Next Generation of Internet of Things

Agriculture expansion is critical to the economic prosperity of any country. Agriculture employs more than 60% of the Indian population, either directly or indirectly. Nowadays, monitoring the crop is the challenging task in the world. In this article, data has been collected from various sensors to propose an IoT-assisted hybrid machine learning approach for obtaining an effective crop monitoring system. Crop monitoring system here means predicting as well as detecting diseases of crops. This study is about leveraging existing data and applying regression analysis, SVM, and decision tree to predict crop diseases in diverse crops such as rice, ragi, gram, potato, and onion. Among the applied methods, SVM outperforms regression, DT methods. The training and testing accuracy of Gram has 96.29% and 95.67%, respectively.

This paper presents the results of an investigation and analysis of the effects of broken squirrel-cage bars. In the investigation, a comprehensive time-stepping coupled finite element approach was fully used to compute stator current waveforms, torque, and magnetic flux density waveform. The harmonic component of air-gap flux density is analysed. From these data, the faulty signatures are extracted. The present method has been designed and implemented using Finite Elements Method depending on time stepping. The proposed method produces an efficient technique in terms of time and accuracy to detect the faults and effects on the operation of electric machine. The early detection of faults in electric machine gives enough time to decrease the probability of electric machine faults. The differences in motor torque waveform timing in each case associated with the stator current waveforms give the flux spreading in the suggested technique. The obtained results show fast fault detection and technique is founded to extract the induction motors faults.

With the advent of mobile technology, the modern paradigm of “connected everyday objects” was built over the current network. The tremendous development of networked devices had increased its reach over the primitive network topologies. This significant change has launched the revolution after flat-page. The surge in the global urban population is placing new demands on people's daily lives in terms of pollution, public safety, road congestion, etc. To accommodate this rapid growth, new technologies are being developed and smarter cities are being built. Incorporating the Internet of Things (IoT) into everyday life makes it possible to develop new smart solutions such as services and applications for industries like hospitals, surveillance, forestry, etc. Research on Artificial Intelligence (AI), Deep Learning (DL), and help of Data Visualization have shown how IoT performance can be improved with some technological aids. This creates a rapid demand for addition and works in terms of Big Data with first-class technologies that we have around us, so in this paper, we will talk about such things and show a comparison on this basis with the other works that are under it, with the deep learning and artificial intelligence models. This study will help us to show how it overall contributes to the growth of the Internet of Things in society to provide a better life for future generations. Finally, we will outline the existing obstacles and problems that occur during the smart city growth facilities.

Investigate the vibration of the rectangular plate related to the liquid, normal body frequency and modes shapes of cantilever plate without and with hole and perforate plate with 169 hole in air and contact of the water surface and immersed in water are presented using finite element method via ANSYS15 software. Acoustic model in three dimension domain is considered using APDL program to take the variables. The gotten comes about detailed the 6th plate normal body frequency and mode shapes which are based upon the behavior of the plate. In all cases, there are diminished with in the natural frequency of the fluid–structure framework. It can be concluded from our work that the exactness of the predicted frequencies utilizing our demonstrate is either exceptionally great or at slightest adequate for commonsense purposes.

The machine learning (ML) and Internet of things (IoT) technologies are increasingly focussed on decision tree classification algorithm. Its use is expanding through numerous fields, solving incredibly complex problems. DTCA is also being used in medical health data using computer-aided diagnosis to identify chronic kidney diseases like cancer and diabetes. Deep learning is a class of machine learning that utilizes neural networks to solve problems and learn unsupervised from unstructured or unlabelled data. The DL used to deep stacked auto-encoder and softmax classifier methods is applied for CKD. In this work, based an R Studio and Python Colab software using random forest, SVM, C5.0, decision tree classification algorithm, C4.5, ANN, neuro-fuzzy systems, classification and clustering, CNN, RNN, MLP is used to predict multiple machine and deep learning techniques, discover an early diagnosis of CKD patients. In this work, classify the chronic kidney disease various stages are identified.

Nowadays, social media outlets involve peoples’ opinions, reactions, and emotions. Sentiment analysis classifies the text from those sites into negative or positive polarity. Over the years, a multitude of researchers studied Arabic sentiment analysis but most of them focused on standard Arabic language. However, the Arabic dialects should have much concentration by researchers. Therefore, the main focus of this research is the sentiment analysis of the Iraqi Arabic dialect. The data sourced from Facebook platform (Posts and Comments), the most popular social media site in Iraq. Then, the data passed through several preprocessing steps and weighting methods. The processed data then passed into comparative experiments with six machine learning algorithms including Naïve Bays, Support Vector Machine (SVM), Logistic Regression (LR), Decision Tree (DT), Random Forest, and K-Nearest Neighbor (KNN). The results indicated the highest accuracy achieved by Naïve Bays with 81.2%, followed by SVM and LR with 74%, while DT and Random Forest achieved accuracy 64% and 63%, respectively. The worst result was achieved by KNN algorithm of 57%.

Sentiment analysis is used to analyze the impact of Git commit on various open-source Java projects. In order to understand the positive, negative, and neutral impact of sentiment on developer community, we have reviewed many research paper based on real-time projects. It is observed that emotion is having a wide impact on software product quality by considering different factors like software requirement, team management, and work distribution which contributes toward negative commit comments. Code commit analysis is either done on daily or weekly basis by considering 8 classifications of emotions. In order to do a detailed survey, different open-source projects like Eclipse, JEdit, ArgoUML, and JUnit are considered for commit analysis based on software refactoring activities. In order to give an exact statistical result on sentiment analysis, different machine learning and deep Learning classifiers can be used. Based on the survey, it is also concluded that code refactoring is highly influenced by positive or negative impact of developer’s emotions.

IoT is growing at a fast pace, and billions of devices are now associated with the amount expected to reach trillions in the coming years. The Internet of things and the individual systems participate closely in launching the fourth industrial revolution and form alliances and develop the goods of the next generation. The foundation of Industry 4.0 is the transforming innovations. The convergence of robot and IoT agents contributes to Internet of Robotic Things concept, where creativity in automated devices generates different possibilities, both in business and science. It covers a range of sectors, including agriculture, manufacturing, health, education, and surveillance though the application of various technologies. The study discusses the new Internet of Robotic Things developments, which have an influence on the area of health, science, agriculture, manufacturing, education, and surveillance and the key open problems of introduction of robot technology into intelligent spaces. Internet of Robotic Things technology and frameworks are often addressed to highlight their effect on daily life and to promote more study on remote and automatic applications.

