with cluster head and also gives the comparison between clustering and quad clustering. This technique represents total energy which is transmission energy, receiving energy, and data aggregation energy through a graph. This proposed work examines that enhanced distance-based technique increases the life time of sensor network for the advancements in the WSN applications.

Bhawnesh Kumar, Naveen Kumar, Harendra Singh Negi, Rakesh Kumar Saini
Emerging Communication Technologies for Industrial Internet of Things: Industry 5.0 Perspective

The Internet of things (IoT) has emerged into various application areas like agriculture, healthcare, defense, transportation, and manufacturing. The transformation of real things in the physical world to the Internet of things given a rise to industrial IoT (IIoT). IIoT applications are intended for the automation of the manufacturing industry, called Industry 4.0. Also, due to needs of end-user personalization, Industry 5.0 is becoming popular nowadays. Industry 5.0 is intended to inject artificial intelligence (AI) into human lives to improve capabilities and productivity. To make Industry 5.0 a successful revolution, IIoT must provide better efficiency, improved productivity, and better asset management. In this context, device-to-device communication plays an important role. IoT devices must be enabled with seamless communication technologies over heterogeneous networks. In this paper, communication standards, technologies, and various published research contributions are reviewed. Further, an analysis is presented to formulate challenges and opportunities for designing communication methods for IIoT. The paper also provides general directions for developing communication techniques in perspective to Industry 5.0.

Nagesh Kumar, Bhisham Sharma, Sushil Narang
Explainable Artificial Intelligence (XAI): Connecting Artificial Decision-Making and Human Trust in Autonomous Vehicles

Automated navigation technology has established itself as an integral facet of intelligent transportation and smart city systems. Several international technological organizations have realized the immense potential of autonomous vehicular systems and are currently working towards their complete development for mainstream application. From deep learning algorithms for road object detection to intrusion detection systems for CAN bus monitoring, the functioning of a self-driving vehicle is powered by the simultaneous working of multiple inner vehicle module systems that perform proper vehicle navigation while ensuring the physical safety and digital privacy of the user. Transparency of the vehicle’s thought processes can assure the user of its credibility and reliability. This paper introduces explainable artificial intelligence, which aims to converge the decision-making processes of Autonomous Vehicle Systems (AVS). Here, the domain of Explainable AI (XAI) provides clear insights into the role of explainable AI in autonomous vehicles and increase human trust for AI based solutions in the same sector. This paper exhibits the trajectories of transportation advancements and the current scenario of the industry. A comparative quantitative and qualitative analysis is performed to compare the simulations of XAI and vehicular smart systems to showcase the significant developments achieved. Visual explanatory methods and an intrusion detection classifier were created as part of this research and achieved significant results over extant works.

A. V. Shreyas Madhav, Amit Kumar Tyagi

Advanced Computing Technologies

Frontmatter
An Empirical Study of Design Techniques of Chatbot, a Review

In recent times, evaluation of the informal coordination in the form of communication between the human being and the electronic brain is making the good progress. Human being or the electronic brain protected system is being used extensively for logical language/terminology development procedure. Chatbot acts as electronic brain which permits human being through the electronic brain applying logical terminology. Chatbot coordination is being used in the different areas like travel, e commerce, customer service, etc. Representation of the chatbot requires different procedures. Hence, by this study or in such work, authors or we introduce summary of the procedures which are used to layout the chatbot. Some steps of the chatbot layout are shown by generally reviewed knowledge as in what way chatbot layout works and which forms of methods are useful for the evolution of chatbot. By fast evolution of the chatbot technology, we can expect chatbot which can enhance human being limitations as well as maximize efficiency.

Akanksha Yadav, Namrata Dhanda
An Approach for Cloud Security Using TPA- and Role-Based Hybrid Concept

In the cloud environment, enormous amount of the data is shared on the server for the availability of access to the employees or customers related to the organization. Two main issues which are generally faced—when data is shared in cloud environment, first is authenticating the user who can access the data, and secondly, to secure the data itself. Seeing the concern, we proposed the hybrid concept which involves the role-based security as well as TPA-based security. By making use of role-based security, first we have authenticated the users using the graphical authentication in which first the image requires to be selected, then the image gets segmented into image blocks, which when selected then the pattern is formed which is used for the authentication purpose, after that when the user shares the file, then he/she specifies the role who can access the file.

Pooja Singh, Manish Kumar Mukhija, Satish Kumar Alaria
Decision Tree Algorithm for Diagnosis and Severity Analysis of COVID-19 at Outpatient Clinic

This study investigates the feasibility of decision tree algorithm like CART recursive method for classifying participants into test-based positive cases and negative cases to detect COVID-19 in the outpatient and suggest admission or home isolation according to the evaluated parameters. It also evaluates the severity of the outpatients using the values of RTPCR test and Chest X-Ray imaging results. A theoretical and predicted decision tree is proposed in the study after focus group interview with a clinical physician. Primary data was collected from the survey of patients visiting a physician for treatment of COVID-19 during the first wave. CART algorithm was applied for predicting the required decision tree. According to the predicted decision tree, it was determined that the most important feature while treating a COVID-19 patient is their history of contact with the positive coronavirus patient. Based on the valuation of dataset, the predicted decision tree provided similar results to that of the conceptual tree. Thus, comparing both trees, it can be evidently said that the predicted decision tree is a subset of conceptual decision tree and can be used by physicians for diagnosis and severity analysis of COVID-19.

Ritika Rathore, Piyush Kumar, Rushina Singhi
CSBRCA: Cloud Security Breaches and Its Root Cause Analysis

Over the last three decades, the computing world has shifted from centralized to distributed systems, and we are now returning to virtual centralization like Cloud Computing (CC) systems. An individual user has complete control over the data and operations in his or her machine. On the other hand, there is CC, in which the operation and data preservation are given by a vendor, leaving the client/customer oblivious of where the activities are operating or where the information is saved. As a result, the customer has no management control over it. The Internet is used as a communication medium in CC. When it comes to data security in the cloud, the vendor must provide some guarantee in service level agreements (SLA) to persuade the user on privacy concerns. As a result, the SLA must specify several degrees of security. Their complexities are based on the services in order for the customer to comprehend the security measures that are in operation. Regardless of the supplier, there must be a defined method for preparing the SLA. The results obtained here indicates management of the SLAs using the Vitrage plugin of the Open stack public cloud eco-system. The results readings were based on the CPU utility of the cloud IaaS resources. The proposed approach shows how the prediction of the resources are managed in real-time when the occurrence of the attack will be identified.

Vivek Kumar Prasad, Vipul Chudasama, Akshay Mewada, Madhuri Bhavsar, Asheesh Shah
A Mobile-Based Patient Surgical Appointment System Using Fuzzy Logic

The advent of artificial intelligence in medical field is playing a significant role in improving healthcare services. In healthcare, there is always need for an intelligent method to schedule resources and patients in order to reduce patient waiting time. The treatment process of patients from their arrival to the starting time of consultation is accompanied by uncertainties. Therefore, this study developed a fuzzy and a mobile-based solution for patient surgical appointment system based on some relevant input variables. The proposed system was simulated using MATLAB fuzzy inference system with a triangular member function. The range of the fuzzy inputs was then fed into the developed mobile-based application for an optimal patient surgical appointment system. The evaluation findings revealed that the proposed framework is efficient in terms of scheduling patient surgical consultations.

