Skip to main content
Top

2022 | Book

Cyber Security, Privacy and Networking

Proceedings of ICSPN 2021

Editors: Prof. Dharma P. Agrawal, Dr. Nadia Nedjah, Dr. B. B. Gupta, Gregorio Martinez Perez

Publisher: Springer Nature Singapore

Book Series : Lecture Notes in Networks and Systems

insite
SEARCH

About this book

This book covers selected high-quality research papers presented in the International Conference on Cyber Security, Privacy and Networking (ICSPN 2021), organized during 17-19 September 2021 in India in Online mode. The objectives of ICSPN 2021 is to provide a premier international platform for deliberations on strategies, recent trends, innovative approaches, discussions and presentations on the most recent cyber security, privacy and networking challenges and developments from the perspective of providing security awareness and its best practices for the real world. Moreover, the motivation to organize this conference is to promote research by sharing innovative ideas among all levels of the scientific community, and to provide opportunities to develop creative solutions to various security, privacy and networking problems.

Table of Contents

Frontmatter
A New Modified MD5-224 Bits Hash Function and an Efficient Message Authentication Code Based on Quasigroups

In this paper, we have proposedUmesh Kumar V. Ch. Venkaiah (i) a hash function and (ii) an efficient message authentication code based on quasigroup. We refer to these as QGMD5 and QGMAC, respectively. The proposed new hash function QGMD5 is an extended version of MD5 that uses an optimal quasigroup along with two operations named as QGExp and QGComp. The operations quasigroup expansion (QGExp) and the quasigroup compression (QGComp) are also defined in this paper. QGMAC is designed using the proposed hash function QGMD5 and a quasigroup of order 256 as the secret key. The security of QGMD5 is analyzed by comparing it with both the MD5 and the SHA-244. It is found that the proposed QGMD5 hash function is more secure. Also, QGMAC is analyzed against the brute-force attack. It is resistant to this attack because of the exponential number of quasigroups of its order. It is also analyzed for the forgery attack, and it is found to be resistant. In addition, we compared the performance of the proposed hash function to that of the existing MD5 and SHA-224. Similarly, the performance of the proposed QGMAC is compared with that of the existing HMAC-MD5 and HMAC-SHA-224. The results show that the proposed QGMD5 would take around 2 $$\mu s$$ μ s additional execution time from that of MD5 but not more than SHA-224, while QGMAC always takes less time than that of both the HMAC-MD5 and the HMAC-SHA-224. So, our schemes can be deployed in all the applications of hash functions, such as in blockchain and for verifying the integrity of messages.

Umesh Kumar, V. Ch. Venkaiah
Leveraging Transfer Learning for Effective Recognition of Emotions from Images: A Review

Emotions constituteDevangi Purkayastha D. Malathi an integral part of interpersonal communication and comprehending human behavior. Reliable analysis and interpretation of facial expressions are essential to gain a deeper insight into human behavior. Even though facial emotion recognition (FER) is extensively studied to improve human–computer interaction, it is yet elusive to human interpretation. Albeit humans have the innate capability to identify emotions through facial expressions, it is a challenging task to be accomplished by computer systems due to intra-class variations. While most of the recent works have performed well on datasets with images captured under controlled conditions, they fail to perform well on datasets that consist of variations in image lighting, shadows, facial orientation, noise, and partial faces. For all the tremendous performances of the existing works, there appears to be significant room for researchers. This paper emphasizes automatic FER on a single image for real-time emotion recognition using transfer learning. Since natural images suffer from problems of resolution, pose, and noise, this study proposes a deep learning approach based on transfer learning from a pre-trained VGG-16 network to significantly reduce training time and effort while achieving commendable improvement over previously proposed techniques and models on the FER-2013 dataset. The main contribution of this paper is to study and demonstrate the efficacy of multiple state-of-the-art models using transfer learning to conclude which is better to classify an input image as having one of the seven basic emotions: happy, sad, surprise, angry, disgust, fear, and neutral. The analysis shows that the VGG-16 model outperforms ResNet-50, DenseNet-121, EfficientNet-B2, and others while attaining a training accuracy of about 85% and validation accuracy as high as 67% in just 15 epochs with significantly lower training time.

Devangi Purkayastha, D. Malathi
An Automated System for Facial Mask Detection and Face Recognition During COVID-19 Pandemic

The coronavirus (COVID-19) pandemicDr. Swati Shinde Pragati Janjal Gauri Pawar Rutuja Rashinkar Swapnil Rokade is an ongoing pandemic of coronavirus disease-2019. It is still spreading continuously across the globe, causing huge economic and social disruption. There are many measures that are suggested by the World Health Organization (WHO) to reduce the spread of this disease. In this paper, we are proposing a system in which people wear masks or not in public and recognize faces who do not wear masks. We detect the people who are monitored by using Webcam and those who are not wearing masks, and the corresponding authority is informed about the same by using convolutional neural network (CNN) with a mobile net and Haar cascade algorithm. The proposed model will help to reduce the spread of the virus and check the safety of surrounding people.