Role of web security has become an important topic as web and web application became quite demanding and people started using them every day. Last year we saw slight growth in web attacks and exploitation of their vulnerabilities. For example: Recently on the website of one of the largest airlines, Air India, an attack was performed where user’s credit card and sensitive information was captured. Input validation issue has resulted in many web application vulnerabilities. Broken access control, improper error handling, etc., are some of the examples of web security vulnerabilities. It is quite easy to acknowledge and eschew many of the web vulnerabilities. Regrettably many of the web developers are not security aware, which causes many of the web pages and websites to be vulnerable on the Internet. In this paper, Authors have discussed some popular web security vulnerabilities to detect them by providing some security mechanisms and some results to safeguard our web applications. It is a challenging and difficult task to protect our web application from vulnerabilities as nearly every day new techniques, tools, and methods are being invented by the attackers. It is one of the essential parts to know what various types of vulnerabilities are and how to detect and secure them from attackers.

In this present digital world, smart phones play a vital role, especially in the healthcare sector. The various mHealth applications installed in smart phones can solve primary health-related issues at a fingertip. However, the utility of those applications will increase if they meet user satisfaction and can solve the tasks effectively and efficiently. Thus, usability evaluation of mHealth applications is a matter of concern. Many usability evaluation techniques have been used till now but a majority of them are not unified and do not safeguard all usability aspects especially for mHealth applications. The purpose of this study is to recognize specific attributes that might help tremendously in assessing the usability of mHealth applications and examine the features of various usability evaluation methodologies for evaluating mHealth applications featured for diagnosing Type 2 diabetes mellitus (T2DM).

The coronavirus, one of the deadliest virus erupted in Wuhan, China in December and has claimed millions of lives worldwide and infected too. This virus has off-late demonstrated mutations thus making it difficult for the health professionals to adopt a uniform means of cure. Many people due to lack of support have confined themselves at home. The hospitals too are running short of equipment and support systems. Thus, computational connectivity between the patients at home and the hospitals needs to be established. The objective of this paper is to propose a framework/model that connects all the stakeholders so that either in regular monitoring or in emergency cases help can be provided to them. It has been well established through research and case studies that critical factors associated with this disease are oxygen level (SPO2), pulse rate, fever, chest infection, cough causing choking, and breathlessness. Data shall be collected, stored, and analyzed for the above symptoms and for this cloud storage and blockchain technology would be used. It has been established through various studies that non-clinical techniques like AI and machine learning prove to be effective for the prediction and diagnosis of COVID-19. Using this theory as the standard basis, machine learning models like SVM, Naïve Bayes, and decision trees can be used for the analysis, diagnosis, and prediction. Using IoT and its variants, remote monitoring of patient, and consultation can be provided to the patient. Appropriate action would be taken. In addition, a mobile application would enable the patients to gather or read about experiences of other patients. Thus, it would be established through the proposed framework, that an integrated approach of technologies has a great potential in such applications and offers several advantages.

Diabetes is a long-term illness that has the potential to disrupt the global healthcare system. Based on the survey report of International Diabetes Federation (IDF), there are around 382 millions of people, who are affected by diabetes worldwide. This number will have increased to 592 million by 2035. Diabetes is a disease characterized by an increase in blood glucose levels. Elevated blood glucose is characterized by frequent urination, increased thirst and increased hunger. Diabetic consequences include kidney failure, blindness, heart failure, amputations and stroke, to name a few. When we ingest food, our bodies turn it into sugars or glucose. Machine learning is a new field of data science that investigates how computers learn from their prior experiences. The objective of this study is to develop a system that can detect diabetes in a patient early and more accurately using a combination of machine learning techniques. The objective of this study is to use four supervised machine learning algorithms to predict diabetes: Support Vector Machine, logistic regression, random forest and k-nearest neighbour. Each algorithm is used to calculate the model's accuracy. The model with the best accuracy for predicting diabetes is then picked. This paper proposes a comparative study for accurately predicting diabetes mellitus. This research also aims to develop a more efficient approach for identifying diabetic disease.

In the field of technological development, Internet of Things (IoT) presents diverse mechanisms to companies to expand their business transactions with customers anywhere and anytime. Processes of selling and buying are getting to be much easier using IoT; therefore, IoT will change the customer's ideas and help them choose the appropriate products. This research article presents the advantage of the Internet of Things and RFID technology in the E-commerce system. From the last year, RFID technology becomes one of the few methods for managing products to guide and help the customer in buying the foremost suitable product which answers his need. This paper explains the use of IoT-supported architecture for E-Commerce with the goal to automate the sale transaction process with physical interaction. It solves problems that companies and customers face in shopping, as knowing our customer desire or knowing which the successful products within the supermarket are.

This work seeks to explore solutions to the challenges posed in sharing/integrating/transferring EHR data across heterogeneous healthcare institutions in developing nations, particularly in Nigeria. This blockchain-based algorithm model for EHR interoperability seeks to address EHR interoperability challenges such as semantics across heterogenous healthcare institutions, the proper infrastructure and consensus structure for the sharing of EHR across healthcare institutions, privacy and security of patients’ records, etc. Hence, this algorithm model, when adopted, is adjusted and contextualized to fit developing nations by addressing the underlining fundamental challenges. EHR interoperability would become a reality across the heterogeneous healthcare institutions in these nations. Future works on EHR will focus on the aggregate blockchain model. Also, the Artificial Intelligence (AI) model inculcating blockchain technology would be an attractive option to dive into future works on EHR.