Femi Emmanuel Ayo, Sanjay Misra, Joseph Bamidele Awotunde, Ranjan Kumar Behera, Jonathan Oluranti, Ravin Ahuja
Implementation of Green Technology in Cloud Computing

The increase in the use of modern technology worldwide in various sectors like business, software development, automation, etc., has made human life considerably easy, comfortable, and far from any danger. For this reason, there has been a rapid growth in technologies and upcoming trends. Cloud computing is one such paradigm which has rapidly developed in recent years. It has a range of applications in various domains due to the features it offers, like scalability, elasticity, and cost saving with low maintenance, security, and reliability. However, the production and consumption of these advanced technologies have a negative impact on the environment causing massive energy consumption and generating carbon footprints. Due to this, the concept of green technology has gained a lot of attention in order to make positive changes to the environment. Green computing is the implementation of environmentally friendly concepts in computing to increase power and energy and reduce carbon content. In this paper, we review that how adopting a cloud-based architecture has reduced the energy consumptions levels, and we further survey various methods and algorithms which can make clouds greener and more energy efficient.

Soha Bhatia, Anushka Shrivastava, Radhika Nigam, Punit Gupta
Concurrency Control in Distributed Database Systems: An In-Depth Analysis

Distributed databases are databases that spread across multiple locations, often crossing geographical boundaries. It has been a popular research topic because of the novel set of problems it brings to the table. One of the problems is maintaining consistency in the database. Concurrent access to the database gives rise to consistency and integrity issues that need to be resolved. Various methods have been put forward, and this paper explores some of those methods, particularly on-lock and timestamp-based techniques. It also analyzes all these methods based on various parameters.

Husen Saifibhai Nalawala, Jaymin Shah, Smita Agrawal, Parita Oza
House Pricing Prediction Based on Composite Facility Score Using Machine Learning Algorithms

Various features of a house play some role to determine its price. Out of these, location is the dominant feature to determine the price. Besides location, there are some other features which affect the price of a house like area, sports facility, hospital, 24 × 7 security, etc. In this paper, 40 features, available in dataset of houses, are taken from Kaggle platform and have been considered for prediction of house prices. The data of six different cities of India has been included, and these are Delhi, Bangalore, Hyderabad, Kolkata, Mumbai, and Chennai. Here, we endeavored to develop a predictive model for anticipating the price dependent on a specific number of highlights that influence the price. Six machine learning algorithms are used to develop models and compared based on their accuracy of prediction, and the most accurate model is used to determine the price of houses.

Santosh Kumar, Mohammad Haider Syed
Malicious Website Detection Based on URL Classification: A Comparative Analysis

Phishing has been one of the most frequent cyber threats in the recent decade, prompting an increase in anti-phishing research and the development of numerous solutions for detecting and preventing phishing assaults. This paper identifies the system’s vulnerabilities and adversaries’ tactics to deceive Internet users into trusting the malicious email or website and providing sensitive information and credentials. For this study, the relevant URL features are retrieved from the collected dataset that includes phishing and legitimate URLs of websites. The correlation among different features is studied that can help users to identify fake web URLs by scanning phishing specific properties. This paper also analyzes the performance outcome of the machine learning, ensemble, and deep learning techniques on the collected dataset. Each model’s performance is compared and measured, and random forest and gradient boosting with XGBoost are found to be the best optimal model for phishing binary classification problem in terms of accuracy (97.3%).

Swati Maurya, Anurag Jain
Attribute Selection, Sampling, and Classifier Methods to Address Class Imbalance Issues on Data Set Having Ratio Less Than Five

Many modern approaches to classification presume that the underlying training set is uniformly distributed. In a class unbalanced grouping, where the minority class is generally the more fascinating class, the majority class's training set greatly outweighed the minority class's training set. The authors of the current study discuss the difficulties that occur when training the machine as a result of the class imbalance. Research on current techniques checks the performance of the techniques on various parameters. Doing this will help us in getting idea how the standard techniques perform while minority class is suffering from one or more kind of issues. The purpose of this article is to present a comparative analysis of techniques for contemporary imbalance data analysis techniques, with a focus on data pre-processing, attribute selection, and algorithmic analysis, as well as a comparison of these techniques in the context of different data distributions.

Aarchit Joshi, Kushal Kanwar, Pankaj Vaidya
Timely Prediction of Diabetes by Means of Machine Learning Practices

In the past few decades, the quality and quantity of medical data generated by digital devices have been significantly improved, which makes data generation cost-effective and simple, thereby increasing its leading position in the field of big data and machine learning. There is a huge application of machine leaning and artificial intelligence in health care sector. The use of machine learning to train the machine to classify the medical cases taking care of the historical data can be a boon in medical studies. In this paper, we have analyzed many machine learning algorithms and classifiers which are used to make prediction on the diabetes based on the chosen features and attributes of the dataset. The implementation of the algorithms and its performance are compared in terms of accuracy. The proposed model uses soft voting ensemble techniques to the standardized Pima diabetes data to best fit the data and high accuracy.

Rajan Prasad Tripathi, Punit Gupta, Mayank Kumar Goyal

Data Analytics and Intelligent Learning

Frontmatter
Detection of Brain Tumor Using K-Means Clustering

Machine learning has been playing a vital role in the field of computer vision. It has many applications in the field of detection of diseases, especially brain tumor diagnosis. In brain tumor detection, segmentation has an important role. In this study, an efficient approach has been adopted. Segmentation has been done using the k-means clustering method. The main idea behind this color-based segmentation approach with K-means is to convert a gray-level MR image into a color space image and then use K-means clustering and histogram clustering to differentiate the position of tumor objects from other items in the MR image. Experiments reveal that the method can successfully achieve segmentation for MR brain images to help pathologists distinguish exactly lesion size and region.

Ravendra Singh, Bharat Bhushan Agarwal
On Efficient and Secure Multi-access Edge Computing for Internet of Things

The explosive growth of the Internet of Things (IoT) and smart devices currently has been drastically encouraging the development of edge computing. To improve the quality of service (QoS) with low latency in IoT applications, edge computing acts as a promising paradigm that transfers the data from cloud to edge nodes. The importance of minimal delay in critical IoT networks is being considered a highly prioritized task. The current review focuses on minimizing the gap between the resource allocation layers and resource consumer devices and nodes. Reduction in these gaps reduces the communication flow to external sources thereby involving real-time communication and reduction in delay. The remarkable development of edge computing leads to the ignorance of security threats in edge computing Therefore, keeping in view the security threats, we have addressed the security challenges on the edge of a network. The most recent security preserving mechanisms have also been taken into consideration for the secure transmission of data on edge.

Akshita, Yashwant Singh, Zakir Ahmad Sheikh
Execution Survey and State of the Art of Different ML-Based Ensemble Classifiers Approach Contextual Analysis of Spam Remark Location

The digital podium is proving as an increasingly important area for the contemporary development of civilization. However, it additionally engenders a rudimentary conundrum. Spamming is one of the most solemn quandaries that puts state-of-the-art security to the test. Spam wires, which send offensive messages to an immensely voluminous number of recipients, conventionally have become an apperceived security peril. There are various ways spam security issues can be addressed, including utilizing a machine learning (ML) complement system. Ensemble classifier is one of the most commonly used ML approximations. Ensemble methods use different models to amend execution. In various examination fields, like computational erudition, stats, and machine learning uses ensemble classifiers. This paper surveys traditional and verbally express-of-the-art ensemble approaches, accommodating a comprehensive overview for both practitioners and newcomers. In customary outfit strategies like Ada boost, Bagging classifier, extra trees sorts the ensemble techniques; gradient boost; logit boost; random forest; real Ada boost. This investigation is fixated on the ensemble frameworks to slant toward the spam (channel spam or ham remarks) security issue. Remark datasets are utilized for a fascinating judgment of over 41k comments and not for spam. We can split the experimental dataset into two parts. The first uses 30k for training, and the second utilizes the remaining 10k for testing. End-of-heuristics evaluation utilizing accuracy, precision, recall, f1 score, AUC score, model preparation time, and mean squared error reveals that Extra Trees outperforms numerous models in various exhibit metrics.