Swati Shinde, Pragati Janjal, Gauri Pawar, Rutuja Rashinkar, Swapnil Rokade
ROS Simulation-Based Autonomous Navigation Systems and Object Detection

Autonomous robots are becomingDr. Swati Shinde Tanvi Mahajan Suyash Khachane Saurabh Kulkarni Prasad Borle popular and are being used in many industries, due to their autonomy features. They are just like humans who have the ability to make decisions on their own without any human help. As the need for such robots is increasing, our paper aims to present an ROS autonomous navigation software system for autonomous robots, which is capable of creating 2D and 3D maps of the Simulation environment, localizing the robot in that environment and further performing path planning of the robot along with object detection using ROS. Moreover, various algorithms used for creating maps along with detailed internal working of the packages used for path planning are being discussed in this paper.

Swati Shinde, Tanvi Mahajan, Suyash Khachane, Saurabh Kulkarni, Prasad Borle
Robotic Assistant for Medicine and Food Delivery in Healthcare

This paper presents the use of a three-wheel holonomic motion drive system for medicine and food delivery to patients in hospitals. The mechanical design of the holonomic drive is discussed along with the control system used for the robot. The inverse kinematic model of the three-wheel omni drive is analyzed and used for programming the navigation system. This paper analyzes the use of a gyroscope in robot heading and position control. Along with gyroscopes, the use of rotary encoders is also discussed. The encoders are used to keep track of the position of the robot while navigating. Due to the COVID-19 pandemic, mobile robots gained high demand. These robots are used to supply medicines and food to patients and staff, medical equipment required by doctors and nurses, thereby optimizing communication between doctors, hospital staff members and patients and reducing the contact between healthcare staff and patients which is useful in preventing the spread of diseases through direct contact.

Akash Bagade, Aditya Kulkarni, Prachi Nangare, Prajakta Shinde, Santwana Gudadhe
Privacy-Preserving Record Linkage with Block-Chains

We present a method for privacy-preserving record linkage of partially de-identified data. The key novelty of our proposed method is the use of a block-chain to store user information, preventing privacy leaks even if all system databases and cryptographic protocols have been compromised by an adversary, including databases containing identifying information. While satisfying this stringent privacy constraint, the system remains able to deterministically link and retrieve de-identified records to unique identities. With logarithmic time complexity to perform linkage and retrieval operations, our method is highly scalable for large-scale use cases. Designed to be HIPAA (Health Insurance Portability and Accountability Act) compliant by default, such systems are easily adaptable to large real-world healthcare systems.

Apoorva Jain, Nisheeth Srivastava
Performance Analysis of Rectangular QAM Schemes Over Various Fading Channels

There is an increased need for higher data rates within the bounds of existing bandwidth. Quadrature amplitude modulation (QAM) is a much-favored modulation scheme used in wireless communications due to its ability to achieve a high data rate while utilizing existing bandwidth. Especially, the rectangular QAM (RQAM) has received more attention due to its tactical advantages and its usefulness across various communication models. However, as the data rate is increased, the signal becomes more prone to error in the channel. There are several fading channel models used to represent channel conditions in various types of wireless communication links. Here, the radio frequency communication (RF) and optical communication links are taken into consideration. In RF, the Rayleigh, Rician, and Nakagami-m fading models are widely used models and considered in this paper. In optical communication, the log-normal fading model, which is best suited to represent weaker turbulence conditions in the channel, is considered. In this paper, the error probability of the rectangular QAM modulation scheme, subject to the fading channels of RF and optical communication, is analyzed and derived. The results are verified in Python.

Siddhant Bhatnagar, Shivangi Shah, Rachna Sharma
New Symmetric Key Cipher Based on Quasigroup

Stream ciphers that use the XOR function for mixing the plaintext and the keystream are vulnerable to attacks such as known-plaintext attack and insertion attack. To overcome such shortcomings of the existing ciphers, we hereby propose a new stream cipher that uses AES. The proposed cipher is based on a large-order quasigroup. It is resistant to brute force attack, due to the exponential number of quasigroups of its order. It is also analyzed against the chosen-ciphertext, chosen-plaintext and known-plaintext attacks, and it is found to resist these attacks. The output of the cipher is subjected to various statistical tests, such as the NIST-STS test suite, and the results show a high degree of randomness of the ciphertext. Hence, it is resistant to correlation-type attacks.

