Intelligent Cyber Physical Systems and Internet of Things

In today's world, when the internet is pervasive, everyone gets news from a variety of online sources. As the use of social media platforms has grown, news has travelled quickly among thousands of people in a very less duration. The propagation has been far reaching for the fake news generation in repercussions, from altering election outcomes in support of specific politicians, creating prejudiced viewpoints. Furthermore, spammers use appealing news headlines to make cash through click-bait adverts. In today’s world knowingly or unknowingly fake news spreads around the world through internet. This has a great impact on the people who blindly believe whatever the internet provides. Hence, fake news identification has become a new study subject that is attracting a lot of attention. However, due to a lack of resources, such as datasets and processing and analysis procedures, it encounters several difficulties. This research uses a non-probabilistic machine learning models of computational prototypes to address this problem. Furthermore, the comparison of Term Frequency-Inverse Document Frequency (TF-IDF) is done, for the purpose of determining the best vectorizer used for detecting fake news. In order to raise the accuracy, stop words of English are used. To predict bogus news, a Support Vector Machine (SVM) classifier is deployed. According to the simulation data, the SVM and the TF-IDF produce results with high accuracy.

Aquaculture comprises an important part of agriculture which has been the pillar in the primary sector of India. Many factors such as temperature, natural calamities and artificial chemicals when monitored irregularly can severely impact the survival of the species. There is more time and energy invested by various scientists across the world to monitor these species in an efficient manner. This is where the idea of an utopian system is initiated to monitor various parameters such as temperature, turbidity, pH and much more through a union of various sensors connected to a microcontroller. The principles of Internet of things and machine learning assist the system to estimate the probability of survival of the species by using the data collected from the sensors. The data which is gathered from the sensors is predicted using the services of IBM and is visualized through a web application that helps the user to interpret data in a coherent manner.

The advancement in technology and the rising demand for the internet have witnessed an increased cyber threat which can be highlighted as a major challenge. Cyber criminals make use of malware programmes to fulfil the malicious purpose. Malware is software coded and designed to bring harm to the target machine by various means. This paper focuses on describing memory analysis techniques under malware analysis with a comparative analysis of different tools. The present paper highlights the application to the case study of CRIDEX malware for a better understanding through a practical implementation of appropriate tools. With its initial appearance as CRIDEX, the bank stealing trojan has evolved in the past decade and has been witnessed to spread malware infection through its new variants as discovered in the past year for which it has been selected as an appropriate candidate to be analysed for understanding its basic working.

Theft of the world’s most valuable trees, such as sandalwood, lumber, teak, rosewood, and pinewood, presents a huge danger to forest resources, causes substantial economic harm, and has a terrible impact on the ecosystem across the globe. These trees are very pricey and scarce around the globe. These trees are employed in both medicinal and cosmetics research. To stop such smuggling and conserve the world’s forests, various preventative measures must be created. Many incidences of tree cutting occur as a result of the higher amount of money involved in selling such trees. This study presents an anti-poaching system based on IoT and WSN technologies. Three sensors are used in the structural framework: a tilt detector (to detect the tendency of a tree while it is being cut), a fire sensor and a smoke detector (to detect timberland fires), and a sound detector Detection of even the sound of a tree being hacked down may be used to catch illegal loggers, for example WSN technology is commonly employed in remote monitoring applications (where monitoring is difficult).

Recognizing human activities plays a substantial role in human-to-human and human-to-computer interactions. Recognizing human activities from video sequences or pictures is a difficult task because of troubles, such as history clutter, partial occlusion, modifications in scale, viewpoint, lights and look. Human action is difficult to classify as a time series. Predicting a person’s movements is a part of this. In this paper, the KTH video dataset is used for designing the system. Feature extraction methods like optical flow and spatiotemporal techniques are being utilized to extract the features. Triple stacked autoencoders are used for clusterization to reduce the data dimensions. An efficient BoW vector feature extraction method is used for extracting text data, by which data is obtained for training the model. A deep learning algorithm such as VGG19 is used to determine and classify the activities of a human. The objective of this efficient model is to apply as an ATM surveillance as a camera module fixed in the room to perform constant surveillance. The Police department can have an mobile application through which they can monitor and desist any unwanted human activities happening in the ATM.

There is no universally accepted definition of terrorism. Terrorism and its ramifications have every once in a while caused massive death and destruction around the world. Current cutting-edge technologies, such as machine learning and deep learning, can predict and classify such attacks efficiently. The major difficulties observed in implementing these strategies are a lack of consistent and clean data, as well as programming knowledge in Python and R. Inconsistent data can be resolved by incorporating graph database features into the dataset, and Python programming can be replaced with the orange data mining tool. As a part of data processing and manipulation software, orange data mining tool employs a machine learning model in a non-coding context. This research study has attempted to replicate the results by using the orange tool and Neo4j Sandbox. In this study, a non-coding approach was used to classify terror attacks by using the orange data mining tool, and the use of graph embeddings as dataset features have assisted in eliminating the problems associated with inconsistent data. The dataset was then subjected to machine learning techniques such as Random Forest, Decision Tree, Support Vector Machine, Naive Bayes, Gradient Boosting, KNN, and Adaboost to classify the terror attacks. Random Forest and Gradient Boosting are the models that can achieve an accuracy score, recall, precision, and F1 score greater than 90%.

The COVID 19 pandemic has given rise to a new normal. This includes wearing masks and maintaining social distance. Nowadays sudents don’t focus in offline classes. Also, students with masks in offline proctored exams find ways to roll their eyes at others’ work for malpractice. The systems designed to date are not accurate to detect facial features with mask. These problems have motivated us to develop a reliable, robust model to detect mask, eye location, eyeball location, eye status, and head pose of people wearing and not wearing a mask, all at once. We have used 3800 masked, unmasked images to train our model using MobileNetV2, a convolutional neural network, with 99% accuracy. The output of this model is processed using image processing, facial landmark analysis, EAR, and deep learning to detect the facial landmarks accurately. Ultimately, a unique method is used to detect head pose of person.

The aim of the design is to implement Artificial Intelligence in the Healthcare domain and find suitable results. This work is an AI based web application which enables four features 1. Medicine Prescriber 2. Diabetic Analyzer 3. Covid-19 Predictor and 4. AI based chatbot. Provide online solution for the patients like prescribing medicine based on the symptoms, Analyzing the blood sugar (Mg/Dl) and suggesting food diet based on the age and sugar level, Predicting Covid-19 Positive or negative according to the X-ray of the chest and AI based Chatbot which acts as a support agent guides the user in using this application and tells interesting facts on Covid-19. The UI design of the web application is crafted using Adobe XD. Machine learning and Deep learning techniques are used to predict results for these features, these Machine learning and Deep Learning models are deployed as Web application using a framework called Flask. IBM Watson Assistant which is used to create the chatbot, allows you to integrate conversational interfaces into any app, device, or medium, as well as add a natural language interface to the app to automating conversations with your customers.