Biswajit Mondal, Subir Gupta
Real-Time Eyesight Power Prediction Using Deep Learning Methods

This paper describes a real-time eyesight power prediction using deep learning methods. Artificial intelligence (AI) based on deep learning algorithmic methods has been widely adopted by researchers in health care for speech recognition, image processing, etc. Similarly, these deep learning methods can be useful in predicting the eyesight of a person. In order to check eyesight, we need refractometer, but due to its high cost, it is very rarely available in rural areas. Also, there are no online means to check your eyesight which leads to more difficulties. So, in this paper, we have proposed a model on the data we have collected from the survey which contained questions set by an ophthalmologist Vikas Saraswat for the prediction of axis, spherical power, cylindrical power, and addition power. These are the essential factors for the detection of eyesight power. Experimental results were shared with graphical representation comparing the traditional methods for the detection with the proposed models.

Amit Saraswat, Abhijeet Negi, Kushagara Mittal, Brij Bhushan Sharma, Nimish Kappal
An Unsupervised Machine Learning Approach to Prediction of Price for Taxi Rides

Taxi services are the primary method of transportation in urban areas. With the advent of technological sophistication and digital innovation used by companies like Uber and Ola, taxi businesses are undergoing a rapid transformation. Various methods have been developed by product engineers of software companies in the past, but they did not consider the demand for a customer’s ride in a particular region. In this paper, a machine learning-based model has been proposed having the capability to automatically classify booking points into different areas based on optimizing the within-cluster sum of squared distances to estimate the taxi demand in different geographical zones of a city. A robust and accurate price prediction model has been developed which would assist in predicting the price of rides from one fixed location to another fixed location based on the time and location of booking.

Ankit Kumar, Kunal Jani, Abhishek Kumar Jishu, Visaj Nirav Shah, Kushagra Pathak, Manish Khare
Facial Landmark Features-Based Face Misclassification Detection System

Issues of face spoofing that can evade the verification system by placing the photo of real user on camera have been discussed a lot in the literature survey. By detecting the person through misclassification, the problem could be minimized. Therefore, in this paper, robust face misclassification detection system is proposed using ABT mechanism. The proposed system provides the additional level of security before face recognition module. Face landmark features such as eye, nose, and mouth movements are used for generating challenges for detecting fake users from genuine users using misclassification. The reliability of system is tested by placing photographs and videos from Replay-Attack database and live database. Proposed system gives good results under spoofing attacks such as eye imposter attack and mouth imposter attack. The results show that system detects the fake user when implemented on all types of attacks and confirms the 79.6% misclassification detection.

Aditya Bakshi, Sunanda Gupta
Predictive Model for Agriculture Using Markov Model

With upcoming technologies in farming and new varieties of seeds, it is a new challenge to adopt new ways of farming with changing climate conditions. So IoT gives us a new way to evolve with upcoming challenges. Production of crops is always influenced by the weather conditions; climate change has drastically changed the scenario. In new generation, new seed is evolved to get better yield, but they come with their own climate and water requirements. So in this work, we have tried to train the machine with the life cycle of the crop and make suggestion to the farmer with automated maintenance to maintain the environmental requirement of the crop at various stages of life cycle of the crop. The highlighting feature of this smart farming project is to maintain and meet the dynamic requirement of the crop and maintain the environmental condition based on the life cycle of the crop rather than making static threshold-based system. The farmer can check for new suggestions based on the growth of the crop which is directly proportional to the condition and height of the crop. Controlling of all these operations can be handled through any computer connected to Internet and connecting sensors with Intel Galileo 2.

Punit Gupta, Sumit Bharadwaj, Arjun Singh, Dinesh Kumar Saini
A Comparative Analysis of Edge Detection Using Soft Computing Techniques

Detecting edges is one of the most significant aspects of computer vision. Typical methods for edge detection like Sobel and Canny are robust and fast, but they are sensitive to noise. Soft computing techniques such as particle swarm optimization (PSO), ant colony optimization (ACO), genetic algorithms (GA) and fuzzy logic system (FLS) have extensive application in edge detection of images because of their adaptive behavior. Edge detection is identifying the discontinuities in intensity of the pixel and grouping the contour of edges. The quality of edges in ACO-based edge detection majorly depends on the choice of constants, pheromone evaporation rate, number of iterations etc. In PSO-based edge detection, the quality of images depends on the values of acceleration coefficients and inertia weight. However, thresholding is major stakeholder in determining the fitness of the chromosomes. The population contains 2-D chromosomes. Fuzzy systems are most suitable for designing edge detection hardware. This paper presents a thorough comparative study of soft-computing-based edge detection techniques and highlights their key features. The factors affecting quality of edges are compared, and the actual outcomes of the approaches are systematically arranged for better understanding.

Ankush Verma, Namrata Dhanda, Vibhash Yadav
A Comprehensive Study of Pose Estimation in Human Fall Detection

According to a study, unexpected fall is one of the main causes of sudden demise in elder persons. Therefore, it is very important to take immediate safety measures for the people having age 65 or above, or the people who are physically or mentally disabled. A powerful fall detection system to identify and provide immediate assistance to senior citizens or the people who is prone to falls is needed. A medical alert system with fall detection allows the user to summon assistance without pressing the call button. This review paper identifies the comparison in the approaches used for fall detection based on machine learning algorithm. A brief discussion on the methods used in pose estimation like OpenPose and PoseNet, which are majorly used to detect the fall and non-fall of a person is done. Moreover, we have also discussed the privacy concern of a person while using camera-based technique for detecting fall.

Shikha Rastogi, Jaspreet Singh
Study and Develop a Convolutional Neural Network for MNIST Handwritten Digit Classification

The goal of this analysis has been on the development of handwritten digit recognition with the use of the MNIST dataset. In the latest days, the identification of handwritten digits has become a challenging research topic in machine learning. Due to physically formed digits having varying lengths, widths, orientations, and positions. It may be utilized in several ways, such as the amount and signature on bank checks, the location of postal and tax papers, and so on. This research used CNN for recognition. Total four steps followed by pre-processing, feature extraction, training CNN, classification, and recognition. Along with its great higher accuracy, CNN outperforms other methods in detecting essential characteristics without the need for human intervention. On top of that, it incorporates unique levels of convolution and pooling processes. Through CNN, 97.78% accuracy was obtained.

Disha Jayswal, Brijeshkumar Y. Panchal, Bansari Patel, Nidhi Acharya, Rikin Nayak, Parth Goel
Unravel the Outlier Detection for Indian Ayurvedic Plant Organ Image Dataset

Image-based outlier detection has been a fundamental research problem for machine learning and computer vision researchers. This paper unravels the outlier detection process for the data preparation framework of the Indian Ayurvedic plant organ image dataset. While creating dataset the outlier images might get introduce due to human or device errors. Identification and rectification of such outlier images are crucial part for creating clean dataset. This paper evaluated and compared four well-known and state-of-the-art outlier detection algorithms, namely Isolation Forest, Local Outlier Factor, Histogram-Based Outlier Score, and One-Class Support Vector Machine for detecting the outliers from the dataset of Indian Ayurvedic plant organ images. For this experiment dataset containing 690 images of “Centella asiatica” was used and augmented to generate more image samples. In total, 21 morphological, geometric, color, and texture features have been extracted from each plant organ image. The experiment shows the isolation forest giving superior results with 91% accuracy, at the same time Histogram-Based Outlier Score proves to be the fastest in execution time.