Umesh Kumar, Aayush Agarwal, V. Ch. Venkaiah
Validate Merchant Server for Secure Payment Using Key Distribution

Nowadays, growing and competitiveness of a financial sector and e-healthcare application developed rapidly and quicker ever so that mobile payment plays a vital role to make easier and quicker. In the existing paper, the cloud transmits the hash value straight to the merchant server without checking for fraud. This paper aims to provide a secure authentication mechanism between the cloud and the merchant server to avoid fraudulent merchant servers. For trustworthiness, our proposed system introduces secure key distribution between cloud and merchant server named as validate merchant server key distribution protocol (VMSKDP). Our proposed system improves security for mobile payments and reduces attacks like man-in-middle attacks, denial-of-service, etc.

A. Saranya, R. Naresh
Extractive Text Summarization Using Feature-Based Unsupervised RBM Method

A methodology for creating shorter and meaningful summaries for single documents is provided. With a lot of content to be had on the web, it’s far simply no longer possible to go through each information source in complete detail. Consequently, a great mechanism is needed to extract relevant information. To overcome these challenges, information in the form of text is summarized with the objective to get relevant knowledge without loss of any information. A methodology for extractive text summarization for single-document summary is devised and developed in this work. It uses a restricted Boltzmann machine to choose essential phrases from the text. The text documents used for summarization are in the English language. Various aspects are used to generate meaningful phrases, and the restricted Boltzmann machine is being utilized to enrich and abstract those features to improve the consequent accuracy without sacrificing any significant information. The sentences are scored, and an extracted summary is created based on those enhanced features. The result indicates that the presented methodology tackles the problem of text overload by producing an appropriate summary. The result of RBM has been compared with the Text Rank, Lex Rank, LSA, and Luhn algorithm. The experimentation is carried out, and the summary is generated for eight different document sets and the result is evaluated using the ROUGE-1 score.

Grishma Sharma, Subhashini Gupta, Deepak Sharma
Depression and Suicide Prediction Using Natural Language Processing and Machine Learning

Depression has always been one of the prominent concerns of mental health worldwide. In the worst-case scenario, someone suffering from depression may lead to drastic measures such as suicide. According to the World Health Organization, depression and anxiety affect one out of every five people worldwide, costing trillions of dollars each year. In the COVID-19 pandemic, the situation has worsened alarmingly as more people suffer from depression. It has become essential, more than ever, toManoj Kumar Gupta Harnain Kour maintain the mental health profiles of our people and to predict any unfortunate event. Depression can be prevented and treated at a very early stage and a low cost, given early detection and identification of the causes. With advancements in machine and deep learning models, it has become possible to identify such behaviour through social interactions such as posts, tweets, and comments. This paper aims to detect user behaviour that can conclude whether a person is suffering from depression and suicidal tendencies based on the user’s social media tweets. The research work proposes a classifier with a hybrid technique in preprocessing using Natural Language Processing (NLP) and machine learning techniques with an accuracy of 75% to identify such traits in a person through his/her tweets.

Harnain Kour, Manoj Kumar Gupta
Automatic Detection of Diabetic Retinopathy on the Edge

The uncontrolled blood sugar levels in diabetes patients lead to an eye disease called diabetic retinopathy. The high sugar levels in the blood vessels of the retina cause blockage of some blood vessels due to which fluids like plasma leak easily into the eye causing the lesions which appear in the eye and may cause severe vision problems. A severe vision problem can be preventedZahid Maqsood Manoj Kumar Gupta by detecting and treating it at an early stage. In India alone, about 80 million people suffer from diabetes, and there is one ophthalmologist for every 100,000 population. Due to this serious shortage of well-trained ophthalmologists, it becomes difficult to diagnose the severity of diabetic retinopathy in some rural areas of India. Since most of the AI solutions for detecting diabetic retinopathy are cloud-based, therefore, it becomes difficult to deploy these frameworks in rural areas where there is no connectivity and no proper Internet connection. This paper focuses on energy-efficient and real-time detection of the severity of diabetic retinopathy on the low-powered edge device without any proper connectivity. In this paper, various deep transfer learning methods were investigated for DR detection, and these include ResNet50, Inceptionv3, EfficientNet-B5, EfficientNet-B6, and VGG19. These CNN models were trained on preprocessed APTOS dataset. To increase training data and to overcome overfitting, various data augmentation techniques were used. The highest accuracy of 86.03% was achieved by EfficientNet-B6.

Zahid Maqsood, Manoj Kumar Gupta
A Survey on IoT Security: Security Threads and Analysis of Botnet Attacks Over IoT and Avoidance

IoT is an emerging technology that provides humans very handy support in various aspects and applications. This technology faces various threats in various aspects. The proposed work will analyze the various levels of threatsM Vijayakumar T.S. Shiny Angel and combating the threats. Various levels of threats are identified to the IoT. Over twenty-five, different levels of threats are identified for the IoT in different aspects. As the IoT is an emerging technology, it has to overcome these hurdles. In this paper, a nitty dirty review of the security-related challenges and wellsprings of peril in IoT applications is presented. Within the wake of talking around the security issues, diverse emerging and existing developments focused on finishing, and also, mainly Botnets-based threats feature over IoT is been provided solution as it is most vulnerable comparing other threats. Combating features are recommended.