COVID-19 virus named CORONA is a vigorous disease spread all over the world very quickly and creates a pandemic situation to the human beings normal life. As per the doctors and researchers from the laboratory point of view, it will spread to a huge volume when humans are not followed certain principles. Moreover, this disease is easily transferred to neighbors and others in a short period which leads to death. To rectify the remedy for this virus, various spread countries and research peoples are creating the vaccines and some precautionary methods for living hood situation. Recent techniques are used to detect and monitoring the COVID-19–affected person’s lifestyle and insisting they take precaution steps for early pre-pandemic life. IoT is a framework that is used to generate data from the human body from the sensors opted for human conditions. Wearable devices have been created with these sensors and communicated with human bodies directly or indirectly. The generated data will send through the server using any connectivity techniques such as Bluetooth or Wi-Fi. Analytics will be done at the server side for taking actions like the human body is affected by the COVID-19 virus or not. Finally, the generated data from a human can continuously store in real time in a cloud server which will be managed as a framework efficiently. This research work proposes a framework for data management in the early detection and monitoring of COVID-19 persons through IoT wearable devices in a pre-pandemic life. The experiments have been done at different zones, and the results are shown symptoms of COVID-19 disease. Parallel work reveals the data management in a cloud server since data have generated continuously in real time and tracking details also stored genuinely. Data management is the typical process in this research because all the data were generated in real time and analytics will be done whenever required. For that large amount of space and effective retrieval technique is required for data extraction. This research work data set is derived from various Internet sources like government web sites and mobile applications, and then, results have displayed the COVID-19 disease details accurately in real time.

Data is generated at a much faster pace, and it is increasing exponentially day by day. Machine learning methods are being used to extract patterns and trends from data to streamline different business activities for more profit with fewer resources. Machine learning models need to be trained with lots of data before being deployed for predictive analysis (Lecture notes in Computer Science, 2012 [1]). Training time depends upon the complexity of an algorithm. We are analyzing the space and time complexity of various machine learning algorithms so that it becomes easier to select and deploy the most efficient and appropriate model for a particular dataset. This research work primarily focuses on data analytics for supervised machine learning algorithms in industrial research domains.

Biometrics are vital in security. Facial recognition, fingerprints, and iris recognition are all examples of computer vision biometrics. Unique authentication based on iris structure is one of the finest approaches for iris identification. This research provides an iris-based biometric identification system combining CNN and Softmax classifier. The system consists of picture augmentation by histogram equalization, image reduction by discrete wavelet transformation (DWT), segmentation by circular Hough transform and canny edge detector, and normalizing by Daugman's rubber-sheet model. Each picture is adjusted before being fed into the DenseNet201 model. The Softmax classifier then sorts the 224 IITD iris classes into 249 CASIA-Iris-Interval classes, 241 UBIRIS.v1 iris classes, and 898 CASIA-Iris-Thousand classes. The performance of our suggested system is determined by the setting of its deep networks and optimizers. In terms of accuracy, it exceeds existing approaches by 99%.

Advancement in technology and connectivity has led to the development of the Internet of Things such that in the present era it is now being deployed into daily-use objects. The idea is to control the basic functions and features of these objects through a computing device either automatically or remotely. The class of such objects that come under the umbrella of a home of a person is called smart home devices, and one of them is the smart doorbell. Smart doorbell offers the main feature of security, remote control, and other features depending on the sensors and functionalities it packs. Security from intruders or an unknown person is the primary worry of a homeowner, and hence, smart doorbells can be a solution to ease out this worry. This paper reviews the developments in the field of IoT-based smart doorbell proposed by researchers and also discusses the results, limitations, and possible future developments.

This paper presents a new electronic system that uses technology which is the student enrolment system (SES) used at the University of Technology (Baghdad). The work aims to save time and effort for both the direct admission staff and the student. Application (SES) was evaluated using several variables: (student score, school branch, student desire, role of success in the high school, admission channel, application date and number of seats allocated to each admission channel in each department). The depth-first search algorithm was used to search in data as a tree or graph data structure and speed up the admission process for first-year students at University of Technology (Baghdad). The results obtained from the (depth-first search) algorithm was tabulated using the SPSS statistical programme (statistical version 20) to analyse variance to determine which of the factors most affected the results of student enrolment to the departments. The results showed that the electronic system (SES) included the acceptance of the highest rates in the scientific departments according to the student's desire and within the qualitative capacity of each department. The results showed that the best average was present at the student model 3 (0.218567 ± 0.0044792), which means that the student made the best choices. The results of (ANOVA) showed that when the value of (P ≤ 0.05), there will be statistical significance, as it was found that the student score (p = 0.001), the role of success (p = 0.003) and the number of seats available for each department (p = 0.004) are among the most influential factors on student acceptance, followed by the student branch (p = 0.033) and application date (p = 0.060), while the student's desire (p = 0.549) and acceptance channel (p = 0.56) have the least influence on the student's acceptance.

In the field of computer science, nowadays, there is a major drift in the field of the Internet of things and is rightly labeled as a smart revolution combining the Internet and things around the world. Using IoT, it is possible to allow connections between billions and trillions of devices and things with the help of the Internet and thereby allowing for the exchange of information and making the daily life of human beings easier and smarter and thereby improving the quality of service. Cloud computing on the other side is handy and scalable access to the network that allows access to on-demand services and allows computing resources to be shared those accounts for dynamic information integration from different sources of data. Merging cloud computing with IoT calls for open questions relating to the merger and confrontations that seek attention and discussed to allow for the successful merger of the Internet of things and cloud. Both the technologies are interdependent for their further growth and publicity. Cloud computing offers a huge number of computing resources that can be used by IoT, on the other hand, cloud computing allowing IOT to use the resources available on cloud accounts for dynamic and distributed data integration. The objective of this paper is to give an informative overview of the merits and demerits of the merger of smart devices and cloud computing. The paper also addresses the challenges faced by allowing for the merging of cloud computing and IoT. In the end, the future research ideas about open issues of the merger resulting in a new paradigm SmartCIOT (cloud-based IoT) are also discussed.

With the advancement of the Internet of medical things (IoMT) technology, security and privacy have become more important, as this technology allows access to medical equipment or resources from anywhere and at any time. Various methods have been created to provide security and privacy to IoMT, but in this paper, we offer a data gathering methodology that leverages the one-time-pad and symmetric key to protect our data and safeguard the privacy and usage of medicinal e-Records. This technique reduces computing time, increases processing speed, and has a lower implementation cost. The suggested technique is implemented in widely-utilized Python technology, which yields superior findings. Finally, the suggested technique increases security and protects our medical data.