Meera Kansara, Ajay Parikh
A Review on Service Delivery in Tourism and Hospitality Industry Through Artificial Intelligence

AI in service industry like tourism and hospitality is changing at an impressive pace and has uncovered new research opportunities. It has been progressively reshaping the service industry and has led to significant innovations in this sector. This study focuses on the systematic review of artificial intelligence in delivery of service in the field of tourism and hospitality. The purpose of the paper is to explore and signify the relevance of artificial intelligence in tourism and hospitality industry in the contemporary times to meet the challenges posed by the pandemic and to ensure speed and accuracy in service delivery for enriching guest experience and sustaining competition. Paper mainly discusses about the optimum use of artificial intelligence through the adoption of AI technology in service delivery in tourism and hospitality industry. The use of AI-enabled tools like chatbots, smart rooms with voice control system, facial recognition technology, robots, operational analysis, and virtual reality in hospitality industry has been analyzed. This study also explores the acceptance of AI by the customer in the tourism and hospitality industry.

Yashwant Singh Rawal, Harvinder Soni, Rakesh Dani, Purnendu Bagchi
MegaMart Sales Prediction Using Machine Learning Techniques

These days online shopping and MegaMarts record their sales and purchase data of each and every item. As the competition between various stores is increasing rapidly, it is necessary to predict future demand of each product at various stores for the customers. This data contain various attributes related to product like its ID, store ID, weight of product, visibility percentage of product, its fat content, its type, location of store, etc. This data are then analyzed to detect the further, anomalies and frequent patterns in the data. After analyzing data, it is processed so as to give us exact report for sales of each product. Then, final data can be used for predicting future sales using different machine learning techniques. We apply different machine learning models like ‘linear regression’, ‘decision tree’, ‘random forest’, ‘ridge regression’, and ‘XGBoost model’ to predict outlet sales. We found out that XGBoost gives us the best accuracy. With this predicted sales, MegaMart can observe the various patterns that should be changed to ensure its success in business.

Gopal Gupta, Kanchan Lata Gupta, Gaurav Kansal
Collaborative Filtering-Based Music Recommendation in View of Negative Feedback System

Recommender systems (RS) are information filtering algorithms that suggest users items that they might be interested in. In this paper, the authors have proposed a content-based approach that maintains fresh recommendations in a music recommendation ecosystem that improves by suggesting new recommendations. A collaborative filtering system has been proposed alongside a negative feedback system (NFS). This results in a much newer array of song recommendations based only on the songs which the user likes, and due to NFS, it can be easily recognized by the user with the precision of 16.78%. Analysis of the results reveals that the song recommendations made by the newly proposed system have a significantly lower intersection with songs that users play from general playlists and available music datasets. Thus, the proposed system allows users to discover new recommendations every time they use the NFS recommendation algorithm and thus performs better compared to the old content-based algorithms, such as popularity-based filtering mechanisms.

Jai Prakash Verma, Pronaya Bhattacharya, Aarav Singh Rathor, Jaymin Shah, Sudeep Tanwar
Internet of Things-Based e-Health Care: Key Challenges and Recommended Solutions for Future

Internet of Things (IoT) has changed the way of living today. Today, Internet connected things (ICT) are increasing at a rapid rate and connecting with devices to reduce load from human being. Irrespective of the sector, the IoT devices are everywhere taking care of everything from the agriculture sector to the sector of manufacturing. But, due to the global COVID 19 pandemic, the sector of health care demands the major use of IoT today. Due to the prevailing pandemic, healthcare professionals also choose to treat the patients virtually rather than treating them physically. IoT plays a major role here. But, most of the application providers or service providers or any other system involving IoT devices for generating and storing data may become a way of leak of information or stolen by a third party for black mailing or financial gain thus leading to privacy and security leak of the user. This work includes all such views with various issues and recommended solutions for the same. Also, other security and privacy requirements and corresponding solutions are also included to provide future researchers a solid base and a clear depth in knowledge regarding the security and privacy issues and solutions required.

Gadiparthy Harika Sai, Khushboo Tripathi, Amit Kumar Tyagi
Deep Learning and Machine Intelligence for Operational Management of Strategic Planning

Currently, industries and businesses are adopting the concept of AI in the technological frontier, while some are opposing the progress. Many financiers are putting more money into AI businesses intending to just see AI adoption in marketing grow at a rapid pace since they are ready to pay for AI equipment, applications, and interfaces. Facebook, Google, and other Internet behemoths are developing tools to kick-start targeted advertising and improved searching. Nevertheless, gaining an understanding of how conventional businesses in the retailing, medical, and telecoms industries spend their own money on AI initiatives is important. Concerning machine learning, the next digitalization frontier is intended to be unleashed using AI. As a result, businesses should be prepared for this type of development since it provides a real-world edge to the corporate sector as a result of the forthcoming digital changes. This article focuses on five AI technical innovations: self-driving cars, computerized visions, robotic systems, deep learning, and virtual assistants, which cover a wide range of current AI breakthroughs and acquiring knowledge. AI development is rising all the time, with Baidu and Google now leading the market. Globally, we estimate that the technical behemoths spent around $2 trillion on AI alone and in 2021. Approximately, 90% of the total funds has been committed to R&D, with the remaining 10% going into AI acquisition. Grants from PE and VC firms, as well as financing and startup funds, have grown considerably. Deep learning continues to be a growing technology investment with a substantial presence in both the internal and external corporate worlds.

Anupam Kumar Sharma, Prashant Singh, Prashant Vats, Dhyanendra Jain
Machine Learning-Enabled Estimation System Using Fuzzy Cognitive Mapping: A Review

With a growing interest in Explainable Artificial Intelligence, the fuzzy cognitive maps (FCMs) have proved to be a simple yet powerful tool for causal reasoning and decision making. It is a hybrid methodology that combines the aspects of recurrent neural network and fuzzy logic. In this paper, we elaborate a FCM technique for Web effort estimation which is a critical challenge in software engineering. Web applications market size is giant and ever-growing. Today, Web applications are becoming more refined as it is not only for uploading and fetching the data but also gathering data from various sources and analyzing by the ML techniques. The mean square error (MSE) is measured and analyzed to show the superiority of FCM estimation technique. It is analyzed that the project characteristics presence should not be ignored by the effort estimation technique selection. On software estimation technique recommendation, there is 70% success probability by the FCM approach.

Ashutosh Sharma, Alexey Tselykh

Latest Electrical and Electronics Trends

Frontmatter
Energy Efficiency in IoT-Based Smart Healthcare

IoT-based smart healthcare is a new technological shift toward efficient, convenient, cheaper, and faster medical services using artificial learning, big data analytics, sensor technologies, and cloud computing. Smart healthcare reduces the time and cost to avail the services region-wise rather than to get the clinical services at distant locations. Along with this, smart healthcare poses many challenging issues. One of the greatest challenges in smart healthcare is to accomplish the energy-efficient services as smart nodes are energy constrained. Therefore, this study addresses the energy consumption challenges in smart healthcare sector. Then, we focus on energy preserving mechanisms to reduce the energy consumption. The various energy-conserving mechanisms such as intelligent techniques, duty cycling techniques, collision resolution techniques, and edge techniques in context to smart healthcare have been discussed for reducing energy consumption.

Pallavi Sangra, Bharti Rana, Yashwant Singh
T-Shaped MIMO Microstrip Patch Antenna for C-Band Applications

The demand of high capacity in the wireless communication systems is increasing as the users and applications are continuously increasing. The capacity of the wireless systems can be enhanced using multiple input multiple output (MIMO) technology. This article presents the design of a T-shaped MIMO microstrip patch antenna (TSMMPA) for C-band applications. The T-shaped structure provides the high gain (GN) and directivity (DY), and the two orthogonal T-shaped structures make the antenna (ANA) suitable for 2 × 2 MIMO communication systems. The ANA structure is simulated and optimized using CST microwave studio software (CSTMSS). The proposed ANA provides the wide bandwidth (BH), wide 3-dB beamwidth, high efficiency (EFY), and easy to fabricate. The presented TSMMPA operates at the C-band frequencies.