M. Vijayakumar, T. S. Shiny Angel
A Coherent Approach to Analyze Sentiment of Cryptocurrency

In this paper, we have tried to analyze the real-time Twitter data of some popular cryptocurrencies, like Bitcoin. Bitcoin has been by far the largest cryptocurrency in terms of market size. Its market capitalization currently sits at over 1 Trillion US dollars. Even though it has gained popularity during this decade, still the cryptocurrency has seen many significant price swings on both daily as well as long-term valuations. In recent times, the influence of social media platforms like Twitter can be seen on cryptocurrency as well. Twitter is being used as a news sourceAyush Hans Kunal Ravindra Mohadikar Ekansh for many users who need to buy or sell Bitcoin. Therefore, understanding the sentiment behind the tweets which have a direct impact on price direction can help the user to trade in cryptocurrency better. The real-time data extracted using the APIs can help the user get the tweet sentiment at the time of purchase which can provide him/her an advantage in buying/selling the cryptocurrency. By combining the commonly used sentiment analysis tools like VADER and TextBlob into a single model, we can get a more accurate sentiment analysis of the tweets.

Ayush Hans, Kunal Ravindra Mohadikar, Ekansh
Supervised Machine Learning Algorithms Based on Classification for Detection of Distributed Denial of Service Attacks in SDN-Enabled Cloud Computing

Software-defined network (SDN) is a networking technology that separates the data and control planes and enables centralized network control. This technique encapsulates lower-level functionality, allowing network managers to configure, manage, and regulate network behavior. While centralized monitoring is an important benefit of SDNAnupama Mishra Neena Gupta, it can also be a serious security risk. The attacker gains access to the entire system if he successfully penetrates the central controller. Therefore, integration of SDN with the cloud itself provides insecurity to the cloud consumers. To skillfully implement DDoS over the controller, an attacker must gain access to multiple systems to launch multiple DDoS attacks. These DDoS attacks will deplete the controller’s resources, causing its services to be unavailable. In order to detect controller attacks early on, it is critical to expand the coverage of the network. There are many existing techniques. However, relatively little research has been done in the area of SDN. A number of solutions fall under this category, including the use of machine learning algorithms for the task of classifying connections as either valid or invalid. To detect suspicious traffics and connections, we employ classification supervised machine learning algorithms, the Naive Bayes and support vector machine (SVM), which also achieved a promising result in order to verify the proposed work.

Anupama Mishra, Neena Gupta
Edge Computing-Based DDoS Attack Detection for Intelligent Transportation Systems

Vehicular ad hoc networks (VANETs) are a critical component of intelligent transportation systems (ITS). Because VANET allows the transmission of critical and life-saving information between vehicle nodes, any effort to compromise the network should be recognized immediately, if at all feasible. The distributed denial-of-service (DDoS) assault is one kind of cyber-attack that affects VANET systems’ availability.Akshat Gaurav B.B. Gupta Kwok Tai Chui As a consequence of the DDoS assault, vehicle nodes are unable to transmit vital information. In this context, this experiment proposed edge computing-based DDoS detection techniques. The proposed technique uses packet entropy to distinguish DDoS attack traffic from normal communication. To determine the entropy values, we performed an in-depth study of five different machine learning methods. precision, accuracy, f1 measure, and recall are used to assess the performance of different machine learning methods

Akshat Gaurav, B. B. Gupta, Kwok Tai Chui
An Empirical Study of Secure and Complex Variants of RSA Scheme

In today’s cyberRaza Imam Faisal Anwer space, where large amounts of data are being exchanged and stored on remote storages, Cryptography plays a major role. Public key cryptography such as RSA is one its effective type, which uses two keys, one for encryption and one for decryption. Concerning the recent advancements in the domain of cryptography, many cryptographers have proposed various extended and enhanced form of RSA algorithms in order to improve the reliability and efficiency of the information security world. In this paper, we studied several extended forms of RSA algorithm. We implemented several and compare them in terms of its efficiency in terms of key generation, encryption and decryption time. Finally, we suggested a multipoint parallel RSA scheme to improve the overall algorithm execution speed compared to standard RSA. This method will also prove to be computationally less costly and more secure as compared to standard RSA.