In this paper, two algorithms are compared based on their time complexity. The time complexity is defined by encryption and decryption of different message lengths. Time varies for different lengths of messages. We statistically analyzed the time complexity of the algorithm and compared their results.

Internet of Things is a domain which has gained quite a lot of momentum in the past decade. The ability of things communicaiting via a network or the internet is both a fascinating and a challenging concept. The authors in this study aim to introduce users to the concept of Internet of Things and demonstrate a use-case wherein data can be uploaded to the cloud; specifically, here the Thingspeak cloud is used as a storage for the data that is uploaded from an ultrasonic sensor via a microcontroller NodeMCU and WiFi standard. This usecase will help new users understand the scope and capability of communication in the Internet of Things. The exeprimental setup shows successful communication between nodeMCU and thingspeak cloud through the internet.

Cloud computing (CC) is evolving computing model with a vast array of heterogeneous autonomous systems by modular computational architecture. Load balancing of activities on the cloud environment is an essential part of distributing services from the data center. CC is agonized by overloading demands because of dynamic computing through the internet. Load balancing must be done to ensure maximum use of the resources in all virtual machines (VM). Task scheduling is a crucial step for improving cloud computing's overall efficacy. Task scheduling is therefore significant to minimize energy usage and increase service providers’ benefit by reducing the time required. This work provides a detailed study about the cloud computing architecture, load balancing (LB) mechanism, task scheduling (TS) framework in the cloud environment. Various meta-heuristic optimization techniques have been implemented to manage the load over virtual machines using task scheduling and load balancing terminologies. Various research gaps and issues have been identified from the literary work done by various researchers. This comprehensive study has motivated and provided us future direction to do work in this field.

Nature is the one of the best inspiration for solving problems around us. Nature inspired computing algorithms are inspired from nature. These algorithms have inbuilt features of self learning, self motivation, co-ordination and collective behavior for solving a particular task. These nature inspired computing algorithms are specified by local search (exploitation) and global search (exploration). As per the effects of nature inspired computing algorithms, depends the searching capability. This paper focuses balance of local and global searching attribute of nature inspired computing algorithms. The people in a group are managed by leaders. By applying PSO algorithm, the best leader in the group is selected using efficient fitness functions. And this proposed fitness function is compared with the benchmark fitness function for providing better optimum results in the balancing of exploration and exploitation.

Organ transplantation is one of the most effective medical procedures to save lives. An individual’s organs can save up to nine lives. However, individuals refuse to donate organs because of lack of awareness and trust in the procedure, leading to the reduction in the number of organ donors. Individuals who wish to donate organs have to go through a complex administrative process, and sometimes these donated organs are managed by unauthorised individuals. To encourage individuals who wish to donate organs, we need a secure, efficient, and scalable platform. In this article, we present our perspective on the blockchain based organ donation management, in particular, for organ donation between organ donors and patients. The proposed platform uses the smart contract to automate the organ donation process and reduces the overall time of organ donation process. The proposed blockchain-based Organ Donation Platform (ODP) help patients in finding a matching donor efficiently. The ODP facilitates the process of organ donation by a decentralized network ensuring security, integrity, and transparency that eliminates the intermediaries. We comparatively evaluate the performance of the proposed ODP with the state-of-the-art literature. The proposed ODP is not only secure and scalable, but also efficient and reliable to find matching donor without revealing their identities.

In digital image processing image enhancement techniques are used to improve the observation, perception and interpretability of the image information through human visual system. Image enhancement is implemented in spatial domain approach that operates directly to the pixels and frequency domain approach that operates through Fourier transform of an image. Image enhancement techniques in frequency domain are useful in different fields like early detection of physiological disorder, remote sensing, forensic science and biometric science. This implementation is done in frequency domain approach that means sharpening, smoothing and homomorphic filters are designed, implemented and image is enhanced to give the better input to the image processing automated techniques like biomedical, recognition and matching applications. In this research work, frequency domain Fourier transform techniques are designed and implemented for fingerprint detection and matching social applications. The frequency domain analysis performance is measured with performance measures which are peak signal to noise ratio and contrast to noise ratio with mean and variance. The fingerprint detection and matching progression technique is usually disintegrated into image pre-processing, matching and extraction. The designed, implemented and presented results of this research work will be very useful in social forensic science fingerprint matching different applications to improve the accuracy before applying to matching process.

With the help of machine learning algorithms including KNN, SVM and LDA; it is possible to determine which crops are to be grown, when the soil needs more water and fertilizers, and what pests are present in the crops. The proposed system is designed to collect the current soil conditions and to calculate soil nutrients by analyzing the current soil conditions. In the proposed model, SVM offers 100% precision and LDA offers 95% accuracy in soil prediction. IoT camera sensor modules will assist farmers in determining whether their crops have been infected with pests so that they can take appropriate measures. In the application, the farmer can get alert notifications as well as other relevant information about crops such as soil type, crop types, nitrogen, potassium, and phosphorus, based on the soil and weather conditions. Farmers can also determine which crops to plant based on the soil and weather conditions. This empowers the farmer to take appropriate action for minimizing crop loss and maximizing crop yield.

In the last decade, the İnternet Of Things(IoT) increased dramatically until it became an integral part of our daily lives. These devices are interconnected to the internet without the need for human intervention. Due to the weak configuration and unique characteristics of the internet of things has become a robust target for cyber-attack that worry the general user of these devices. Furthermore, IoT security challenges are increasing day by day and are subject to a variety of attacks. The traditional security measures, such as authentication, access control, network security, and encryption, for IoT devices and their vulnerabilities, are insufficient, ineffective, and cannot process these issues. Existing security methods must be improved to protect the IoT environment. ML/DL provided many solutions that assisted solve the challenges of the IoT and provided safety for it. The goal of this paper is to provide a study on the attacks in IoT architectures such as the sensing layer, network layer, and application layer, then present ML and DL that contributed to the solution in attack detection. In addition, we discuss the challenges of IoT architectures.

A low-power, energy-efficient circuit is essential for IoT edge devices, which increasingly perform data-intensive applications. Nanometer technology nodes push standard CMOS to its limits, which include increased leakage and increased power consumption. Appropriate algorithms for low-power circuits include adiabatic logic and approximation computing. It is possible to construct circuits that are more energy efficient by using adiabatic logic. The adiabatic logic's dual-rail construction and power clock approach, on the other hand, increase the overall footprint. More power is conserved by lowering the circuit's complexity and size while utilising approximation computing. For the Internet of Things (IoT), energy efficiency, and security, adiabatic circuits have the potential to work together. IoT-RF-powered devices can benefit greatly from adiabatic circuits even though they have been around for more than six decades, as demonstrated by some of the recent advancements. These enhancements are described in detail, with an emphasis on the main design challenges and opportunities associated with adiabatic circuits.