The research work depicts an intelligent cloud and IoT-based voice-activated vehicle that responds to voice commands. Noise and distance handling, on the contrary, will require further progress. The car is controlled using speech instructions that are straightforward such as forward, backward, left, right, and stop. An android application sends these commands to the Bluetooth module. We can monitor the car and process the data with the aid of an intelligent cloud computing device. The Bluetooth device and control unit work together to retain and monitor voice recognition. This vehicle then responds to commands obtained from an Android application, allowing the user to control the vehicle via Bluetooth or voice commands, and monitor it using a real-time GPS tracking device. The microcontroller analyzes and executes this instruction. Image processing is applied in the vehicle to become aware of the shadows and obstructions. This car will function without the use of any hard manpower; simply attach your phone to the device, enter the password, and use it as a voice controller or Bluetooth controller. This research has been limited to short-range; however, using long-range modules could result in long-range communication with the vehicle.

Software testing refers to a testability method which has test support to improve and predict the software testability. Various types of method have been adopted by researchers and practitioners to improve the testability mechanism in software testing domain. This paper main objective is to reviewing the body of knowledge in this domain and provides a comprehensive overview to new readers and researchers about the software testability. This review selected eighteen papers as evidence to discuss the benefits, limitations, and proposed methods in the domain of software testing. We believe that this short review will give a quick overview to new researchers and readers in the field of software testability.

Deaf and dumb people use a sign language that can only be communicated through hand gestures to express their ideas and views. This coded language is mainly used by people who have speech and/or hearing impairment. The sign language is constructed by various movement of hands, arms, legs, or facial expressions to express their opinions. Meanings are communicated for every movement or position of gesture. Hand gesture plays a significant role to make mother tongue of impairment people for daily communication. The captured image feature can be extracted to translate the hand gesture communication to text\voice format to minimize the gap between the deaf and normal persons. This work considers the images of sign numerals to classify the numbers 0–9 and the alphabets for A–Z (including space).

Linear induction motors (LIM) are applied obviously in industrial applications, mainly in linear motion models. In fact, these machines have two general problems which are low efficiency and low power factor. These issues cause different side effects such as high power consumption, high input current, and occupation. This research studies the dynamic behavior of a single-sided linear induction motor by changing the basic design parameters. The process of selecting the equivalent circuit elements was also carried out using a new algorithm to obtain a modified model to meet the requirement, which is to improve the efficiency and power factor simultaneously. The final equivalent circuit of the modified model was adopted as an input to a simulation program using MATLAB/SIMULINK, which helped build an integrated model using mathematical equations with the specified inputs and the required outputs. Also, the process of testing and comparing the results is implemented for the proposed model in different cases with and without load. Then, analysing the results of motor performance and response speed using a voltage source inverter (VSI) with and without using a PID controller to improve the dynamic performance of the model, and speed control. Finally, the optimization results are validated, and the results are compared. The outcomes of proposed algorithm are shown by the genetic algorithm (GA), particle swarm optimization (PSO), and Cuckoo search with respect to the power factor and efficiency.

Agriculture is backbone of any country’s economy, and also, good crop yield is highly essential for supporting the growing demand of increasing population. By using machine learning, we will be able to predict the crop yield and also the right crop that can be grown in a particular area by analyzing the soil data and the weather data of the particular location. This study mainly focuses on how supervised and unsupervised machine learning approach help in the prediction. Different machine learning algorithms include KNN algorithm, SVM, linear regression, logistic regression, NB, LDA, and decision trees. Taking different dataset preprocessing operation is performed, and missing data are modified so that it does not affect the prediction. Then, the processed data are utilized by the machine learning algorithms for making the prediction. The dataset is divided into training set and test set, and the accuracy of prediction is verified. There are different performance metrics which can be used to evaluate the accuracy in prediction of the algorithms like MSE, MAE, and RMSE, coefficients of determination metrics (R2), confusion matrix, accuracy, precision, recall, and F1-score.

Advances in mobile technology have made significant improvements in helping the visually impaired. The glasses project helps blind people detect and recognize objects in their environment that they see through a small camera attached to their glasses. This method helps a blind person establish a connection with a nearby object by sending a voice message to an earpiece worn over the blind ear. The goal is to develop intelligent systems that can mimic the human eye. To do this, we use a small device called the “Raspberry Pi 4” that works in a way similar to the human brain, using a camera. It is known as a convolutional neural network algorithm for object recognition using deep learning algorithms. Finally, the moment the features of the image are recognized, the sound of each object is transmitted so that the visually impaired can know about the object in front of them. Python was used to create this project. The results showed that the blind CNN classifier achieved 100 curacy on the COCO dataset.

CBIR—content-based image retrieval is commonly known as the process or technique to the “image retrieval” problem that is the problem of analyzing and searching for a real content of images. Image search is a search technique that uses images as a source to retrieve an image that is similar to the given image. This technique has many applications in various fields and industries: securities, banking, education, business, etc. For the past decade, a variety of approaches has been introduced to solve image search problem, and one of the approaches which proves to deliver the highest results is the deep learning CNN model. This paper presents the approach of using multi-deep learning algorithms and similarity measurement. The problem is solved with three pre-trained CNN deep learning models: RestNet50, RestNet101, and VGG19 to extract features, then based on these features to calculate the cosine similarity between images to find the mostly similar images with the given query image. We obtain some encouraging results from several experiments on flower dataset. The results show that the CNN method has succeeded in supporting the retrieval task and therefore has huge potential for practical applications.

Meander line antenna (MLA) with defected ground structure, which resonant at 2.5 GHz, has been designed and examined in this paper. The antenna was built on a FR4 (r = 4.5) substrate with a thickness of 1.1 mm and a loss tangent of 0.025. To evaluate the antenna's performance, features were used operational bandwidth, gain, return loss, and radiation pattern. We achieve a return loss of −17 dB, a bandwidth of 57 MHz, and a gain of 3.21 dB using defective ground structure (DGS). The antenna is 34 28 1.1 mm3, which is a relatively small space.