Raza Imam, Faisal Anwer
Text Normalization Through Neural Models in Generating Text Summary for Various Speech Synthesis Applications

Machine learningP N Varalakshmi K Jagadish S Kallimani, like neural network methods, has been implemented in the natural language processing of virtually every domain. With speech utilizations like vocabulary to speech organization, coalescence challenge that has been adequately immune directly toward successful machine attainments learning approaches is fundamentals normalization. Considering example, in this application it must be determined that 123 is verbalized in signatures as one hundred and twenty-three but in sovereign potentate Ave as one twenty-three. Modern industrial systems for this role are heavily dependent on hand-recorded penned double speak-specific stratification. We introduce neural interconnection miniatures well-known regard text notarization for as a streak to progression problem, where input is admission taking in token in history, and turn out gain would be that token’s verbalization.

P. N. K. Varalakshmi, Jagadish S. Kallimani
Classification of Network Intrusion Detection System Using Deep Learning

Over the pastNeha Sharma Narendra Singh Yadav one decade, there has been a continuous rise in the usage of Internet services all over the world. However, numerous challenges emerge since malicious attacks are constantly changing and are happening in exceptionally huge volumes requiring an adaptable solution. This has led to a desperate need not only of detection and classification of attacks at host as well as network side but also the detection being automatic and in a certain time frame, as a result of which the world has seen many developments in this field with machine learning and deep learning playing a huge role in it. Because of the dynamic effect of malware with constantly changing attack techniques, the malware datasets accessible openly are to be updated efficiently and benchmarked. In order to develop an effective intrusion detection system, machine learning or deep learning techniques are also becoming more advanced day by day, and it is important to utilize their benefits in this field. This paper focuses on the development of network intrusion detection systems (NIDS) using deep learning. This paper uses UNSW-NB15 dataset as it is one of the most recent and improved IDS datasets. It has been improved on many factors from its predecessor KDD CUP99. Convolutional neural network and recurrent neural network have been implemented to compare the results. The classifications implemented in this paper are both in binary and multiclass with the major focus regarding maximum macro precision, recall, and f-score for the multiclass approach.

Neha Sharma, Narendra Singh Yadav
Toward Big Data Various Challenges and Trending Applications

WithBina Kotiyal Heman Pathak the continuous growth in the data and its use as a resource for analytic knowledge continues to help companies develop, the need for innovative approaches, tools, and strategies to extract actionable insights is becoming increasingly important. The data collected from myriad sources like social media, search engines, and the Internet of Things has developed substantial opportunities concerning the business to business industrial organizations. Big data (BD) computing is classified as batch and stream computing based on the processing types. Batch computing is performed when data is at rest, whereas real-time computing is performed when data is in motion. In the present era, real-time stream processing is in demand as the massive data generated has to be handled speedily to meet the business or organization requirements. BD analytics is used to get the big insight from this data. However, cleansing, interpreting, and analyzing such massive databases present hurdles in marketing, particularly in terms of making real-time decisions. This paper throws light on the various issues and challenges associated to BD. Most of the challenges are associated to the preprocessing phase of BD. It also presents the diverse applications of BD.

Bina Kotiyal, Heman Pathak
Convolutional Neural Network-Based Approach to Detect COVID-19 from Chest X-Ray Images

COVID-19 is a worldwide pandemic that poses serious health hazards. COVID-19’s diagnostic test sensitivity is restricted owing to specimen processing abnormalities. The discussed technique might be used in clinical practice as a computer-aided diagnostics approach for COVID-19. The use of chest X-ray pictures for detection is life-saving for both patients and clinicians. Furthermore, in nations where laboratory kits for testing are unavailable, this becomes even more critical. This work aims to demonstrate the application of deep learning for high-accuracy COVID-19 identification utilizing chest X-ray images. Image-based applications have reached a pinnacle in the last five years thanks to the widespread usage of convolutional neural networks (CNNs). CNN gathers information from images by extracting features. The enormous popularity and efficacy of CNNs have sparked a new rise in interest in deep learning. The image data space is littered with CNN models. They excel in computer vision tasks like image categorization, object identification, and image recognition. This research work attempts to discuss the CNN-based approach for detecting COVID-19 from chest X-ray images.

P. Pandiaraja, K. Muthumanickam
Classification of Medical Health Records Using Convolutional Neural Networks for Optimal Diagnosis

Pneumonia is considered to be one of the lungs affecting inflammation for small air sacs. Dry cough, chest pain, fever, and breathing difficulty are some of the common symptoms during this situation. The seriousness of the condition of the patient is variable based on several parameters. Viruses, bacteria, and by other microorganisms usually cause pneumonia. Some of the risk factors during this situation are: cystic fibrosis, chronic obstructive pulmonary disease (COPD), asthma, diabetes, and heart failure. Sometimes a weak immune system may also increase the severity of the situation. The medical diagnostics using machine learning powered by computer vision and deep learning will help us to extract useful information by filtering out the non-essential and insignificant information from the diagnosis report. Computer vision, neural networks, and artificial intelligence methods like convolutional neural network will lead to identify and extract the useful information from the diagnosis report, and in turn, it will help to assist in medical diagnosis. In this regard, the main objective of this work is to classify disease based on symptoms. Clinical and laboratory symptoms are considered as the basic for this investigation.