Agriculture is the backbone of our society; hence, the Indian economy is heavily reliant on farmers. Farmers are in charge of crop cultivation and account for around 51% of all agricultural production. Despite their contributions, individuals do not realise the benefits or earn sufficient profit for a variety of reasons, including a lack of understanding and supply chain management inefficiency. Smart technologies that require supply chain management models are used to tackle these difficulties. This aids in the financial transaction's monitoring at each stage. Blockchain technology has recently emerged as a transparent supply chain management platform. The goal of this study is to show how various supply chain management systems can track their transparency. We show the various problems with the current system in this review for a traceable transaction that can help farmers in tracing the financial transaction. Furthermore, AI (Artificial Intelligence) is suggested for future research direction.

Many applications, like crossing points, banking, and mobile banking, are now using Face Recognition (FR) systems. The widespread usage of FR systems has heightened concerns about the security of face biometrics against spoofing assaults, in which a picture or video of a valid user's face is employed to attain unauthorized access to resources or activities. Even though numerous FAS or liveness detection techniques (which identify if a face is live or spoofed at the moment of acquisition) have been developed, the problem remains unsolved because of the complexity of identifying discriminatory and operationally affordable spoof characteristics and approaches. Furthermore, particular facial sections are frequently repetitive or correspond to image clutter, resulting in poor overall performance. This paper proposed a neural network model for face-anti-spoofing which outperforms the other models and shows an accuracy of 0.91%.

AdHoc On-Demand Distance Vector (AODV) is a notable and broadly utilized protocol for MANETs. The Mobile AdHoc Network or MANET, without any infrastructure, is an assortment of remote nodes conveying and communicating over a wireless network. All wireless devices working in AdHoc mode inside range are allowed to have communication with each other in absence of base station. The routers are capable to roam and communicate arbitrarily and organize themselves as per the requirements when the nodes structure themselves into a random topology. Radio signals possess range limitations due to which multihop communication in MANETs is inevitable. The performance of traffic situations utilized in a mobile AdHoc network is responsible for the transmission and gathering of data between source and destination in a MANET.This paper provides a comprehensive comparative analysis of the routing protocols with respect to variation in node configuration.

Nowadays individual’s identity verification is required at many places for authenticity like Government Sector, Private sector, Public Sector etc. Many existing systems are based on either physical documents or on biometric parameters. Biometric system has become a more convenient way for authenticity check. In biometric system, some samples of biometric parameters are collected and stored at the server side for further use. These samples would be used for the verification of the identity of the person. The actual required biometric parameter will be compared with all the existing samples available in the system to match with the registered person. If it matches with one of the existing sample, then that will be authenticated by the system and allowed to perform further operations. But what if, the samples collected by the authority get misused? That needs security from the owner of the system. So, the identity of the samples must be hidden from the operators by some way, which has been focused in this work. A new approach called as Cancelable Biometric Technique (CBT) using Modified Multi-Line Code Generation Algorithm (MMLCGA) is used for storing biometric samples using template. The cancelable method converts the gathered samples and stores it into the system for hiding its original identity. In the verification phase, the system will convert it back to the original sample to be used for identity matching of the person or user. This technique provides more privacy, because of which privacy threats can be reduced. The time and accuracy of the proposed technique is better by 15% and 1.4% respectively, when compared to the existing technique Multi-Line Code Generation Algorithm (MLCGA).

The twenty-first century has seen considerable implications towards social trends, as well as advances in technology and industrial achievements. The fourth industrial revolution is the consequence of a move towards automation and a decrease in the amount of human participation in most industries (Industry 4.0). Through an in-depth assessment of the existing available work, this article investigates the influence of technology on the Internet of Things (IoT) has and continues to have on Supply Chain Management (SCM) during the era of Industry 4.0. This analysis of the relevant literature not only provides a plethora of fresh information on Industry 4.0 but also indicates areas that require more attention and gives suggestions for the future. Investigating academic developments related to Industry 4.0 will assist in the process of closing this knowledge gap. A comprehensive review of academic articles on the impact of IoT on supply chain management in Industry 4.0 was conducted up to the end of April 2022.

Flooding in cities is a worldwide occurrence that presents a significant problem to municipal administrations and urban planners. The loss of the life, delays in public transportation, damage to public and private property, the interruption of services such as the water supply and power supply are some of the effects of urban flooding which leads to economic losses as well as public health issues. The motive of this research paper is to review the various strategies for managing urban floods and to determine the research scope in terms of smart city development. The flood is one of the most prevalent natural catastrophes that may strike any city. Rainfall, water level, temperature, humidity, drainage water level, water discharge, as well as other parameters are generally viewed in flood prediction models including artificial neural networks (ANN), fuzzy inference processes, regression models, deep learning, gradient boosting decision trees, and self-organizing feature mapping networks (SOM). Real-time flood parameters were considered in the flood detection and warning system. Real-time flood characteristics were considered in the flood detection and warning system, and the system was constructed utilizing IoT. The accuracy of flood prediction of computational intelligence techniques is only 76.48% in average.

As the modern world is applying smart technologies in every day and in every sphere of life, the advent of Industry 4.0 standards increase the risk of privacy loss while using the cyber physical systems. There are many popular methods of privacy preservation in CPS but they are too complex to implement and not have high degree of data utility. In keeping all these reasearch challenges, we have applied a novel approach of distributed constraint optimization for minimizing the privacy loss in a cyber physical system. The main advantage of our proposed mechanism is that the balance between local privacy level and global privacy level is maintained so that the data utility is not degraded.

Charcot-Marie-Tooth is a neurodegenerative disease which causes muscle degeneration and loss of sense in the appendages. Patients affected with the disease find holding an object in their hand difficult, since they have no sensation in their skin and have no idea how well they are grasping it. Over time, it is possible that their grip falls, and the object slips from their hand. To counter this, a device which improves the quality of life of people affected with Charcot-Marie-Tooth has been presented in this paper. The proposed device is in the form of a glove, which uses force sensors to continuously track the force applied by the user on the object they are holding. If it falls below a certain threshold, it warns them through haptic feedback-vibration motors placed in the forearm where the users’ sensation of touch is maintained. The sensors are paired with analog signal conditioning circuitry to obtain as good a linearity as possible without sacrificing resolution.