Piyush Kumar
Eye Disease Detection Using Transfer Learning on VGG16

Deep learning has emerged as a breakthrough technology in varied fields like health care, computer vision, natural language processing and many more. Ocular infections like diabetic macular edema (DME), choroidal neovascularization (CNV) and DRUSEN are commonly found eye diseases in humans and can lead to temporary or permanent loss of eyesight. The optical coherence tomography (OCT) technique is often used for the preliminary screening of mentioned ocular ailments and provides high resolution cross-sectional imaging. In this work, we have focused on classification of normal and abnormal optical coherence tomography by making use of visual geometry group (VGG16) convolution neural network (CNN) model for prompt diagnosis and timely proper medical treatment of the eye diseases mentioned. OCT image is high resolution imaging technique capable of capturing microstructures within human eye. Here, we endeavored to develop a CNN model for classifying OCT images into normal and abnormal category. Our model achieves an accuracy of 99% and precision of 98.8% which is quite improved results in comparison with other state-of-the-art works that we reviewed.

Aditi Arora, Shivam Gupta, Shivani Singh, Jaya Dubey
Text-Based Automatic Personality Recognition: Recent Developments

The use of computation in personality recognition has been explored for several decades now. As such, it is possible to derive personality from the data available on social media, telecommunication signals, and every signal obtained from human–machine interaction. Personality computation has been explored in two major domains: social signal processing and human–computer interaction. Automatic personality trait recognition from textual context is an emerging research topic that has gotten considerable attention in the area of natural language processing (NLP). In this survey, we reviewed the existing works in the field of automatic personality detection from texts and provided a comparative analysis. We identified some open research gaps and discussed major issues presented in existing literature, including issues with current datasets, techniques, personality features, and personality models employed, as well as how they can be bettered in the future.

Sumiya Mushtaq, Neerendra Kumar
Use of a Precious Commodity—‘Time’ for Building Skills by Teachers for Online Teaching During Pandemic by Using Decision Tree and SVM Algorithm of Machine Learning

The competency to perform a particular task effectively and efficiently is what we call a developed skill. Skills could be of any type: communication, leadership, interpersonal, problem solving, decision making, etc. This crucial period of the pandemic has brought along the threats and challenges and several opportunities with it. A chance to learn something new, think out of the box, be creative, convert our idle time into a quality one, etc. All this has given rise to using our time for some productive purpose. For months, we have been facing this pandemic, and ‘Work from Home’ is the policy adopted by almost every company, firm, and educational institution. And this has given all the employees working from home an opportunity to put their saved time into something innovative and productive. So, this study has emphasized the usage of time for skill development by teachers of the educational institutions of Mumbai for online teaching during the period of the COVID-19 pandemic through different training programs. This study is based on the primary data that has been collected from the teachers aged from 30 to 60 and above. Also, its results state that the skills which are required by the teachers for their effective teaching–learning process are developed successfully, and the majority of the faculties have improved their technical skills as well, which in turn have enabled them to adopt new and innovative teaching techniques.

Bharti Khemani, Jewel Sabhani, Mala Goplani
Road Lane Line Detection Based on ROI Using Hough Transform Algorithm

Now-a-days technology has become the means of survival. Automotive Sector is also affected by this technology growth. Driver safety is one of the most important concern for the automobile industry. Lack of attention causes the road accidents and may endanger the driver and co-passenger lives at risk. The stats presented by WHO on road accidents shows that approximately 1.35 million people dies annually as a result of the car accidents. And about 20–50 million peoples suffer from non-fatal injuries, but they may cause lifetime disabilities. These road crashes also impact the economy of the countries. Most of the countries suffers 3% of their GDP due to road accidents. The major challenge is to make the technology available in the commercial sector. So various methods and algorithms are introduced to achieve better performance and robustness. One of the major components of autonomous vehicles are road lane detection. Marking the region of interest (ROI) in which car should be driven. Recent advancement in the technology like image processing and deep learning helps in achieving the aim to detect road lane lines. Autonomous cars are now equipped with cameras, radar and LIDAR for tracking roads and track environment. In this paper, road lane line detection problem has been addressed using Open CV library; also, an approach for finding an efficient way for detecting road lanes precisely and more accurately has been proposed. The road images captured by the camera mounted on the vehicle is processed and region of interest is masked. After masking the ROI, it is converted into a pixel matrix using NumPy library. The Hough Transform is applied on the matrix and lanes are detected in between which vehicle runs.

Mohammad Haider Syed, Santosh Kumar
Dimensionality Reduction-Based Discriminatory Classification of Human Activity Recognition Using Machine Learning

Majority of work in activity recognition using different machine learning and deep learning has shown very challenging results to monitor daily activities. Different datasets available on Web have been used to improve the results, still model fitness need to be verified in terms of different characteristics of matrix and error analysis. Dimensionality reduction (DR) of datasets improves the results of models due to pruning of dataset features. In this paper, we have introduced seven different machine learning models to improve the results. Proposed framework has used principle components analysis (PCA) and linear discriminant analysis (LDA) for dimensionality reduction of UCI-ML dataset. Results show that LDA is better than PCA. Kernel–SVM accuracy has increased from 95.39 to 96.23%. Naïve Bayes has shown 96.78% accuracy with dimensionality reduction. Simple dataset has shown low accuracy while dimensionality reduction has improved the performances of models. We have also introduced different challenges associated with machine learning models, fitness value, and future challenges. At the end of this work, we have done comparative study and error analysis of models.

Manoj Kumar, Pratiksha Gautam, Vijay Bhaskar Semwal
SPECIAL SESSION ON RECENT ADVANCES IN COMPUTATIONAL INTELLIGENCE & TECHNOLOGYS (SS_10_RACIT)
Development of Generic Human Motion Simulation Categorization using Inception based CNN

Generic human motion is the steps taken by the person for everyday movement. In this paper, the generic framework has been proposed for gait activity recognition using the inception-based convolutional neural network (CNN) model. The gait pattern is a compound of seven sub-phases. However, sequential execution of these seven sub-phases of left and right legs is called gait cycle. Gait features are used to perform the biometric, biomechanics, and human psychology analysis of human beings. Human walking is very challenging to measure due to high variability. It depends on age, gender, walking terrain, walking speed, mental condition, health condition, etc. The analysis of human gait helps identify human activities, diagnose many gait-related diseases like Parkinson’s and freezing of gait. Due to variability of gait, biometric is difficult to spool. This paper covers to analyzing the existing available wireless sensor data mining (WISDM) dataset and our gait dataset of human activity (GDOHA) dataset and evaluating the performance. Moreover, we compare performance metrics on datasets applying machine learning and deep learning algorithms. The result achieved 99.03% accuracy using the inception-based CNN model. The proposed computational model helps to early detection of gait abnormality and provides the proper mechanism for recovery from abnormal gait.

Ram Kumar Yadav, Subhrendu Guha Neogi, Vijay Bhaskar Semwal
Cryptanalysis on “ESEAP: ECC-Based Secure and Efficient Mutual Authentication Protocol Using Smart Card”

Very recently, ESEAP mutual authentication protocol was designed to avoid the drawbacks of Wang et al. protocol and highlights that the protocol is protecting all kind of security threats using informal analysis. This work investigates the ESEAP protocol in security point of view and notices that the scheme is not fully protected against stolen verifier attack and does not provide user anonymity. Furthermore, the same protocol has user identity issues, i.e., the server cannot figure out the user identity during the authentication phase. Later we discuss the inconsistencies in the security analysis of ESEAP presented by RESEAP.

