Mobile Radio Communications and 5G Networks

Android smartphones have a big share in global market in comparison as it is open-source architecture, high usage and popularity in the community of developers. In general, smartphone becomes a persistent gadget in individual’s life because it is used for various purposes such as office applications, gaming, Internet and vehicle guidance-based services along with basic services like calling and messages. With the high usage of android smartphones and its association with monetary benefits, it has made its point of attraction for attackers. As a result of this, it leads toward the exponential growth in android malware apps. This study has mainly focused upon android platform and aims toward the systematic characterization of existing android malwares. The study includes the types of malwares in android operating system along with characteristics. It also covers the activities that are performed by the malwares. Moreover, case studies of few recently discovered malwares are discussed in detail with their working and threats. This study will help the researchers to obtain the knowledge regarding android malwares and their threats caused to devices. It will help the society to become aware about the malwares that will keep them away from these types of apps.

Electronic voting (E-Voting) has been described as one of the most efficient methods of collecting consensus-based decisions about a particular entity. These systems are useful for a wide variety of application scales, which range from selecting candidates at small corporations to nationwide elections. But voting systems face inherent security and quality of service (QoS) issues, which limits their public-domain deployments. Due to many sources of vulnerability, such as mutability, poor traceability, reduced trust levels, and centralized computing design, these systems are susceptible to attack by hackers and other adversaries. The usage of blockchain-based computing models, in which each set of votes is translated into a transaction, and these transactions are kept inside smart contracts, can be used to tackle these problems. These smart contracts are turned into blocks and kept in a decentralized blockchain database for storage. This database uses an improved unidirectional linked list, where each block is connected to the next block via a unique hash value. Due to the uniqueness of connecting hashes, this model exhibits immutability, which is one of the main reasons for its use in e-Voting systems. Secondly, hashes are generated using a decentralized mining mechanism, due to which the blockchain database is stored on multiple nodes, and is resilient against denial of service (DoS), Sybil, masquerading, and other server-based attacks. Similarly, the blockchain model also possesses transparency, and traceability, which makes it an ideal candidate for e-Voting systems. But the delay of voting increases exponentially w.r.t. the number of transactions, which is due to the fact that the addition of each block requires mining nodes to generate a new unique hash, which requires scanning of the entire blockchain. This limits the scalability of the blockchain model, which makes it unusable for larger-scale networks. In order to remove this drawback, a novel sidechaining mechanism is proposed in this text, wherein sidechains are created and managed using a firefly optimization model, which uses a number of parties and cast votes per party parameters. Due to this dynamic model for sidechain creation and management, the proposed method is capable of reducing transaction delay by 28% when compared with a single blockchain, and 16% when compared with static sidechain methods. Additionally, the model was tested on medium to large-scale e-Voting applications, and it was discovered that, when compared to other cutting edge models, it is capable of improving throughput by 8% and reducing storage cost by 18%. The proposed sidechain paradigm can be used for a wide range of e-Voting application deployments as a result of these benefits.

The task of recognizing or locating an object or group of objects in an image or video sequence is the focus of object recognition in the field of computer vision. It's a problem of matching models from a database with image luminance data representations of those models. The paper presents the novel method for the detection of vessels in complex environments. The present work is inspired by cuckoo search algorithm which was used to optimize the results. Otsu thresholding method was used to binarize the object image and then adaptive cuckoo search algorithm was applied for the detection of vessels in an optimized manner. In this work, we are comparing the proposed results with existing accuracy and time with other methods. The proposed method gives a better accuracy (97.8%) and processing time (3.67 s) as compared to other previous methods used.

Data handling is an important task in cloud computing because the data arrive at a high rate with a large volume. These data are stored in the cloud storage for consistent performance. Cloud offers services to the customer by allocating the virtual machine (VM) for requesting tasks. Cloud provider always satisfies the customer at anytime and anywhere manner. The services depend upon the resources at the data center through virtualization. Cloud follows the strategy called demand-based and pay-as-you-go basis. The customer must pay for the resource consumed using metering services. Customers store large volumes of data which occupies more space in storage. It leads to expensive problems for the customer because of excess payments paid to the cloud provider. This type of storage suffers a high volume with redundancy in the uploaded data. This problem is overcome by using the deduplication technique for keeping only one copy of data in the cloud storage. This achieves less storage, so the cost of services reduces drastically. The integrated data elimination with the cost estimation method has been introduced in this study to achieve better efficiency and availability. This process has been carried out by using two levels of elimination techniques such as local based and global based. This data are integrated into datasets without any redundancy in a data-centric manner. Integration of eliminated datasets is combined as a global level of elimination process by handling newly generated data. Finally, the data are stored in the cloud storage without any repetition in data.