Road accidents are a major cause of death, and many victims die as a result of not reporting such events to the appropriate authorities. Because the event was not reported, there is a lack of emergency medical assistance, which leads to deaths. A computer vision-based traffic observing and revealing strategy can help with giving health related crises continuously, perhaps saving many individuals. Conventional traffic systems, which are outfitted with IP cameras and sensors, are currently set up all around the city to supervise and control traffic. In this paper, we present a better traffic checking framework that perceives and distinguishes moving items like vehicles, cruisers, etc. in live camera, takes care of, identifies accidents of these moving articles, and promptly sends crisis admonitions to the fitting authorities. An innovative architecture for detecting road accidents is given in this paper. The suggested framework uses YOLO to locate accurate objects, followed by accident detection for surveillance data. The nearest police station is notified of the observed accident. On commonplace street traffic CCTV reconnaissance film, the proposed framework gives a reliable procedure to accomplish a high Detection Rate and a low False Alarm Rate.

Today’s confined individuals and patients who have been infected by viruses can profit from robotic and IoT technology. The world has recently been affected by the Covid-19 epidemic. The virus-affected and quarantined individuals feel helpless since caregivers, medical professionals, and other individuals are scared of the dangerous sickness. This project will produce a robotic IoT agent that will help nurses and doctors monitor patient conditions and deliver meals and medications to them. A robot with a camera transmits data to a remote server through video; the droid cam app will be utilized for this. Sensors assist in retrieving body temperature from a remote location; a machine learning model trained on the image of the patient detects them and maintains the patient’s name and body temperature in a centralized database.

Information is the wealth of every organization and in the modern-day when information is shared digitally and via the internet, protecting this treasure has become a top issue. Private photographs need to be protected from unwanted access due to security concerns raised by internet photo transfers. Nowadays, practically everyone shares their personal information online, including photographs, either with other users or in a database that attracts cyber criminals who can use it to their benefit. Steganography can be as a security tool to safely transmit secret information because it is one such technique where the presence of a confidential message cannot be detected. This article compares various steganography techniques, including AES, LSB, DCT, DWT, etc. are compared them with each other and also more advanced techniques involving Cryptography and Steganography i.e., sharing secret data using counting-based s and matrix-based to increase security.

The popularity of social media websites tends to boost the volume of multimedia material. This results in the development of the brand-new industry known as the Multimedia Internet of Things. Lightweight image encryption methods are required to protect multimedia data on these resource-constrained devices. Since chaos theory has grown more common in modern multimedia cryptography, it serves as the foundation for the suggested lightweight encryption technique. 2D augmentation models are used in this chaotic-based multimedia encryption method to provide secure data transit. The suggested method has minimum residual clarity and key sensitivity while, simultaneously maintaining the excellent encryption quality of chaotic maps. The simplified image encryption technique employs the chaotic map model, which has the properties of confusion and diffusion. We have also put forth a novel key generation algorithm for use with the Logistic map and the Rubik’s cube transformation. Using the Elliptic-Curve Cryptography (ECC) Key Algorithm, the initial values are produced. To analyze the computational complexity, the suggested method is applied to medical (binary) and colored images. The histograms of the encrypted images are flat and distributed across all the pixel values, according to the security analysis. These images have an entropy of 7.86046675 with average correlation values of 0.0010575 (horizontal), 0.013994 (vertical), and 0.00235 (diagonal) (encrypted image). The suggested lightweight image encryption hence displays a high level of security.

Road traffic has been increasing rapidly for the past few decades, which has resulted in air pollution, long waiting times in road traffic, frustration, wastage of effective time on the road etc. The driver's poor vision in identifying road edges, especially during the night, may result in major or minor accidents, which all lead to traffic congestion. Finding a parking space during peak hours also results in congestion. With the emergence of machine learning, IoT, computer vision etc., these problems can be minimised. Road edge detection helps to prevent vehicles that are running off the road and parking space allocation helps to allocate unoccupied space for vehicles even at peak times. The paper addressed a study on different methods through which road edge detection and parking space allocation is possible. This literature survey helps researchers to understand the importance of various cutting edge techniques in road edge detection and parking space allocation. Based on the study, a model is being proposed to provide a solution to edge detection of roads and allocate parking slots, thereby optimising the road traffic.

Sensors can now recognize human physical activity with the recent technological advances. Accelerometers, gyroscopes, and magnetometers are some of the sensors embedded in smartphones. In today's data-driven world, human activity recognition is important in a variety of fields, including medical applications, fitness tracking, human survey systems, and so on. This research study analyzes the data obtained from mobile sensors such as gyroscopes, accelerometers, linear accelerometers, and magnetometers. Further, the proposed model predicts the human activity by using the data collected from a mobile sensor. Sitting, standing, jogging, and other such activities can also be tracked. Calorie-Meter is an Android application that calculates the calories burned while engaging in such activities. Using the application's predicted activity, the user’s calories burned and calorie deficiency can be calculated. This research study proposes the utilization of Long Short-Term Memory (LSTM) and a Neural Network (NN) technique for predicting the human activity based on sensor data.

In this era of Industry 4.0, securing file data is very crucial in today’s environment with respect to data transfer of Internet of Things (IoT) devices. Over the years with the evolution of technology and file storage systems, many algorithms have been used for encryption and decryption processes for securing the file data, each with its methodologies, advantages, and limitations. An efficient algorithm has very few limitations thus making it a top choice for usage. In this paper, we have proposed a symmetric key cryptographic algorithm called the Alternate Tiny Encryption Algorithm (ATEA) focusing on a strong approach for safekeeping of the file data and minimizing the weak points of the existing Tiny Encryption Algorithm (TEA) and Extended Tiny Encryption Algorithm (XTEA). The Alternate Tiny Encryption Algorithm (ATEA) is a Feistel cipher that utilizes mixed algebraic group operations. The algorithm is simple enough to incorporate into practically any computer program and can be quickly translated into a variety of languages. It uses a unique key generation technique making the encryption of file data more secure.

The advantages of free software are innumerable, and it is highly encouraging to witness many domestic computerization sources that provide free and efficient software program to Internet of Things researchers across the globe. The developers behind such computerization sources have worked hard to create a solid system foundation that anyone can use. Moreover, these sources are freely used by the people to obtain individual responses. Similar to most of the other software systems, for developing an advanced solution, a strong community that is willing to return it to its initial position is required. These characteristics have motivated to compile and compare a greater number of open-source automation software, which is considered as the most exciting domain of IoT. The user can utilize the open source software based on the suitability of the system available and preferred, programming language support for the system, number of users, desired security level, service protocol, and so on.

Computer networks are prone to cyber as a consequence of global internet use; as a consequence, academics have developed several Intrusion Detection Systems (IDSs). Identifying intrusions is one of the main significant study topics in data security. It aids in the detection of misuse and attacks as a safeguard for the network's integrity. Machine Learning, Bayesian-based method, nature-inspired meta-heuristic methods, swarm intelligent approach, and Markov neural network is some ways to find the most effective characteristics and thus improve the effectiveness of Intrusion Detection System (IDS). Over the years, numerous databases have been used to evaluate various projects. This publication provides a comprehensive assessment of IDS with machine learning approaches.