Mohammad Abdussami, Ruhul Amin, Satyanarayana Vollala
Modeling, Simulation, and Comparative Analysis of Flyback Inverter Using Different Techniques of PWM Generation

Nowadays, the flyback inverter has increased more interest due to its various advantages. Some of these are its simplicity, low cost, and high efficiency. Applying pulse-width modulation techniques to improve the performance of many types of inverters is under research. In this paper, distinct switching techniques generate the gate pulses for switches of the flyback inverter. The paper deals with the implementation and comparative analysis of three different models of flyback inverters: (1) open-loop flyback inverter with the pulse generator, (2) open-loop flyback inverter with pulse width modulation (PWM), and (3) open-loop flyback inverter with a sinusoidal pulse-width modulation (SPWM) technique. The output voltage and current signals are pure sinusoidal when the SPWM technique generates the gate pulses. The SPWM technique used in model-3 improves the nature of the output AC signal compared to the other two models.

Mangala R. Dhotre, Prashant V. Thakre, V. M. Deshmukh
Industrial Rod Size Diameter and Size Detection

Thermo-mechanical treatment (TMT) rods are the walloping production of the steel industries, and there are giant machines that will make the task of cutting TMT rods easier for industries. While cutting the rods, photoelectric sensor, manual labor, and complex computing machine are used that need huge maintenance of the machine, and it is a time-consuming process. In the last decade, research on digital image processing and computer vision has seen much progress. In this paper, we propose an adaptable methodology for the industries in measuring the TMT rods much more efficiently, maximizing the efficiency, robust toward cram-full of rods and minimizing the error rate. The captured digital image first undergoes the preprocessing phase, where the first step is image enhancement and then edge detection, which extracts the TMT rods edges then followed by the diameter calculations (pixels per metric). An experiment has been conducted with various challenging conditions to demonstrate the capability of our approach to a good measure of success.

Swathi Gowroju, N. Santhosh Ramchander, B. Amrita, S. Harshith
Sentiment Analysis of Twitter Data Using Clustering and Classification

Data mining helps in collecting and managing data besides performing analysis and prediction analysis. The process that is implemented to discover useful data patterns may have different names. Statisticians, database researchers, and professional organizations were among the first to use term data mining. The fundamental steps for sarcasm detection are dataset collection, feature extraction, and classification. This work puts forward a new model of sarcasm detection formed by fusing K-mean, PCA, and SVM classifiers together. With respect to common evaluation metrics like accuracy, precision, and recall, the architecture designed for this work is especially productive.

Santanu Modak, Abhoy Chand Mondal

Security and Privacy Issues

Frontmatter
Image Distortion Analysis in Stego Images Using LSB

One of the most pressing matters at hand of today’s communication networks is privacy. Privacy is essential where information to be transmitted to the desired target without being intercepted by third parties or bringing them in a way that they cannot understand. Lossless text encryption technique for gray scale or RGB images using least significant bit (LSB) is the most commonly used method to hide data. LSB technique is preferred as it is difficult to draw a distinction between the cover object and stego object if few LSB bits of the cover object are replaced. However, image distortion is a major issue which needs critical investigation as it can change the original image. This paper proposes an algorithm to embed text in RGB images and also evaluate the distortion in image at various bit positions. Simulation results suggest that distortion in the image is least at 8th-bit and most at 1st-bit position.

Shubh Gaur, Swati Chaturvedi, Shiavnsh Gupta, Jay Mittal, Rohit Tanwar, Mrinal Goswami
Towards a Secured IoT Communication: A Blockchain Implementation Through APIs

Years ago, intercommunication among systems was not at all possible, but with the development of the application programming interface (APIs-traditional) in the 1960s, this was possible. With the evolution from the 1960s, these APIs have advanced, and the birth of APIs (Modern) took place in the initial 2000s. Many conventional techniques and tools that possess diverse security issues and are operating in a fully functional manner can be solved to a greater extent with the help of APIs if integrated with modern and SMART technologies. One such example is a sub-group of artificial intelligence (AI) i.e., IoT, that possesses diverse issues and challenges and can be solved by the peer-to-peer technology blockchain. Here, blockchain is regarded as the SMART solution. The diverse attacks on IoT spectrum are illustrated in this paper to attain a better scope and security for IoT devices. Moreover, the implementation of blockchain is also illustrated in this paper to depict a convenient approach for solving the issues of IoT with decentralization attribute. The reason is that IoT illustrates a centralized architecture.

Rajat Verma, Namrata Dhanda, Vishal Nagar
Application of Truffle Suite in a Blockchain Environment

Advancement is a term that never stops with something so is technology. Blockchain technology is an old term but with updation, it has become a key-buzz term in the technological market. The advancements in blockchain led to the formation of distributed and decentralized applications. This advancement is only possible in the applications phase of blockchain. A simplified ecosystem in which decentralized apps (DApps) can be built is truffle suite. Majorly, three constituents that are completing the truffle suite are truffle, drizzle and ganache. DApps are those that operate among the users and are not monitored by a central authority. In DApps, the ecosystem of a peer-to-peer network works as a complete operating system. This paper focuses on these DApps using truffle suite and its different scenarios. A quick depiction of blockchain with its connection to the truffle suite is also highlighted in this paper. Moreover, this paper also illustrates a popular use case of a DApp with its real-time issues and concerns.

Rajat Verma, Namrata Dhanda, Vishal Nagar
Assessment of Compliance of GDPR in IT Industry and Fintech

The general data protection regulation (GDPR) is the tough privacy and security policy toward protection of data. The policy was drafted by the European Union but imposed obligations for any organization which collects data related to the people in the European Union. This policy came into effect from 2018. If any organization violets, the law will levy huge fines. Consumer driven companies in the areas like IT services and Fintech likely to affected by the GDPPR and have to comply. The research paper seeks to explore the implication of GDPR on these two industries. The challenges faced by the two industries in planning, implementation, and complying to GDPR, the overlaps and contradictions with the existing industry frameworks which will last post GDP6R implementation, the pre- and post-GDPR scenario analysis, and lastly the trial process for the data breach of the two industries. Based on the comprehensive study and research on the aforementioned areas, this research paper then delves into building a hypothesis through qualitative and quantitative data gathered which provides a solution for the two industries to prepare, plan, implement, and comply with GDPR across industry level with respect to user data management centers and Fintechs. We have used empirical methodology and collected responses through the questionnaire. Through the research study, we have found that how important is data encryption, not only because it is mandated in GDPR but also since any sort of data revelation to a criminal party can cause a lot of damage.

Pankaj Pathak, Parashu Ram Pal, Rajesh Kumar Maurya, Rishabh, Mayur Rahul, Vikash Yadav
Digitally Signed Document Chain (DSDC) Blockchain

This paper suggested the framework of digitally signed document chain (DSDC) blockchain which focuses on digitalization and decentralization of educational certificates storage, authentication, authorization, confidentiality, ownership, and privacy. Blockchain generation has lately emerged as a capacity suggests for authenticating the record verification and a sizeable device to fight record fraud and misuse. This research paper diagnosed the safety subject matters required for file verification with inside the blockchain. This paper also suggested the framework for modifications and deletions of data under circumstances by competent authority only which is against the principal of immutability of blockchain by utilizing chameleon hashing.

Udai Bhan Trivedi, Santosh Sharma
Algorithms of AI in Deciding Optimum Mix Design of Concrete: Review

The preparation of mix design of concrete requires a knowledge of design mix proportioning. Various properties like slump value, compressive strength are considered while preparing mix design. This traditional mix proportion method is a time-consuming and costly process. It is also done manually which may lead to different errors. To overcome this, use of artificial intelligence has been brought into this field to predict the design mix of concrete in limited time, low cost and minimum error due to use of computational algorithms as compared to traditional methods. In this paper, the studies using different algorithms of artificial intelligence are reviewed. Estimation of properties like compressive strength, slump value of concrete is done. Further, this paper also presents comparative analysis between different algorithms of AI. This research paper will be of great help to concrete technologist to explore future possibilities of AI techniques in concrete industry.