M. H. Chaithra, S. Vagdevi
Smart Farming Using IoT Sensors

Agriculture is the backbone of India. Agriculture is the sector where water usage is more with irrigation accounting 75% of global water usage. If we don’t make any improvement in efficiency of usage of water, it is expected that usage of water for agriculture will increase by 20%. As population is increasing day-by-day, agriculture is becoming more important factor as it feeds many of the lives and we should be blessed for that. In the traditional approach, farmer will not get to know how much amount of water does the plant needs, so plants will not get required amount of water because of which we may end up in either more or less water supplied to the plants. In this research work, by using smart farming technology using Internet of Things (IoT) we will get to know soil moisture level, according to the soil moisture level the farmer can supply required amount of water through water pump. By this work we can overcome two extremes of the problem either reducing the wastage or too much consumption of water. Once water resource is used efficiently, the next stage is to check whether the banana plant is healthy or not. The banana plant health checkup is monitored through its leaf. Here, we are checking whether the leaf is diseased or not using SVM algorithm. From this research work, we are saving the water resource and plant health state in its initial stage. Thus, we have improved the efficiency of usage of water to the banana plant.

J. Y. Srikrishna, J. Sangeetha
Securing the Smart Devices in Home Automation System

Security is the major concern in every infrastructure such as offices, banks, hospitals, etc. Due to lack of security in the existing home automation system, hackers can easily access and collapse the system. Hence, in this research work, we are providing security to the remotely controlled home infrastructure and to reduce the energy consumption of smart devices. The smart device consists of the appliances such as four lights and two switches operated using the web-based application. This application allows only authorized users to remove these smart devices which are not in use for longer time and also the hacked devices. Thus, we are reducing energy consumption and securing our smart devices from hackers. Adding to the benefit of the user, through this application we can change the status of the devices either to on or off state. The status of the devices is stored in the cloud in the encrypted form (ciphertext) using the encryption technique such as the AES algorithm. Through the server using Wi-Fi module, this ciphertext can be accessed by the user credentials and decrypt it through web-based application. In this research work, major security issues like authentication and verification are taken care, and this work focus is on reduction of energy consumption by removing unnecessary devices and protecting the smart devices from hackers in the existing home automation system.

Syeda Sabah Sultana, J. Sangeetha
Dual-Channel Convolutional Recurrent Networks for Session-Based Recommendation

Jingjing Wang Lap-Kei Lee Nga-In WuRecommender systems assist a Web application user in satisfying their needs or interests based on the user profile and past activities. Yet due to privacy and other concerns, some applications and services only keep anonymous information. A session-based recommender system (SRS) predicts the next item by exploring only anonymous user-item behavior orders during ongoing sessions. Recurrent neural networks (RNNs) and their two variants have dominated the research on SRS. However, there are two shortcomings in these RNN-based methods: (1) RNNs easily generate false dependencies because RNNs assume all adjacent items are highly dependent on each other; (2) the sequentially connected architecture of RNNs can only capture the point-level dependencies but ignoring neglecting the union-level dependencies. This paper proposes a Dual-channel Convolutional Recurrent Neural Network (D-CRNN) model to address these problems. This hybrid model leverages RNN to explore complex long-term dependencies and combines CNN to extract the union-level context features, which help to reduce the noise. The hybrid model was evaluated on three commonly used real-world datasets. The experimental results on Diginetica dataset D-CRNN showed an improvement of 5.8% and 4.8% respectively in terms of Recall@10 and MRR@10, demonstrating the effectiveness of D-CRNN on the session-based recommendation.

Jingjing Wang, Lap-Kei Lee, Nga-In Wu
Reuse Your Old Smartphone: Automatic Surveillance Camera Application

The life cycleLap-Kei Lee Ringo Pok-Man Leung Nga-In Wu of a smartphone is decreasing rapidly, which leads to a lot of electronic waste and causes damages to the environment. Home security has also become an important concern due to the frequent occurrences of burglaries and home accidents. Surveillance cameras are commonly deployed to improve home security. This paper presents the design of an automatic surveillance camera application called iEye. It aims to reuse components of an old smartphone including the camera, sensors, and microphone to automatically monitor users’ home security. iEye is a cross-platform application, which contains a Java application as a server, and an Android app installed on old smartphones as cameras, and installed on the user’s smartphone or personal computer as a viewer, respectively. It turns an old smartphone into a home security camera, supporting real-time streaming such that users can see live video of their home on their smartphones anywhere and anytime. iEye also allows users to define abnormal events based on its detection components, namely motion detection, face recognition, human detection, and light detection. Upon an abnormal event, the camera video will be recorded and the user will also be immediately notified by email. iEye is particularly suitable for solitary elderly, working parents, and pet keepers. It also helps to reduce electronic waste.