In India, Agriculture is remaining as the major occupation among people. Farmers cultivate various crops depending on the type of land and season. Additionally, farmers also irrigate agricultural land while cultivating. However, farmers cannot predict how much water they will use for irrigation. Currently, farmers are supplying water to the field without knowing the moisture content of the soil. If this continues, there will be a scarcity of water, and underground water will also be depleted and may not be available for future generations. Furthermore, the crops are destroyed if more water is supplied than the required amount. This research study has utilized an advanced technology called Internet of Things (IoT) to design and develop agricultural monitoring systems for evaluating soil moisture content and other agricultural parameters. This system includes soil moisture sensors for measuring the soil moisture content. Rain drop sensor, water level sensor, and DHT11 sensor are used for measuring rain, field water level content, temperature and humidity. The outputs from the sensors are sent to the ESP32 module, which then sends it to the motor (used for irrigation and ejection). These values can also be visualized in the ThingSpeak cloud platform.

RFID (Radio Frequency Identification) transponder is a small beacon tag responsible for transmitting the radio waves in the range of 30 kHz to 3 GHz to an antenna. The antenna is then connected to the reader, which sends and receives signals from the antenna. The use of multiple tags in a perimeter increases the chance of tag collision. Tag collision occurs when multiple tags send signal to the reader at the same time. Such a situation can lead to miscommunication between tags and readers present within the same perimeter; such miscommunication between reader and tags can lead to the failure of the entire RFID system. To avoid such a situation, an algorithm mechanism helps to avoid collision risks. The main goal is to collect data and identify various ranges and proximity in which an RFID tag collision may occur, as well as to conduct preventative analysis to avoid such failures.

BizGuru 1.0 is an online learning platform using mobile devices known as mobile-based learning. It is a modernized alternative to acquiring knowledge which is suitable with the current digitalized environment. BizGuru provides learning materials that promote business-related knowledge, focusing on Digital Marketing. However, in this study, the mobile application design will be focusing on the elder’s group to cater for their needs. The target users are people aged 60 years old and above, who use an Android smartphone and are interested in gaining new knowledge. The purpose of the proposed application is to help these retired elderlies find an alternative that enables them to gain income at late age to continue supporting their living expenses. With the current pandemic situation and how they are often related to poverty, both circumstances result in the elders having to struggle to survive financially. Therefore, by using BizGuru, the elderlies do not only get to familiarize themselves with modern devices, but also they could look for other alternatives to gain income and avoid poverty which helps to fulfil the 1st goal of Sustainable Development Goals (SDG) on the eradication of poverty issues. Besides, this proposed application also provides learning opportunities for elderlies who have the desire to gain knowledge at late age which can help fulfil the 4th goal of SDG which is promoting life-long learning opportunities for all.

The ever-increasing amount of e-medical data poses a security risk because of technological advancements in the healthcare business. An unstructured and large amount of unstructured data is generated by the healthcare data management system because of the wide variety of data formats that are used to capture patient information. Branches of hospitals can also be found in different parts of a city or state. Health information on patients that is kept in multiple places must be merged from time to time for research purposes. Cloud-based healthcare management systems can be an effective solution for storing and managing health care data more effectively. However, security is the most pressing issue with a cloud-based healthcare system. Elliptic Galois Cryptography (EGC) is used in this study to encrypt medical data files, and the value of the Galois field is determined using the Mayfly Algorithm. As a result, the proposed model is referred to as a “optimal EGC”. Use of the elliptic curve over a Galois field in elliptic curve cryptography reduces rounding errors. The healthcare data is protected in terms of both confidentiality and integrity when it is shared via the health cloud. Experiments have shown that the ideal solution can be computed more quickly in terms of file upload and download speeds as well as key generation and generation time. Additionally, it protects healthcare data from being tampered with during transmission via the health cloud.

More and more people are increasingly using multimedia on their mobile devices, such as smartphones, tablet PCs and smart watches as a result of technological improvements. The most significant elements of a mobile device are the battery life, memory, bandwidth, and CPU performance. When such computationally extensive tasks are performed on a mobile device, the battery quickly drains. However, offloading such tasks to a proxy and executing those results in significant power savings in mobile devices. We compare the ways of offloading to a proxy from a mobile device in this study based on power usage, energy, and execution time. A thorough examination of the offloading process is also presented. The findings show a significant reduction in the amount of energy consumed by mobile devices.

According to a survey, 137,726 accidents have occurred in the year 2018 alone. In order to avoid such accidents and also to provide other services, research on Vehicular Adhoc networks started. VANETs have hardware called OnBoardUnits (OBU) situated on vehicles and Fixed infrastructure situated on the Roadside called RoadSideUnits (RSU). Standards have been developed that allows communication between vehicle to vehicle and Vehicles to RSUs. Also, separate Bandwidth standard has been defined called Dedicated Short Range Communication (DSRC) to provide communication during emergency. However, we are yet to identify the associated risk of privacy and security in such communication. Then only successful implementation of Intelligent Transportation System can be achieved. Hence study of various attacks and its impact on Vehicular network is very much important. In this paper, we will perform the simulation of VANET and study the various performance metrics by introducing Sybil attack and analyse its impact over the network.

In the light of recent war crimes and data piracy conspiracies, privacy is of utmost importance to an organization and even to an individual. The majority of the population is dependent on third-party services for their daily communication. Albeit these major corporations advertise “secure” means of chat transfer, they install various kinds of backdoors to sell the user’s data to advertisers. Under the notion of going “Aatmanirbhar” i.e., Make in India, we have developed an indie solution without incorporating any third-party services or APIs. “Aatmanirbhar Sanchar” aims at providing users with a real-time off-the-grid, secure, and anonymous messaging service. It features an End-to-End encrypted transmission of messages and data files likewise. This is achieved by combining the open-source AES algorithm with a self-developed XOR encryption process.

Cloud computing is an up-to-date model for distributing information processing utility and provides a large amount of resources through the internet. The major challenges affecting a cloud computing environment include resource provisioning and security. In this paper, we focused on resource provisioning mechanisms using Meta-heuristics techniques such as spider monkey optimization (SMO) and simulated annealing (SA). A simulated annealing algorithm helps to give a fine solution along with statistical promises for uncovering the best solution, yet it cannot notify whether the best solution is found. So it requires another method to overcome this drawback. This paper presents the Spider Monkey Optimization algorithm with Simulated Annealing (SMO-SA) to generate the best fitness value possible. The aim of the proposed hybrid algorithm is to generate the minimum fitness value by combining spider monkey optimization with simulated annealing to provision the resources dynamically. This paper also presents the step-by-step mathematical working of our proposed hybrid algorithm by applying it to the relevant data set and calculating the speedup factor as well as mean square error (MSE) value along with fitness value, which shows the effective impact of our proposed SMO-SA algorithm.