In wireless sensor networks (WSNs), the protocols for routing have a great effect in performance of network as network’s lifetime, good energy organization, etc. These protocols are developed based on the different schemes like clustering, chaining and cost based. The WSN contains a large number of nodes which are sometimes difficult to manage. So, the best way is to make a cluster by combining various nodes, and this technique is known as clustering. By doing so, the energy exhaustion by nodes can be restricted. A node is nominated as cluster head (CH) to handle the communication amongst nodes and managing of nodes in the cluster. An ACO-based probability rule is used for choosing the CH amid the cluster nodes. The data are sent from cluster nodes to the CH; then, it further sends the relevant details to the base station (BS). ACO-DEEC: ant colony optimization-based distributed energy efficient clustering protocol is used for probability rule calculation for CH selection depending on the metrics, i.e. efficiency of the nodes and distance amongst nodes. The algorithm proposed enhances the parameters like energy consumption by cluster nodes, finding dead nodes in cluster and quantity of packets sent to the BS as compared with existing DEEC protocol.

The Named Data Networking (NDN) is a future Internet architecture to support content centric delivery. Among all the features offered by NDN communication such as effective network bandwidth utilization, name-based content routing, and so on, in-network caching plays a major role for achieving energy efficient content delivery. The caching of requested contents at network edge is incredible due to reduced content retrieval delay and network traffic a requester has to experience. Although, a number of caching strategies for content caching at edge exist in literature, still caching of popular contents with certain lifetime at edge network are less studied. Therefore, this paper proposes a novel idea for NDN node to take cache decision based on requested content popularity and its data lifetime. The data with the highest popularity and greater residual lifetime would be preferred for caching. The proposed approach has been evaluated in Icarus simulator for different performance metrics. The results retrieved from experimentation proved outstanding performance of the proposed approach over the existing state of art strategies.

This article represents performance of segmentation algorithm of an unsupervised image based on hidden Markov random field (HMRF) model and random walk segmentation algorithms for gallbladder MRI. Detection of lesions in human brain is an important task aimed at saving precious lives. The novelty of the method of segmentation is the use of random walk algorithm together with entropy maximization. We employ entropy maximization to automatically identify the seed points that are to be used in random walk algorithm. Detected early diseases can be healed by eliminating slices of human organs. Specific symptoms of the disease are missing, and the cancer remains undetected until it has spread to all the organs. Hidden Markov random field model is used to segment gallbladder lesions. Expert medical opinion is required to determine whether the lesions have cancer. Here, we analysis the performance of both the algorithm HMRF-based unsupervised segmentation and random walk segmentation.

Breast cancer is a very common type of cancer found in women. Approximately, 43,000 deaths are recorded per annum worldwide due to breast cancer. With the advancement in medical technology, computer-aided detection and diagnosis (CAD) system is being used widely for the early detection of cancerous cells. Rapid development in deep learning has made the task of detecting cancerous cells accurate and trivial. In this paper, researcher used convolutional neural network (CNN) for classifying cancerous cells. CNN is a type of neural network which is extensively used for image processing, classification, and segmentation. The proposed system has achieved 82% accuracy by successfully classifying cancerous cells into benign and malignant which are the two common types of cancer cells found.

Machine learning provides a plethora of approaches for tackling categorization problems, i.e., determining whether or not a data item belongs to a particular class. When the objective is to accurately classify new and unknown data, neural networks are frequently an excellent choice. The widespread availability of growing processing power, along with the development of more effective training algorithms, has enabled the application of deep learning principles. To provide superior learning solutions, deep architectures leverage recent breakthroughs in artificial intelligence and insights from cognitive neuroscience. Convolutional neural networks (CNNs) are a subclass of discriminative deep architectures that have demonstrated acceptable performance when processing 2D data with grid-like topologies, such as photos and videos. In this paper, we compare the performance of deep CNNs and machine learning for diabetic foot disease categorization. Diverse machine learning approaches were implemented for classification, viz., decision tree (DT), support vector machine (SVM), quadratic discriminant analysis (QDA), K-nearest neighbors (KNNs), AdaBoost (AB), Gaussian Naïve Bayes (GNA), logistic regression (LR), extra trees (ET), random forest (RF), histogram gradient boosting (HGB). Further, deep convolutional neural networks (CNNs) and transfer learning-based Inception ResNet V2 algorithms were used to analyze in the context of deep learning implementations. In this work, the classification of diabetic foot disease through plantar thermograms was conducted using deep learning implementations. The data augmentation is also done to address the paucity of data. Here, deep learning-based models with augmented dataset prove better outcomes.