Lap-Kei Lee, Ringo Pok-Man Leung, Nga-In Wu
A Model of UAV-Based Waste Monitoring System for Urban Areas

This paperDalibor Dobrilovic Gordana Jotanovic Aleksandar Stjepanovic Goran Jausevac Dragan Perakovic presents an approach in using unmanned aerial vehicles (UAVs) with remote imaging for urban waste monitoring. The system is designed to monitor green areas, public trash cans, and unregulated landfills and to detect possible violations of garbage disposal rules. Public green urban areas, such as parks, green surfaces, sport terrains, and bathing areas, are gathering places for people and therefore prone to unregulated waste disposal. The proposed solution describes the real-time monitoring of the area using drones and the detection of irregularities in a garbage disposal. The cameras mounted on drones are used to take images of public targeted areas at pre-mapped points. Visual data collected by supervisor drones are used for further processing and notification of authorized personnel and institutions.

Dalibor Dobrilovic, Gordana Jotanovic, Aleksandar Stjepanovic, Goran Jausevac, Dragan Perakovic
A Secure Multicontroller SDN Blockchain Model for IoT Infrastructure

IoTK.Janani Dr.S.Ramamoorthy is making significant progress in a variety of fields, including health care, smart grids, supply chain, and so on. It also makes people’s daily lives easier and improves their interactions with one another and their surroundings and environment. There is a variety of research on decentralized computing for IoT develops a decentralized IoT-based biometric facial recognition solution for COVID-19 lockdown cities. They propose a three-layer architecture (application layer, control layer, and data layer) and then create a blockchain framework on top of it to entirely restrict public movements. The software-defined network is the most widely utilized solution for establishing secure network interaction and building secure IoT infrastructures. They give a solid and dependable framework for dealing with dangers and issues like security, scalability, and confidentiality. This study provides a blockchain-based software-defined IoT framework for smart networks that are optimized for energy efficiency and security. Indeed, multicontroller SDN blockchain (MC-SDNBC) has been extensively used to manage vast-scale networks which are, though, subject to a variety of attacks, include false data injection, which causes regulator topology inconsistencies. Every software definition network domain is administered with a single master controller who communicates with both the masters of the other Internet via blockchain. The controller unit generates blocks of dynamic network modifications, which are subsequently evaluated by redundant controllers using a reputation technique given by the control system. The popularity system uses continuous and coupled reactive fading repute algorithms to score the controllers, for example, the voter’s maker and block, during each voting activity. The analysis findings show that false flow rule insertion may be detected quickly and efficiently, keeping more secured IoT Systems.

K. Janani, S. Ramamoorthy
A Recent Survey on Cybercrime and Its Defensive Mechanism

CybercrimeGarima Bajaj Saurabh Tailwal Anupama Mishra is one of the severe issues in today’s world that is increasing day by day due to unawareness of people about the harm it can cause. The main reason against the augmentation of cybercrime is the lack of education or knowledge about the impact it can lead to. Cybercrime can be done against individuals, society, or any organization whether it is private or government. The aim of the paper is to focus on what cybercrime is, its types, related work, and its defensive mechanism. Defensive mechanism against cybercrime includes the ways or measures that how can any individual or any organization protects them against cybercrime. This paper also includes the related work which includes some points about the work which has been done so far on cybercrime.

Garima Bajaj, Saurabh Tailwal, Anupama Mishra
A Hybrid Feature Selection Approach-Based Android Malware Detection Framework Using Machine Learning Techniques

With moreSantosh K. Smmarwar Govind P. Gupta Sanjay Kumar popularity and advancement in Internet-based services, the use of the Android smartphone has been increasing very rapidly. The tremendous popularity of using the Android operating system has attracted malware attacks on these devices. Detecting variants of malware features that change their behavior to hide from being detected by the traditional method of machine learning is being an incapable and challenging task. To overcome these issues of malware feature detection, an efficient feature selection plays a crucial role in detecting malware features and reduces the dimensionality of a huge dataset and removes the unnecessary features that are not useful and keeps those relevant features that improve the classification accuracy and detection rate. To address the above issues, this paper proposed a novel framework in which a hybrid feature selection using wrapping feature selection (WFS) with the combination of random forest and greedy stepwise (RF-GreedySW) framework is devised to optimize the malware features. The proposed framework is capable of reducing a large number of attributes into an optimal feature to enhance the performance of the machine learning model. The framework used the three most popular ML classifiers such as random forest (RF), decision tree (C5.0), and support vector machine radial basis function (SVM RBF). The performance of the proposed framework is evaluated using the CIC-InvesAndMal2019 dataset. The DT (C5.0), RF, and SVM RBF model achieves better accuracy of 91.80%, 91.32%, and 82.33% on static layer, respectively. Similarly, the accuracy is 72.41%, 75.10%, and 62.07% on the dynamic layer by DT (C5.0), RF, and SVM RBF, respectively. Our model highlights good results on the CIC-InvesAndMal2019 dataset in terms of classification accuracy and increases the robustness of the model.