The procedure of manual testing takes a lot of time. Automation is particularly desired since testing process is also error-prone due to its repeated structure. Robot Framework is a flexible tool that makes use of the keyword driven testing methodology. It is straightforward to use when high-level keywords may be created from current keywords. Although a command line interface makes it easy to integrate new test libraries, it is also possible to create customised test libraries in Python or Java using a straightforward library API. All these capabilities guarantee that Robot Framework can be used to execute test cases in a timely manner. This paper explains how a WebApp tool could be used to quickly automate testing procedures by reducing expenses and raising the overall functionality of the software. Results comparison of manual testing with automation testing shows that automated tests run on average 80.46% faster than manual tests.

Wireless Sensor Network plays an important role in collecting data from different environments where human involvement is deemed fatal or unnecessary. Apart from its usefulness, security threats and vulnerabilities exist as a common problem. A robust IDS (intrusion detection system) in WSN will be helpful in detecting and classifying the types of attacks, so as to remove or nullify the security threats. In this paper, we proposed a method to detect Mirai and GAF-GYT attacks in WSN using CNN along with f_classif function and normalization. Further, implementation of the proposed method has been carried out considering the scenarios: CNN without normalization along with f_classif function and CNN without normalization and without f_classif function. It is seen that the method that uses CNN along with f_classif function and normalization exhibits better performance in terms of parameters such as TPR, PPV, TNR, NPV, FPR, FDR, and FNR.

Network forensics is a branch of Digital Forensics concerned with analysing the network traffic to see if any anomalies are present that may indicate an attack or could lead to one. The goal is to figure out what kind of attack it is by capturing the details, store them in a forensically sound manner, analyse, and then present them in some visual form. A model based on traceability and scenarios, with proven literature and justification is desired. This study offers a professional digital framework in which the investigative process model enhances the systematic tracking of offenders. Cyber fraud and digital crimes are on the rise, and unfortunately less than two per cent is the conviction rate worldwide. Continuous and scientific research in this area is crucial to ensure safe and secure internet usage especially for money transfers and confidential personal communication. This paper examines the essential development phases of a Network forensics investigation model, and compares different network and digital forensic methods, and also offers a systematic model of a digital forensic model for cybercrime investigation. The survey also includes classifications based on infiltration detection systems, trace backs, distribution models, and attack maps. The aim of this study is to facilitate the digital forensic process and identify improvised practices. The Systematic Network Forensic Investigation model (SNFIM) aims to establish appropriate policies and procedures for practitioners and organizations.

Coronavirus has infected billions of individuals worldwide, with the number of persons infected continuing to rise. Humans contract the virus through direct, indirect, or close contact with infected individuals. This proposed work introduces a new feature, an intelligent community distance system, that allows people to maintain community distances among others in the indoor and outdoor areas, to avoid exposure to COVID-19 and to delay its spread locally and internationally, to help prevent the spread of COVID-19. The proposed research intends to monitor an IoT-based portable monitoring device that is designed to measure COVID-19 signals. Furthermore, by monitoring real-time GPS data, the system automatically notifies medical authorities concerned about any confinement violations of patients who may be infected. Also, figure out what new tool will be beneficial for tracking and predicting COVID-19 collections. To support in the analysis of COVID-19, the solution incorporates a mobile system coupled with a portable device that is equipped with clever IoT capabilities (complex data analysis and intelligent data detection) embedded within the system. A comparison of various machine learning classifier algorithms such as SVM, Random Forest, KNN, and Decision Tree is presented as the best model for making predictions and determining accuracy. We observed that KNN performs better, with a 95% accuracy rate. COVID-19 will be used to prevent the spread of diseases in future global medical problems using an automatic social distance monitoring and contact tracking system.

Advancement in the internet of things to meet the requirement of human beings and society makes integration of multiple devices into a single system. The integration of hardware and software needs to be provided with security to avoid the stealing of the data. Otherwise, the hacker may gain control over the devices and change the functioning of the system which may lead to malfunction. In order to provide security for the data transfer in IoT, the security algorithm need to be embedded with the IoT. The algorithm should provide high security and at the same time, it should be efficient. In this paper, an attempt is made to design a synthesizable Deoxyribonucleic acid (DNA) based Nano Advanced Encryption Standard (AES) Intellectual Property (IP) Core which can be used as a crypto engine in an IoT system. The crypto engine developed is optimized in terms of power, area and delay. The developed design when compared with the conventional design has given an area advantage of 81.6%, power of 21.17%, gate delay of 88.44% and path delay of 99.64%.

Internet of Things (IoT) is the term used to describe a network of physical things such as mobile devices and household appliances that are embedded with electronics, software, sensors, and network connection that enables these objects to gather and exchange data. Sensors, recognition and remote control of items are all made possible by the Internet of Things (IoT). Once this property is combined with sensors and actuators, it becomes an example of a cyber-physical system, which includes technologies like intelligent power grids (grids), intelligent homes (smart homes), smart cities (smart cities), and intelligent transportation systems (ITS). Integrating MANET and WSN with IoT is covered in this study. Technology and protocols needed to deploy the Internet of Things (IoT) are explored in this article.

In recent years, chatbots have become more widely used in a variety of fields, including marketing, customer service, support systems, education, healthcare, cultural heritage, and entertainment. Live chat interfaces have gained popularity as a way to engage clients in real-time customer care in many e-commerce contexts. Artificial intelligence-based chatbots, commonly referred to as conversational software agents, are created to have natural language conversations with human users. They are progressively taking the place of human chat operators (AI). In this study, we’ve covered a chatbot’s fundamental definition as well as its numerous sorts and qualities. In this study, various methods for creating chatbots are also mentioned and described.

A vehicular ad hoc network (VANET) is a wireless network that connects a group of moving or stationary vehicles together. VANETs were primarily used to provide safety and comfort to drivers in automotive environments until recently. Clustering is an important concept in vehicular ad hoc network (VANET) where several vehicles join to form a group based on common features. Clustering increases the complexity of data. The assessment of road accident strategies in Machine learning is presented in this paper. A road collision is the most unwanted and unexpected occurrence that may happen to a vehicle, due to the fact that they happen regularly. The goal of this study was to investigate the correlation between the concentration of collisions and injury. There are several factors that influence crashes, including weather, road conditions, driver distraction, and misread vehicle signals.