Rajat Verma, Uzair Khan, Binod Kumar Singh, Rizwan A. Khan
A Review of Integration of Data Warehousing and WWW in the Last Decade

The data warehouse (DW) is a powerful technology to store and analyse huge volumes of historical data supporting business intelligence. The World Wide Web, or simply the Web, has revolutionized the way to author, share, search and access information. In the past few decades, a significant amount of research has been done in both the DW and Web domains. Interestingly, the integration of data warehousing and the World Wide Web has led to a variety of new opportunities as well as challenges for the researchers and the industry. The main motivation to conduct this systematic review of the relevant research works integrating DW and the Web in the last decade is to provide the groundwork for the research advancement in this field. A total of 27 relevant research works were identified for the research. An in-depth analysis was performed to find the problems addressed, the most relevant research categories, the tools or techniques applied and the application domains of these research works. Encouragingly, our results yielded seven categories and four sub-categories of research employing the integration of DW and Web. On the other hand, we found some open research issues, and the future research works should focus on generalized solutions for handling semantic heterogeneity, change propagation and quality analysis of identified Web sources for the DW.

Priyanka Bhutani, Anju Saha, Anjana Gosain
WeScribe: An Intelligent Meeting Transcriber and Analyzer Application

In all existing organizations regardless of their type or size, meetings are conducted on the regular basis to invite discussions for organizational decision making. While many organizations, even large ones, still hire employees to perform these tasks, there is no doubt that the results are exposed to human error. Documenting meetings’ minutes is essential for its success and keeping track of the work progress and decisions flow, approvals, while keeping it complete, consistent, and coherent. This project idea was proposed by Aramco to develop a suitable solution for a hectic problem. The process of documenting and taking minutes can be tedious, so we aim to automate audio meeting transcription with the use of technologies that convert speech to text while recognizing the speaker and then process and analyze the most valuable information tagged based on persons in the meeting. This goal can be accomplished through the development of an app that uses speech recognition for conversion, voice recognition for identification of speakers, and natural language processing (NLP) for analysis and then combines them all in a transcription form with considerable accuracy. Further, the proposed system identifies potential events, deadlines, and follow-ups and adds them to the speaker’s calendar upon approval. In the future, we aspire to expand it with some features such as increasing the number of meeting members, creating special sections for each department in the company which adopt WeScribe, and feed our NLP model with more data to develop its performance and increase its accuracy.

Mohammad Aftab Alam Khan, Maryam AlAyat, Jumana AlGhamdi, Shahad Mohammed AlOtaibi, Maha AlZahrani, Malak AlQahtani, Atta-ur-Rahman, Mona Altassan, Farmanullah Jan
Customer Churn Prediction in Banking Industry Using Power Bi

The development of technology in our modern day has led to the generation of huge data. This is evident by the 2.5 quintillions of data generated by persons connected to the Internet per day in 2020. With the expectation of 5.3 billion Internet users by 2023, complex and efficient tools, models, or approaches that will explore, analyze, and produce meaningful hidden information from huge data are needed. In recent years, machine learning techniques such as logistic regression, decision trees, and clustering are beginning to gain relevance, especially in churn prediction. Customer churn prediction is the process of determining the proportion of clients who avoid or might stop using or subscribing to a product or service offered by an organization or company. Though various prediction models have been proposed, most research attention has been given to measuring the efficiency of prediction models, rather than identifying its application for sustainable economic development. In this paper, we investigate the determining factor for customer attrition in the banking sector using Power BI. Dataset from United Bank of Africa (UBA), Nigeria was preprocessed with four key customer variables were used. The decision tree algorithm available in the Power Bi software was employed for training and testing. The results show that customer account balance is a key determining variable for churning. Furthermore, the results show that churning occurs less in male than female clients. This work will provide banks with useful knowledge on building effective customer retention strategies. Building an effective and accurate customer churn prediction model is an important research problem for both academics and practitioners.

Awe M. Oluwatoyin, Sanjay Misra, John Wejin, Abhavya Gautam, Ranjan Kumar Behera, Ravin Ahuja
Issues in Credit Card Transactional Data Stream: A Rational Review

Online transactions are trending worldwide now and in future developments. A big amount of transactional data is generated like networking, stock market, telecommunications, and weather forecasting. This data can be classified for the knowledge extraction and learning. Credit card nowadays is very easy methods for physical and online transactions. Transactions using a credit/debit card having some advantages and flaws. Some of the problems with the credit card transactions are also highlighted here, further in the paper focused on the extensive studies of the various learning methods used by various authors on the imbalanced data stream of the credit card transactions.

Rinku, Sushil Kumar Narang, Neha Kishore
Artificial Intelligence-Based Smart Packet Filter

Packet filtering is a fundamental feature for the firewall as the security of the whole system depends on it, but on the other hand, it should also be fast and efficient and should be able to process queries as fast as possible due to the large number of queries sent to the server together. There is a need for an artificial intelligence-based smart packet filter that can act on the packet in a fraction of seconds of receiving the packet. This research proposes an ensemble model (smart learner) made up of a random forest of max depth 11. The average accuracy obtained by the model is 99.76% using the stratified K-gold cross-validation technique. Earlier papers published on packet filtering are dividing the dataset into test and train where their test and train are fixed which can cause overfitting on the dataset and in real life testing can reduce the accuracy of the model, that is why this research focusses on using K-fold cross-validation technique and to support this technique highest accuracy was achieved with ensemble model of random forest. This research can be implemented in the firewall to make them faster.

Mohit Dayal, Ameya Chawla, Manju Khari, Aparna N. Mahajan
Preserving Privacy in Internet of Things (IoT)-Based Devices

According to the global risk reports, data breaches and cyberattacks are in the top 5 deliberate risks. We all are aware of the rapid advancement and deployment of the IoT. Because these technologies are so tightly linked to individuals, privacy and security are important problems in today’s world. Attackers who try to target IoT must constantly expose communication relations to capture transferred data and identify subtle data since they always rely on formerly gathered information to launch their attacks. Sleep is one of the crucial activities to our health. Depression, difficulty in concentrating, and irritability are a few important concerns that are caused by sleeping disorders. Using a sleep tracker may help a person understand their sleeping behavior and detect many important concerns. There are several dangers connected with information gathering since these IoT-based gadgets, including tracking device stowage, data transmission across a system, and information storage in the cloud. The information gathered by IoT instruments can expose the users’ everyday activities, location, and other delicate statistics. Hackers usually try to attack these and when gadgets or the stowage is hewed, they may get confidential information and facts about their personal belonging and that data can be future used for phishing or advertisements. As a result, the privacy of data gathered by IoT-based devices must be protected.

Dheeraj Sharma, Amit Kumar Tyagi
A Sentiment Analysis-Based Recommender Framework for Massive Open Online Courses Toward Education 4.0

The emergence and confluence of progressive technologies like artificial intelligence, Internet of things, and automation in Industry 4.0 have also driven parallel domains like the education sector. Today’s digital education aligns with the progressive dynamics of Industry 4.0, and with the increasing mix of information and communication technology (ICT), we have entered the era of Education 4.0. The ICT tools gather a lot of data content, which is generated through data generation in the form of text, audio, images, and video in online social networks (OSNs), blogs, posts, and many others. Usage of ICT has facilitated the conduction of open courses to masses of people connected through heterogeneous networked applications. Such courses termed as massive open online course (MOOC) platforms have grown significantly and have reaped high profits. However, users browsing for suitable courses in MOOC platforms are faced with challenges of selecting and filtering courses, based on current demands, effectiveness, and pre-requisite knowledge. Scientifically, it is observed that due to incorrect course selection, users are many times not satisfied with the MOOC course, which results in high dropouts. In the past, researchers have addressed the issue through recommender systems for users, but recommendation systems require effective filtering mechanisms for proper results. Thus, to address the research gap, in this paper, we propose an approach that is based on skills information from users’ LinkedIn profiles combined with ratings and review data of courses. For experimental validation, we consider a Udemy MOOC user public dataset and apply natural language processing (NLP) to contextually organize user reviews, skill-set keywords from LinkedIn and refine search keywords. The proposed results indicate the efficacy of the framework toward correct MOOC recommendations for active learners and users.