The paper proposes an innovative approach to solve the cooperation and path planning problem of multiple mobile robots in clutter environment. The main emphasis of the work lies in designing a multi-objective fitness function for stick-carrying robot pairs to compute a collision-free optimal path. The present context of the paper address the multi-robot cooperation and path planning of two pairs of stick-carrying robots that move from a predefined initial location to a pre-assumed goal position by carrying a stick at either end. The basic cuckoo search (CS) algorithm is modified concerning the step size and the scaling parameter at each step of movement of the robot pairs. The modified cuckoo search (MCS) algorithm is implemented with the robot pair to mimic the egg laying behavior of the cuckoo for producing the next generation solution. The proposed algorithm is validated using computer simulation and has been compared with other existing approaches such as ICFA, CS, SDA, and ABCO. Due to its simplicity and efficacy in terms of path optimality, the proposed algorithm produces an optimal solution both in the static and dynamic environment irrespective of the number of obstacles.

Applying technology to education to improve efficiency is one of the inevitable trends. In higher education, augmented reality (AR) is strongly used due to its advantages. AR provides the ability to apply simulation models in teaching to increase the visualization as well as extend conditions for students to practice more and more. Derived from that fact, we have research motivation for applying AR technology in education, especially for higher education. In this work, we propose a procedure to build up an educational AR application as well as a related technology framework. We also build up an AR application following our proposed procedure and technology framework. This application allows users to interact with machine elements. Furthermore, we model machine elements in automobile major which are available on GitHub and free.

An antenna for wireless communication applications, which is a simple multi-band planar antenna is offered. This work introduces an open slot antenna supplied by an E-shaped monopole for use with mobile and wireless LAN services. The objective of this paper is to design printed antennas suitable for use in LTE mobile stations. The big challenge is to obtain small frequencies from an antenna of a small size because the inverse relationship between antenna size and frequency, safe for human use, suitable for use in DVB, operates for most of the mobile applications, and with a wide bandwidth. The antenna size is 42 × 33.8 × 1.5 mm3. We added three branch lines for the proposed antenna to accommodate the Digital Video Broadcasting DVB. The antenna’s length is adjusted to 500 MHz to support DVB bands; it runs in five bands: 470–530 MHz, 666–750 MHz, 862–980 MHz, 1.37–2.88 GHz, and 3.15–3.52 GHz. The SAR computations are performed using the commercial program CST 2014. It is worth noting that the experimental measurements were compared to the simulation results and show a high degree of compatibility.

Gender recognition based on acoustic attributes plays an important role in various audio forensic level tasks. When we talk about forensic level issues, accuracy is the most prominent attribute that needs to be taken care of. This article shows our attempts to observe the impact of multiple folds applied to the popular KNN classifier on the accuracy of results while recognizing the gender of the speaker. We demonstrate our experiments by using python programming language on the dataset available on Kaggle. The results show that 20-folds KNN can provide maximum accuracy (95.77%) and saturate afterward with the size of the dataset up to 3168. Results also show that changing the number of nearest neighbors in this algorithm will not put any impact on the accuracy.

Cryptography defines different methods and technologies used in ensuring that communication between two parties over any communication medium is secure, especially in presence of a third part. This is achieved through the use of several methods, such as encryption, decryption, signing, generating of pseudo-random numbers, among many others. Cryptography uses a key or some sort of a password to either encrypt or decrypt a message that needs to be kept secret. This is made possible using two classes of key-based encryption and decryption algorithms, namely symmetric and asymmetric algorithms. The best known and the most widely used public key system is ElGamal. This algorithm comprises of three phases, which are the key generation phase, encryption phase, and the decryption phase. Owing to the advancement in computing technology, ElGamal is prone to some security risks, which makes it less secure. The following paper previews combination of three paring function used to enhance the ElGamal algorithm and increase its security. The results showed that the modified algorithm gives 93% accuracy.

In December 2019, a deadly virus named SARS-CoV-2 started spreading in the regions of Wuhan, Hubei, China. The number of coronavirus patients gradually increased in Wuhan, and by 20 December, it reached 60 and 266 by 31 December. Till now, there have been more than 40 Lakhs deaths due to Covid-19. This deadly pandemic gave a setback to most people all over the world in terms of losing their loved ones. Apart from that, this pandemic mentally affected a lot of minds. Social illness and loneliness have been linked to poor mental health by a broad body of research, and data from late March suggests a negative increase in mental health. There had been news of people committing suicides or some going under depression all because their social life was cut down and all they did was question their life choices, their existence, their personality, and their achievements which ultimately trapped them in those intrusive thoughts that kept popping up again and again—which made them disturbed or even distressed. The objective of this paper is to analyze and categorize the mental states of people from all over the world in order to raise mental health awareness, particularly during COVID-19. We used the big data approach to display the surge in sadness and suicidal ideation in terms of the increase in the frequency of certain words. To continue with this problem statement, we will examine text data and learn what words are utilized in virtual suicide/depression notes utilizing two subreddits and NLP tools.

Nowadays, the lithium-ion battery pack (LiB) is used as the main power supply for electric vehicles (EV). The remaining energy of LiB is the very important parameter determined continuously by estimating LiB’s state of charge (SoC). SoC estimation is one of the main functions of the battery management systems (BMS). This article presents the use of two sigma-point Kalman filters (SPKF) to estimate accurately the SoC of the LiB based on the second-order model of the cell. The LiB’s average SoC and the zero bias of the current measurement through the LiB are estimated by the first SPKF, while the second filter is applied to calculate the SoC differences between LiB’s average SoC and the modules’ SoC in the LiB. To improve the SoC accuracy of the LiB modules, a second-order RC equivalent circuit model (SECM) of the cell is used, and the influences of temperature, voltage hysteric, measurement errors, and zero bias of current measurement on the SoC estimation of the LiB are taken into account. To verify the method, the experimental test is conducted in the LiB with cells connected in parallels and series. The simulation and experimental results are analyzed to prove that the SoC estimation of the modules in the LiB is higher accuracy, and the LiB’s average SoC errors are less than 1.5% at different temperatures ranging from − 5 to 45 ℃. The calculation time consuming is shorter, and the calculation complex is reduced significantly.