In the field of network and server security, every organization, no matter how big or small, has to adapt different monitoring techniques in relation with the servers and networks. Server attacks are a major threat in today’s world if the systems and networks and the host servers are vulnerable. Network and server monitors or administrators have to keep an eye over different tools for different purposes. Therefore, this paper offers a methodology with the help of which we propose to develop a tool that would ease up the work of server administrators. This tool would provide all the services that multiple other tools provide. This would help enhance the monitoring section of the web servers in particular by using different tools and modules with the help of python. This paper also explains and concludes that the tool being developed would be effective and efficient, and therefore save a lot of time and ensures security from threats related to servers.

Security is an essential factor that must be addressed in wireless sensor networks (WSNs) since they are resource-constraint, may handle sensitive data, and operate in hostile inaccessible regions. The Purpose of security attacks in networks is mainly to breach the confidentiality, authenticity and, integrity of data which consequently raises the energy consumption rate and lessens the network lifetime. Several methods and techniques have been applied by attackers to achieve their goals. However, one of the easiest methods is to prompt congestion which causes packet delay, packet loss, re-transmissions and ultimately jeopardizes the network. Unfortunately, aggravation of network congestion by the faulty behaviour of malicious nodes has not been discussed to great extent in literature. In this paper, we explore and discuss different types of attacks to analyse their impact on congestion occurrence and related problems through simulation in different network scenarios. According to this analysis, different types of attacks could have devastating effects on networks. This demonstrates the need for developers to create more secure WSNs.

Weather Forecasting is an Internet of Things (IoT) based initiative that attempts to provide weather forecasting via the internet via a website. Using several sensors, our project Weather Forecasting collects data such as temperature, humidity, rain, and pressure. Our project also includes a facility for measuring atmospheric conditions to give information for weather forecasts and to study weather and climate. As a result, Weather Forecasting using the Internet of Things is proposed to assist consumers in accessing weather data anywhere in real-time. For the data storage, a real-time database has been used using Firebase and will be displayed in the dashboard which was designed using React JS.

The Cloud adoption statistics clearly depicts that considerable number of organizations has hosted their infrastructure on cloud, most of the major migrations from on premise server to cloud are done and some are in progress. With this Cloud Computing era, Infrastructure as a service requires more efficient management. It comprises of IT resources for storage, networking, computing etc. Optimized and efficient use of the rented resources is the need of hour as resources are on rented basis and they have cost associated with it. As industry is growing, users are growing, resources are growing followed by volume and traffic. To manage any cloud infrastructure there is need of cloud monitoring solution with proactive alerting. This paper will provide entire overview with focus of providing open-source solution to monitor cloud environment.

In the modern world, wireless networks are more prominent. Due to the dynamically changing nature, mobile networks are more vulnerable to several attacks. Data security is a significant problem since data flow across a wireless channel that is naturally exposed, making possible for malicious attackers to obtain sensitive data. Finding and resolving security problems in wireless networks are essential steps in ensuring secure data transmission. One of the main threats that can be performed against wireless networks to stop legitimate nodes from sending and receiving data packets is distributed denial of service (DDoS). The proposed framework includes implementation, detection and elimination of DDoS attacks in the wireless network. To provide further security, an authentication approach is introduced which will grand access/denial to the receiver based on successful authentication of the nodes. Finally, analysis of the proposed model is performed in 3 modes i.e., normal mode, attacker mode and controlled mode.

In emergent field of networks everyone is able to access data as required. Huge amount of data transmission is done on internet; so data security, integrity and confidentiality become important. Data Security is improved by use of intrusion detection system (IDS).This system allows administrator to monitor network to keep it secure from vulnerabilities. Most intrusions happen in the network are due to an attack. A model based on hybrid structure for intrusion detection system using rule based genetic algorithm for anomaly detection for new identified attacks is presented here. Model makes use of various algorithms of machine learning such as naïve bayes, support vector machine and random forest into ensemble. Ensemble approach helps in performance improvement. Behavior based detection is proposed using rule based genetic algorithm. Model is trained using UNSW NB15 dataset. This allow model to detect all types of modern attacks. Model overall performance is observed with 98.5% accuracy and 0.11% for false alarm.

Email is the most common channel through which cyber attackers commit crimes and initiate spamming attacks. Spamming is a popular method of sending unsolicited messages in order to distribute malware. The disadvantages of spamming include system slowdown, time consumption, and the presence of viruses. The main goal of this work is to identify spamming text, url, and message id and reduce the rate of spamming on e-mail. To reduce and control spam in e-mail, the text content, links, and header information in the corresponding e-mail can be analyzed. In the above scenarios, one of the main driving forces is the use of natural language processing to distinguish between spam and non-spam occurrences. The major goal of the proposed study is to consider three features—text, url, and message id—rather than deploying a single feature. The deep learning models, which play a significant role in predicting the spamming texts than any other traditional machine learning models. The machine learning models also play an important role in predicting the spamming urls and message id. Three standard datasets—Enron for text content, Phishtank for urls and SpamAssassin for Message Id are collected and processed. The aggregation of machine learning and deep learning models are an added advantage in this work. This identifies the spamming e-mail on the real-world datasets by using LSTM, Random Forest, Multinomial Naive Bayes models, which then evaluates the performance. Further, the output from the three models namely text, url, message id are integrated using weighted fusion approach and finally the output will be obtained as spam or ham.

With the consistent spike in the world’s population and continued expansion of urban cities, waste is a highly visible city problem for every prominent stakeholder including the respective State Government of the territory. Waste management in the cities has big expenditures based on the method of waste collection and disposable systems that are opted by a particular city. There are several ways to enhance the waste collection mechanism, however, one of the most appropriate systems with appropriate use of technology is key to getting the waste management system right and making it commercially viable. The latest entrant to smart city waste management is the Internet of Things (IoT). The smart bins comprising IoT sensors and route optimization improve traditional waste management processes in the most efficient manner. Therefore, this paper reviews various IoT-based solutions used for waste collection and route optimization methods for garbage collector vehicles in the past decade. The objective of the paper is to study and analyze the state-of-the-art techniques used for the waste management system with smart IoT-enabled bins. In this paper, analyses of research papers are described in the literature. Research gaps from an existing work have been concluded based on the results of the study. Further, this paper also describes the various challenges and issues of the smart waste management system. This thus, calls for further improvement and innovation.

Internet of things (IoT) devices can store and manage the real-time data created by many restricted Internet-connected devices. If one of the nodes were compromised due to Man-in-the-Middle (MITM) attack, the network might suffer significant damage. Due to the limited resources of constrained devices, it is difficult to incorporate appropriate cryptographic capabilities. Hence lightweight cryptography strives to meet the security needs of situations with few resource-constrained devices. In this paper, the framework is constructed using the smart aircraft environment monitoring system (SAEMS) and created with the help of nodes and the message queuing telemetry transport (MQTT) protocol for communicating the sensor data. A hybrid lightweight encryption algorithm (HLWEA) is proposed to mitigate the MITM attack on IoT devices. The HLWEA comprises (i) Key generation and (ii) encryption and decryption. The proposed method achieves an encryption time of 0.0309 ms; encryption bandwidth is 19.02 kbps, decryption time of 0.029 ms and decryption bandwidth of 19.36 kpbs. The proposed implementation is a smaller key size, minimal time complexity, and enhanced real-time cryptography-capable security.