Santosh K. Smmarwar, Govind P. Gupta, Sanjay Kumar
Security of Big Data: Threats and Different Approaches Towards Big Data Security

In theYashi Chaudhary Dr. Heman Pathak present era, the use of the Internet has extended abruptly. With this abrupt increase, massive data is being created, resulting in big data. Big data means more diverse, more impetus, and more complex data streams. Data is being produced in abundance in exabytes and zettabytes by electronic devices, power grids, and modern software. This big data brings different challenges such as incompleteness, inconsistency, heterogeneity, and security with itself. The presented paper targets the security challenge as it is a very significant feature overseen by various data analysts; thus, data must be secured from dwindling in the wrong hands. This paper discusses the approaches and mechanisms mainly based on anonymization, access control, and encryption.

Yashi Chaudhary, Heman Pathak
Segmentation of Image Using Hybrid K-means Algorithm

Image segmentation is a crucial step to recognizing an object. During the segmentation process, pixels in an image are categorized based on their gray color. In pixel classifications, the K-means clustering algorithm is commonly used. In this approach, the centroid of the segment was measured using arithmetic mean and Euclidean distance. In the proposed paper, the centroid was updated using the hybridization of harmonic and arithmetic means. The proposed algorithm makes use of the harmonic and arithmetic mean features. The experimental results are compared to conventional K-means and harmonic K-means algorithms, demonstrating that the proposed algorithm performs better when checking segmentation consistency.

Roopa Kumari, Neena Gupta
A Chatbot for Promoting Cybersecurity Awareness

Cybersecurity is one of the hot topics nowadays. However, not many Internet users know the cyber risks around them. To promote cybersecurity awareness, this paper presents a chatbot that is a cybersecurity expert. It aims to let its users learn more about cybersecurity. It relies on Google Dialogflow, which is a natural language understanding platform. Our chatbot contains a knowledge base on cybersecurity knowledge. Users can make queries to the chatbot to learn definitions and concepts of different cybersecurity terms. Our chatbot also provides self-quizzes for users to test their knowledge on different cybersecurity topics. It also provides suggestions to users on cybersecurity issues such as how to identify and handle phishing emails. In a survey of twenty users, the majority of the users agreed that our chatbot is easy to use and can increase their awareness of cybersecurity issues.

Yin-Chun Fung, Lap-Kei Lee
An Advanced Irrigation System Using Cloud-Based IoT Platform ThingSpeak

The conventional irrigation systems are manual and thus require human effort and interruption to water the crops, and in such a scenario, water wastage is evident. This work was designed to address these two problems associated with conventional irrigation systems, i.e. manual operation and water wastage. The system designed has been made smart and thus has automatic decision-making ability to reduce human effort and interruption. The designed system has made use of sensors like soil moisture sensor, temperature and humidity sensor, and rain sensor and thus can calculate the moisture content of the soil, read surrounding temperature and humidity, and sense rainfall. It has the feature of making an application programming interface (API) weather call to extract information about rainfall from the OpenWeather webpage and finally makes a decision comparing all collected data and threshold data already set by the user. ThingSpeak, a cloud-based Internet of Things (IoT) platform, has been used for storing the data read by various sensors in the form of graphs for better visualization and future reference. A (Global System for Mobile Communication-Global Positioning System) GSM-GPS module is also taken into work for establishing Internet connection and determining the system location for precise weather data. The system was tested for different threshold values of soil moisture and temperature reading, and based on the comparison with real-time sensor values, it successfully turned ON/OFF the motor and thus found to work fine as desired. The weather data fetched by the system also found to match with the real-world weather conditions.

Salman Ashraf, A. Chowdhury
Backmatter
Metadata
Title
Cyber Security, Privacy and Networking
Editors
Prof. Dharma P. Agrawal
Dr. Nadia Nedjah
Dr. B. B. Gupta
Gregorio Martinez Perez
Copyright Year
2022
Publisher
Springer Nature Singapore
Electronic ISBN
978-981-16-8664-1
Print ISBN
978-981-16-8663-4
DOI
https://doi.org/10.1007/978-981-16-8664-1