Due to light refraction effect, the need to design a proper technique for underwater scene is a critical task. The problem began to be appeared at deeper levels of water and hence, one cannot distinguish nor observe the properties for underwater objects. The properties may include dimensions, color, texture, etc. So, the current paper involves such situation through designing a system prepared for this purpose with the use of two types of water; an ordinary type with an addition of alum and distilled water. The presented work aims to measure object’s properties of three different objects immersed inside such system. Results show a complete match between the actual and estimated values for the relationship between the two used laser spots distance-water depth variation for the two used sources of water. In addition to that, a special convergence appeared clearly upon the actual and estimated values in detecting underwater object's properties especially at water depth that exceed one meter.

Digital image watermarking is the most interesting and active field for research as it prevents unwanted access to multimedia data. The trade-off between imperceptibility, robustness, capacity and safety must be maintained for the conception of an efficient and strong digital picture watermarking system. Different studies have been conducted in order to ensure that these needs are hybridized by many domains, including spatial and transformational fields. An analytical analysis is performed on existing digital picture watermarking systems in this research. The digital information that has resulted in the request for a safe ownership of the information may recently be readily changed, reproduced, distributed and stored. The watermark solution for the authentication of content and copyright protection is quite good. This paper discusses basic concepts and features of digital watermarking, important attacks on watermarking systems, general embedding and extraction processes for watermarking marks, and important techniques for the transformation using machine learning are analysed. The objective of this paper is to provide an ephemeral study and background on the definition, and idea and major contributions of watermarking the techniques are classified according to different categories: host signal, sensitivity, robustness, kind of watermark, essential data for extraction, processing domain and applications.

Finding the shortest path in the road networks is an urgent issue for the vehicle driver to reach their destination in the shortest time which leads to consuming less fuel. In this article, a framework has been suggested to direct vehicle drivers to their destination in a shortest time taking into account the actual distance and the traffic jam of the road network. The framework consists of three stages: data clustering stage, merging stage, and determining of shortest path stage. The experimental results illustrate that the proposed framework has similar accuracy (98%) to find a shortest path as compared with existing work; on the other hand, the running time of the RN-CMS was the shortest running time; it has been achieved (1.05 s) for 100 inquires.

The pandemic during COVID-19 has had a negative influence on the world's fabric, including health systems, travel, living and working habits, and economies in numerous countries throughout the world. Furthermore, it has had a significant negative impact on continuing global attempts to curb excessive usage of plastic materials. The extensive usage by healthcare professionals and the overall community, of masks, sanitizers, and synthetic-based personal protective equipment (PPE) kits, has resulted in massive amounts of plastic trash, with no effective measures or policies in place to reduce its severity. Wearing a face mask as a way of protection against COVID-19 has become commonplace. However, because present mask disposal techniques (i.e., burning and reclamation) produce dangerous chemicals, huge production of contaminated face masks causes environmental difficulties. Furthermore, disposable masks are prepared of a variety of materials that are either non-recyclable or difficult to recycle. Therefore, as a result, it is critical to comprehend the scope of the problem and, equally essential, to devise a viable solution to contribute to the creation of a sustainable civic society.

Development of a method is based on wireless sensor for determining pollutant material in water and agriculture products. The identification is based on a smart algorithm. Pollutant material detection by sensors has many properties such as being easy to achieve and very cheap. And at the same time, the detecting mechanism gives result approach to real analysis with mini difference in concentration. This detection method describes the analysis of different pollutant materials that affect human health and may cause health risk such as cancer disease. The suggested method depends on three different types of sensors to collect the data. The principle work of the system is based on the attitude and behavior of ions at different temperatures; the collected data processed using microcontroller to get digital signal and the reading have been dealing with it to get a comparison between control sample (free sample) and measured sample. A smart system is used to process the data for training the system to distinguish between two pollutant materials by using advanced algorithm. The detecting pollutant materials reach 98% at different concentrations.

The reference generator in the robot control system depends on unknown robot–environment interaction. In this paper, an adaptive reinforcement learning (ARL)-based optimal control is developed to achieve the solution, which considers an unknown environment as a linear discrete-time system. A performance index that establishes the interaction effects of trajectory tracking error and external torque is optimized using admittance adaptation. The proposed reference generator is designed in a completed control system of robot systems under unknown robot–environment interaction. Simulation studies are conducted to show the effectiveness of the proposed solution.

With the advancement in technology, the importance of person recognition in photographs or videos has grown due to its usefulness in the search for wanted persons and criminal identification using various theories and algorithms and various non-hybrid and hybrid techniques to identify the person through the face, and its features have been developed. The paper proposed a hybrid algorithm to improve the performance of person identification in video files. The system works through several steps: first, face detection using Viola–Jones algorithm; second, feature extraction by algorithm local binary pattern (LBP); final, person identification by hybrid proposed algorithm (HPBFF) by hybrid between backpropagation neural network and firefly algorithm. The results show that the system was able to identify and monitor the person with a high classification accuracy rate of 98.4%, compared to 94.7% for the approach without the hybrid. The results of the tests revealed that the system is robust and has a high recognition rate, making it suitable for use in mobile and compact identification and authentication.

With the rising population and vehicular traffic across the globe, driver safety on road has become a huge concern for most governments. Emerging technologies and industrial revolutions have given rise to concept of autonomous cars. The driving systems embedded in these cars identify the traffic signs on the road and then take appropriate action. In spite of all these efforts, the accuracy of traffic sign image detection still remains a challenge for most car manufacturers and drivers, especially under difficult weather conditions. Multiple authors have done research in past and have proposed approaches relevant to identification of traffic sign images. The proposed solutions on traffic sign image detection have been influenced largely by Artificial Intelligence (AI)-based implementation techniques. In this research paper, authors have used Mapillary public traffic image dataset and have proposed an innovative approach using chi-squared ranking algorithm along with ANN for image classification. The effectiveness of proposed approach is compared with some related works. Experimental results showed that the proposed enhanced algorithm based on ANN and chi-squared algorithm provided better results.

This brief presents a completed control structure, including formation control and adaptive reinforcement learning (ARL) algorithm for a multi-agent system of multiple surface vessels (SVs). The ARL strategy is established for each SV with the advantage of handling a non-autonomous system without solving the Hamilton–Jacobi–Bellman (HJB) equation. The additional formation controller is implemented to complete the control structure of multi-SV systems and guarantees the formation tracking problem. Simulation studies are developed to show the performance of the proposed control structure.