Insect identification is one of the most pressing difficulties for Indian farmers, as numerous insect species harm a vast number of crops and hence diminish the quality of harvests, resulting in financial losses for both farmers and the country. However, in agriculture, the combination of IoT and machine learning (ML) allows for ease and innovation, allowing farmers all over the world to better their farming operations. On the other hand, in India, a very little amount of farmers is aware of smart farming and its benefits. Various study on research paper shows that the proper use of IoT devices embedded with the machine learning algorithm can reduce the task of farmer at very early stage of the plant life and thus saving the crops from being degraded, also included the survey of various research done across the globe and identified the potential methods which must be included for the current era farmers in order to minimize the insect effect on the crops. The aim of our experiment is to involve ML and IoT technology to sense the crop conditions in terms of quality and whether it is affected by insect or not for this a experimental study with the help of image processing has been performed thus calculation of results done accordingly. There are various sensors, which are equipped with ML technology like computer vision algorithm, which make the sensor powerful, and images being captured by these sensor can be analysed automatically and thus trigger the automated pesticide treatment systems using a ML-based decision support model. In this paper, study of Convolution Neural Network (CNN), Long Short Term Memory (LSTM), Support Vector Machine (SVM), Grid search based SVM (Grid-SVM), and K-nearest Neighbour classifier has been done Among them based on the required performance, nd the CNN-based model is much accurate for predicting the required treatment. The CNN has achieved up to 88% of accurate classification. Further, the model has been extending by incorporating the regression analysis, which enables the system to recommend the quantity of the required treatment. In this context, study on the KNN regression and Support Vector Regression (SVR) model has been made, among them the KNN regression provides up to 99.8% accurate prediction for treatment quantity prediction.

There is a strong correlation between attendance of school and offices and its attendees’ performance and success. The traditional ways to maintain attendance for organizations are time consuming and cumbersome. A novel way of doing this task is proposed in this paper where attendance of a person is marked based on his/her face and voice based on voice and speaker recognition. Both the biometrics are preprocessed to feed the combination as a datapoint to the Convolutional Neural Network. This ensures that proxy attendances are avoided and the shallow network is able to perform well. The model achieved an accuracy of above 90%. A python based interface facilitates the entire process of person registration, attendance marking and database maintenance.

6G wireless network is going to revolutionize wireless systems by introducing several innovative services such as virtual reality (VR), 16K Video, Vehicle to Vehicle communication, and Internet of Everything (IoE) on a commercial scale to increase end-user experiences. Hence, network infrastructure needs upgrading to provide higher data rates, massive connectivity, and more secure wireless systems to meet the use case requirements. Distributed ledger technology and blockchain, which is known to be a disruptive technology enabler, can address the challenges and functional needs of 6G technology. In this work, we investigate the opportunities of those blockchain-enabled services in the 6G network, along with the shortcomings and limitations that need to be discussed in further researches.

This paper presents the machine learning based automated disaster message classification system. Machine learning is be used to identify such information and provide valuable information for aiding disaster response during emergency events. Disaster management prediction systems (DMPS) are computer systems for determining when and where to deploy mitigation measures in the event of an emerging natural or man-made hazard, while accounting for and mitigating human factors that may compromise operational effectiveness for providing fast services to handle this high volume and velocity of urgent information. Till date diverse techniques used for disaster and pandemic management are available using the technologies like satellite-based systems, cellular networks, Internet of things (IoT), smartphone-based systems, 5G and cellular networks. Linear support vector machines are an efficient way to learn discriminative models, which is especially useful in data where the number of attributes is large or is not known. Linear SVC (Support Vector Classifier) is one of the most successful linear models, not only because it is quite fast to train and compute, but also because it can achieve excellent performance in high dimensional problems. Linear SVMs can make use of a wide variety of learning algorithms.The proposed work uses Linear SVC Algorithm with strategy of self-training that learns from available datasets with the labeled data. Finally, the paper gives the message classification based on the emergency to the relevant disaster.

Heart and kidney are the two major organs in a human cardiovascular disease (CVDs) and Chronic Kidney Disorders (CKDs) are the leading by causing death and health related issues globally. Mostly, cardiovascular diseases or any heart abnormalities can be prevented by addressing some of the risk factors such as using tobacco, unhealthy diet which leads to obesity, physical inactivity and massive consumption of alcohol. CKD means the kidneys are damaged and losing their ability to keep our body healthy by filtering the blood. CKDs can only be treated, by early clinical-diagnosis and treatment. So that it’s possible to slow down or stop the progression of kidney disease. The heart helps to pump the blood filled with oxygen through all parts of the body, including the kidneys. As we know, the kidneys help cleaning the blood, by removing the waste products and extra water in the body. Without the help of kidneys, the blood in our body would contain too much waste and extra water that is unnecessary, which can lead to be fatal infections sometimes. So, only by the proper functioning of both heart and kidney would help maintain our body functionalities properly. It is very important to understand that everybody with renal disease is at risk for heart problems, which can increase your chances of developing heart disease. The Intelligent health care system works as a web application in which users can enter some blood test parameters and blood pressure (BP) in order to find if there is any abnormality or not in their heart and kidney. This application creates awareness about the heart and kidney health. It is better to follow prevention than getting cured is the strategy followed here.

The Industrial Internet of Things is a intricate area which comprises feature like information and operation technology, statistics, and engineering. The industrial data management system uses five basic layers like things layer, edge layer, fog Layer, communication layer, and cloud services to build a system for industrial operation. The cloud assisting in fetching and acquiring vast industrial data generated by several devices in the industry on the shop floor and retrieve necessary information based on context aware approach to create a smart enterprise based on industrial scenario. The paper presented a new solution for industry using IoT for predictive and remote maintenance provision for various industrial environmental parameters and assisting in increasing the work carried out by hand as well as productivity in industry using a machine learning approach. More specifically, this IIoT solution captures air quality, outdoor temperature humidity, boiling temperature from one sensor node and object detection, indoor temperature, humidity, smoke and light intensity data sensors from other sensor node in the system base on controllers and analyses them in the fog layer to provide a timely evaluation of intelligence require to operate the system which is helpful in increasing the productivity in the production line. The proposed experimentation illustrated the design of the IIoT solution, described the prototype industrial plant in normal and abnormal operation, analyzed with supervised machine learning approach and presented the sensor data analysis to create a smart enterprise.