Akhil Bhatia, Anansha Asthana, Pronaya Bhattacharya, Sudeep Tanwar, Arunendra Singh, Gulshan Sharma
Lung Cancer Detection Using Textural Feature Extraction and Hybrid Classification Model

Medical image processing (MIP) offers powerful and promising key developments in modernistic three-dimensional (3D) medical imaging based on science and medicine due to the creation of hi-tech images. Image processing is used to detect lung cancer. Detecting a cancer nodule consists of three levels. CT scans are generally adopted to identify the incidence of cancer affected nodules. To improve the interpretation of information in an image to a human audience, the step of image enhancement is enforced. The next step of segmentation involves segmenting the required area into many sub-areas. The output of this step is used as input for the next step of feature extraction. Cancer, at this stage, is detected on the basis of the abstracted features. This work implements GLCM with a hybrid classifier model to localize and classify the cancer affected area from the CT scan. The hybrid classifier framework constructed by integrating KNN, SVM, and decision tree classifiers is an efficient cancer detection framework work. This work takes three parameters (i.e., accuracy, precision, and recall) under consideration to evaluate the designed hybrid classifier model.

Jasbir Kaur, Meenu Gupta
Overview of Security Approaches Using Metamorphic Cryptography

Initially, researchers employed only one information security technique either cryptography or steganography to secure the communication. But later, researchers stress on the amalgamation of both cryptography and steganography, and this amalgamation is popularly known as metamorphic cryptography. Steganography can be classified on the basis of cover medium. This paper surveys the different metamorphic cryptography approaches which uses image as cover media for securing the data. This paper also covers general concepts of cryptography, steganography, classification of metamorphic cryptography, and evaluation parameters like PSNR, MSE.

Lokesh Negi, Lalit Negi
A Bibliometric Analysis to Unveil the Impact of Digital Object Identifiers (DOI) on Bibliometric Indicators

Digital object identifier (DOI) is often used as an important identifier of scientific contributions. It raises the readers’ awareness toward genuine and authentic work of authors, organizations, and journals. The aim of this study is to identify the scientific contributions with and without DOI information associated with them in multidisciplinary indexing databases such as Web of Science (WoS). This study also sheds light on the contribution of self-citations in calculating the author, organization, and journal informetrics. The result shows that at author level, 82.2% of publications and 81.6% of citations are with DOIs, at organization level, 76.3% of publications and 73.5% of citations are with DOIs, and at journal level, 83.9% of publications and 62.1% of citations are with DOIs. Author level has 7.7% of self-citations, organization level has 13.7% of self-citations, and journal level has 10.3% of self-citations. Decreases in publications and citations have resulted in an average downfall of h-index 2.9 in author data, 15 in organization data and 12.3 in journal data. Finally, stakeholders are encouraged to review the publications, and citations data with DOIs, and note on the self-citations before considering for final informetrics of authors, organizations and journals.

Parul Khurana, Geetha Ganesan, Gulshan Kumar, Kiran Sharma
Cyber Attack Modeling Recent Approaches: A Review

The advancement in cyber technology has enhanced user convenience tremendously hence accelerated its uses. But at the same time, cyber frauds, threats, and attacks have increased with same pace. So, to protect our cyber system and devices from them, cyber attack modeling is quite essential and challenging task. It provides us the chance to detect and protect our system by applying suitable security measures to them. There are many attack modeling techniques available today. This paper provides an elaborate discussion on the two very popular graphical attack modeling techniques, that is Attack graph and attack tree-based approaches. A comparative analysis of various works done in these techniques is presented here.

Neha, Anubha Maurya
A Secure DBA Management System: A Comprehensive Study

The current world is transforming into a digitalized world. Here, every information is taken care of as online data. Data are growing unquestionably rapidly and irksome in managing them moreover overall taking care of securely. Nonetheless, information is put away in gigantic sums, so consequently, security is similarly significant. For this situation, it is the test for the question author to produce an inquiry that aids in shielding information from unapproved access and malignant assault on data sets. Database management system (DBMS) rules accompany a simple illustration of a question generator; by time, database management system (DBMS) is overwhelmed by relational database management system (RDBMS) for their productive work. After some time, the standard method of inquiry writing in social information bases is confronting many new difficulties because a high measure of information comes from different places additionally with blunders, infections, and in numerous others in vindictive structure. The database administrator (DBA) management framework is made to watch out for the security of information/data. This paper presents various security aspects of the database administrator (DB) management system. The article also gives an analysis of various security threats of the last two decades.

Khushboo Jain, Umesh Jangid, Princy Kansara, Smita Agrawal, Parita Oza
Education 4.0: Hesitant Fuzzy SWARA Assessment Approach for Intelligent Selection of Research Opportunities

Reforms and revolutions are periodic in nature, a cyclic process they are and with innovations in technology and analytics, the idea of learning education has transformed into experiencing the education. With every Industrial Revolution, education pattern, its content and delivery got new dimension. Education in India has been categorically segregated as primary, junior, middle, senior, undergraduation, post-graduation, and research degrees. Aspirants are usually aware and well informed till they reach research degrees level and beyond. With macro- and micro-options within specializations, scholarships, center for excellence, availability of experts, eligibility criteria, research trends, technological advancements, procedural steps during the course, live projects associated, financial aspects, and education policies, the whole process becomes a huge unsolved mystery and aspirants might miss better opportunities which otherwise he might have gone for. The work carried is an effort to identify a fuzzy-assisted pattern for the aspirant to make an informed decision about the research option he or she can avail for enhanced growth. Study proposes a data analytics model for the informed intelligent selection toward optimum research carrier. The selection is assisted by number of weighted parameters which an aspirant should take into consideration while selecting a particular domain. SWARA technique was used for estimating weights of criteria based on experts’ preferences further to establish feasibility of the technique proposed, an empirical study of sustainable organization selection taken under hesitant fuzzy (HF) environment. The data used for the research work was obtained from Zenith Ph.D. Training & Consultancy (ZPTC), Jaipur, proposed technique was tested on data from 300 universities. The organization focusses and specializes in doctoral fellowships, examinations, processes, and regulations in India and has huge database for the same. Priority order $${G}^{*}$$ G ∗ was computed, and degree of utility was estimated ( $${\uplambda }_{i}$$ λ i ) for the proposed analysis. The degree of utility came as 99.8% for K3, 93.1% for K2, and 86.4% for K4. With estimated values, it was established that aspirants preferred organizations in following order K3 > K2 > K4 > K1 > K5 > K6.

Pooja Khanna, Pragya, Ritika Gauba, Sachin Kumar
Backmatter
Metadata
Title
Proceedings of Third International Conference on Computing, Communications, and Cyber-Security
Editors
Dr. Pradeep Kumar Singh
Dr. Sławomir T. Wierzchoń
Dr. Sudeep Tanwar
Prof. Dr. Joel J. P. C. Rodrigues
Dr. Maria Ganzha
Copyright Year
2023
Publisher
Springer Nature Singapore
Electronic ISBN
978-981-19-1142-2
Print ISBN
978-981-19-1141-5
DOI
https://doi.org/10.1007/978-981-19-1142-2