Mittal, Shilpi Prasad, KantaWith the advancement of cloud computing and edge computing in traditional vehicular network, it has improved the QoS in VANET. When these heterogeneous networks use resources of clouds for computing and communication, it faces an issue of anxiety and latency due to long distance from the location of data generation. To concern this issue, we have implemented a data dissemination scheme compared with cloud and edge computing services in vehicular network. In this paper, architecture design with static and dynamic vehicle movement is illustrated and a data dissemination scheme is proposed on real-time scenario with the testing of QoS using simulator. The proposed scheme shows better result in terms of low latency and enhanced throughput.

Bondugula, Rohit Kumar Sivangi, Kaushik Bhargav Udgata, Siba K.Human activity recognition helps identify the activity of a person based on data provided by sensors. The wireless wearable sensors provide robust techniques for data collection and classification. Most wearable devices contain heart rate and body orientation detection sensors. In this work, we experimented on a Photoplethysmography sensor used for heart rate identification and accelerometer signals to recognize the subject’s orientation to classify the human activities. We proposed a novel deep learning architecture (MiniVGG) which was able to find the right activity time interval that resulted in overall lower false positives and false negatives. Our proposed model, MiniVGG, gave the highest accuracy of 97.75 $$\%$$ % on the PPG dataset higher than any other existing models. The results of our experiments are compared with other baseline models and at varied sampling time window sizes and have shown greater accuracy. In addition, we report the best combination of the sampling time window size and the appropriate model to achieve the best accuracy, minimum false positives, or minimum false negatives depending on the requirement. This helps in developing a multi-criteria decision-making system for human activity recognition system using wearable sensor devices.

Watermarking implies embedding of data in a manner thatit is easily accessible by a user that’s authenticated rather than other users, bearing in mind that the underlying data may experience some changes as a result of such data embedding. Reversible watermarking has been developed in advance of watermarking, whereby both of data quality and data recovery are guaranteed. For long time, the watermarking scheme to meet image, audio and video data type security was considered ever popular, but at the same time, since last several years ago relational database watermarking was really in mind and under study, even various watermarking approaches have been suggested. Nonetheless, these techniques yet are not sturdy enough against malignant attacks that can result in false insertion, modification or elimination leading to data performance and quality deterioration. So, and with aim of advancing and improving existing watermarking techniques, a new robust and reversible relational database watermarking method is suggested by which better outcomes were achieved.

The paper presents novel the adaptive beamforming in linear antenna array through hybrid optimization algorithm. The hybrid optimization algorithm has been formulated by combining improved version of the whale optimization algorithm (IWOA) and improved version of sine–cosine algorithm. Application of hybrid algorithm in the beamforming is to estimate the excitation weights of the desire signal applied on array elements, different interferences received from different directions and update the position the receiver so as to receive better quality of service by adopting the weight of the input signal. The robustness of the algorithm is confirmed through the simulation in MATLAB, and result shows that performance has been improved by minimizing bit rate, power transmitted, beamforming through different number of array elements using hybrid optimization algorithm.

An emerging technology Internet of things is the backbone for better solution in medical science research, COVID infected sampling analysis, and device integration process. 2020 may be a year of healing not only for our mother earth, but for mankind too. It is a year of change and practice to develop ourselves against adversities. Due to the recent pandemic caused by COVID-19, many lives were affected. COVID-19 has created an adverse effect on the economy, education, mental health, and physical health of humans. It has been witnessed that despite lockdown, the death rate has increased. From several statistics, it can be observed that populations with less immunity have a higher mortality rate. This study has been performed to make a checkpoint on the factors which may be responsible for determining immunity level and based on that a prediction model may be prepared using a machine learning algorithm. This proposed work employs an IoT application to collect real-time symptom data from users to identify suspected coronavirus symptoms. IoT’s sensor-based mechanism adopts for enhance capability of risk minimization specially in surgery cases compilation for COVID-19 type pandemic.

In the current epoch, the Internet of Things (IoT) can be reflected as an important technological revolution related to evolution of smart cities, smart homes, IoT-controlled factories, and IoT for logistics in seaports implementations. With the existence of smart sensing systems in seaports becoming a reality today, different sectors in seaports are working toward a programmed mode. Some of the eye-opening projects related to smart seaports in the IoT era can be found all over the globe. In many of these new architecture implementations, even though the rapid development of IoT enables us to inspire new research works, the challenges in IoT also grow equally in terms of security. Encryption plays a key role in safeguarding IoT hardware and the data from various sensors. The proposed work focuses on reality study on various security issues emerging in the usage of IoT in seaports and suggestions for handling the security issues. Highly secure algorithms need developed in IoT encryption level standards are discussed here. Conclusions regarding the extension of future research prospects in the IoT systems and high-level security in seaports are guided in the final segment of the paper.

The objective of the paper is to scheduling of independent task dynamically among the virtual machines (VMs) in the cloud network on the share resources. Scheduling of task and allocating of the resources from the data center have been performed through the several meta-heuristic algorithms and achieved encouraging results. However, their performance evaluation is far based on the ideal state and needs more improvement. Load balancing in necessary when some VMs are executing more number of task and task need to wait in queue for processing at the same time other VMs are free and not allocated any task or less number of tasks for execution. The problem under consideration is proposed sunflower optimization algorithm with sine–cosine algorithm (SFOA-SCA) for improving the performance of load balancing in cloud network. The experimental result illustrates that the projected procedure is outstripping its opponent in the manner of throughput time, waiting time, response time, execution time, and utilization energy during load balance of task in cloud network.

This article provides an overview of the benefits and the limitations of using exploratory data analysis for research questions surrounding explosion incidents. Most importantly, it provides a starting place for those interested in learning about this type of analysis and considerations that users should be aware of. Those unfamiliar with exploratory data analysis may find the article helpful in understanding what will be covered during their dissertation process. Terrorism is a term we use to refer to violence and intimidation in the pursuit of political aims. Terrorism has been around for about two centuries, but it did not gain much traction until the 1970s. Terrorism is defined as violent acts against innocent people committed to making an impact on society or government. The underlying causes of terrorism are usually motivated by injustice, revenge, oppression, retaliation, conflict, and hostility borne out of inequality among countries. Furthermore, it is associated with religious extremism and usually targets civilian populations to cause death or destruction. As a result of these patterns in manifestation and potential consequences worldwide, terrorism has become a pressing global issue that requires attention at multilateral levels. In this article, we have tried to cover all the aspects of terrorist data analysis using various machine learning algorithms like support vector machine and linear regression. After analyzing the above datasets by the said model, we have figured out which region is mainly affected by attacks and where the cases of terrorists have been seen frequently. Here, we have tried to explain what it means to do this kind of research and provide some considerations for those who want to use this approach for their work.

As a result of their surroundings and lifestyle choices, people nowadays suffer from a wide range of ailments. As a result, predicting illness at an early stage is crucial. Doctors, on the other hand, struggle to make accurate diagnoses based solely on symptoms. The most challenging task is predicting sickness properly. Machine learning plays a key part in forecasting in order to complete this difficult task. To tackle this challenge, machine learning plays a key role in illness prediction. Medical research creates a vast amount of data every year. Early patient care has benefitted from effective medical data analysis because of the rising quantity of data growth in the medical and healthcare professions. In data mining, disease data is utilised to identify hidden patterns in huge volumes of medical data. Based on the patient's symptoms, we created a broad disease prediction. Machine learning algorithms like ANFIS and CNN are used to properly predict sickness (adaptive network-based fuzzy inference system). The collection of illness symptoms is necessary for disease prediction. For an accurate prognosis, this general illness prediction takes into account the person's lifestyle and medical history. When it comes to illness prediction, ANFIS outperforms CNN by a wide margin (96.7%). ANFIS, on the other hand, does not require as much time or memory to train and test because it does not use the UCI repository dataset. There are several libraries and header files included with the Anaconda (Jupyter) notebook that make Python programming more precise and accurate.

With an expansion in the demand for automated medical imaging, the field is getting importance, fast, reliable and efficient diagnosis which can provide insight to the picture image better than human eyes. Brain tumor is the second leading cause of cancer-related deaths in men age 20–40 and 5th leading cause cancer among women in the same group. A diagnosis of tumor is a very important part in its treatment. Identification of a tumor is very important part in its treatment. To obtain the background, this paper covers noise elimination and image sharpening and also morphological functions, erosion and dilation. Plotting contour and c-label of the tumor and its boundary provides us identifying the size, shape and position of the tumor, it helps the medical employee as well as the patient to understand the seriousness of the tumor with the help of different labeling for different levels of elevation.

Because of the importance of software refactoring for software code quality and stability, this research primarily emphasizes whether refactoring can be vital to identify probable software components for future refactoring. Modularity, reusability, modifiability, maintainability, and service-oriented development may all be improved with refactoring. This fact encourages academics to develop a new and improved machine learning paradigm for restructuring OO software. We have made a multi-purpose optimization effort to assess the OOP-based software systems or components refactoring in this work. This research intends to exploit and optimize OOP software metrics to examine code quality by performing refactoring. Our objective is to develop a highly resilient and efficient ensemble computing model for refactoring prediction at the method level into a machine learning framework using software metrics as features. The focus is on applying enhanced state-of-art data acquisition, data preprocessing, data imbalance resilient re-sampling, feature extraction, and selection, followed by improved ensemble-based classification. This work will also focus on the types of project work for different kinds of classification.

Agriculture has a captious part in maintaining a large population. It plays a decisive role to forward our country’s economic development. Crop cultivation has been the most prominent problem in recent days due to changes in weather patterns. This has a significant impact on crop productivity, either directly or indirectly. As a result, new technologies might be brought up for use in order to overcome this problem and uplift crop output. In this research proposal, we have explored IoT by choosing the smart farm and digital technology and explaining the management of heterogeneous data for agriculture. We proposed an IoT-HELE-based smart farming prediction and intelligent agriculture analytics model and a decision support system that effectively predicts crop production by utilizing cutting-edge machine learning and deep learning techniques. In this model, ensemble voting results in a more efficient, sustainable, and profitable agriculture enterprise. The multi-source dataset from the National Research Council (CNR), an ISTAT, and an IoT sensor will be analyzed. This work is presented through a new innovative idea after a rigorous literature review; presumably, it is valuable and increases the productivity of an agricultural firm.

Heart disease becomes one of the most influential diseases that cause a large number of deaths every year around the world. A report by WHO shows in the year 2016 nearly 17 million people gets died due to heart disease every year. The death rate is increasing rapidly day-by-day and it is estimated by WHO that this death ratio will reach the peak of 75 million by 2030. Despite the availability of modern technology and health care system, prediction and diagnosis of heart disease are still beyond the limitations. Currently, the clinical industries and diagnosis centers have a huge of amount data for the diagnosis of heart disease patients. Machine learning algorithms are more useful to find the hidden patterns, discover knowledge from the dataset, and predict correct outcomes. This research proposed an efficient machine learning-based classifier methodology that outperforms the existing similar methodologies. To evaluate the proposed machine learning classifier, we have taken data from the UCI repository. In this study, we have used ZeroR, bagging, M5, and decision table classifier. The M5 classifier produced a good result compared to other classifiers with 0.2726 mean absolute errors.

Encroacher detection machine ought to obtain proper precise and adaptableness to retribution attacks from encroacher. Encroacher detection machine distinguishes among the valid and criminal users and ought to be used with the primary line of security to thwart encroachers and aberrations from inner in addition to outdoor attackers. The encroacher detection machine is a critical asset to PC security, due to the fact that attacker attempts to hide his/her identity, and release attacks through intermediate hosts broadly called stepping stones encroacher. Secondly, converting creation of generation and approach makes it extra hard to come across attacks. These cannot obtain by existing methods, those various community attacks are modeled via way of means of using different types of traffic, that is, expensive and complex. The initiate encroacher detection machine can consequently employ optimization algorithms strategies in gadget studying to come across the unknown destiny assaults. In the exploration, a brand-new version is initiate for detecting the future encroacher attacks. Experimental result confirmed that the improved encroacher detection model mixed with deep notion network efficaciously improves the popularity price of encroacher attacks and reduces the complexity of the network. The adequacy of the start approach is set up contrary to current systems through method-of-method for the utilization of following measurements, for example, Accuracy, Sensitivity, Specificity, Precision, F-score, and Mathew's connection coefficient.