Health monitoring framework based on Internet of things (IoT) has been introduced as of late for improving the quality of medical care administrations. To control glycemia, patients analyzed with diabetes mellitus should screen their blood glucose levels frequently. Subsequently, they must play out a capillary test at least three times each day, also a laboratory tests few times each month. These principal strategies present trouble patients as they need to undergo the invasive method of pricking their fingers to figure the glucose level, yielding uneasiness and trouble. Further, patients affected by diabetes may have a possibility of damage in their nerve conduction system. An Internet of things (IoT)-based system for non-invasive blood glucose monitoring with a simultaneous nerve study system has been proposed in this paper. An Arduino controller with a sensor node is used for monitoring and data transferring to the cloud.

In the today’s world, waste management is a prevalent problem and is increasingly growing with the growth in urbanization. In many nations, waste management is a crucial component of environment development. Enhancing management needs are usually recognized by officials in India like developing countries. Nevertheless, little effort is made to improve the situation and long-term changes. We know that 19.6% of the population of India is equivalent. A smart garbage management system is necessary with the development of intelligent cities throughout India. Since the quantity of waste grows every day. The best approach to dealing with this problem is essential, because the waste generated exceeds 2.5 billion tons. To enable the dumping sites to confirm that garbage is organized of appropriately, the waste must be sorted in a basic method. Waste sorting requires more staff and also takes more time. In numerous techniques, waste could be managed and sorted. Processing of image to analyze and categorize waste can be a highly effective way in the process of waste materials. This paper discusses the model based on MLP and Naïve Bayes for waste segregation. These also discuss the disadvantages and ways to overcome existing systems. The paper also presents a design for system for reducing physical effort and promotes automatic separation of waste.

The healthcare business deals with overly sensitive information that should be handled with care. Names, social security numbers, insurance numbers, addresses, and medical history are all stored in Electronic Health Records (EHRs). Patients, healthcare professionals, medical insurance companies, and research institutes all value such personal information. Patients and healthcare providers, on the other hand, face substantial privacy and security risks as a result of the public dissemination of this extremely sensitive personal data. When health information is transferred, EHR management systems necessitate the use of effective technology. Medical record storage systems are frequently exposed to typical security attack vectors due to current management techniques. As a result, we anticipate the need for new solutions to solve the security and privacy issues that private data poses in healthcare applications. Decentralized, anonymous, and secure EHR handling is possible with healthcare-focused blockchain systems. The needs for privacy and interoperability in healthcare data exchange are examined in this research, and a blockchain-based solution is offered. We used private blockchain technology to examine their suitability for healthcare applications in this paper. The permissioned blockchain structure of Hyperledger Fabric is used in the article. The proposed approach may effectively preserve patient details while maintaining anonymity. The findings of this study show that blockchain technology may be used to promote privacy while also improving interoperability in healthcare information management.

Mobile ad-hoc networks are vibrant and incessantly varying ad-hoc networks, so having centralized checking on the same is impossible. Vehicular ad-hoc network (VANET) is like mobile ad-hoc network (MANETs) where vehicles keep on interconnecting with adjacent cars and roadside units. Here in this paper, we use mobile ad-hoc networks, which combine various mobile nodes, which reduces interruption-however could not remove them. Intrusion detection in MANETs is an assignment associated with the machine learning area. In this paper, our primary focus is mainly on reducing features for achieving extreme precision and reducing the overhead time of machine learning algorithms. The interruption dataset is reserved from the customary dataset of named Network Security Laboratory-Knowledge Discovery and Data Mining (NSL-“KDD”). Here in this paper, the genetic algorithm (GA) is substituted by the gravitational search algorithm (GSA), which delivers the maximum precision and consumes significantly less time for training the model.

The devastation of crops due to various ailments has become one of the biggest threats in the agriculture sector. The necessity of early diagnosis and preventive measures can limit the effect of the disease on overall yield. To predict the condition of the plant, the disease severity is estimated based on the affected area of the leaf. This paper proposed 4-score disease severity classification architecture with the integration of Convolutional Neural Network (CNN), and Long Short-Term Memory (LSTM). CNN has made significant breakthroughs in extracting the feature from an image. The input of convolutional layers is inputted to the LSTM layer. The result analysis shows a promising improvement in the classification performance by including LSTM layers after Convolutional layers in the proposed model. The proposed classification approach is evaluated using a dataset consisting of 1856 images of bacterial blight disease. The images are collected from st, standard online repositories. The accuracy of the classifier is significantly higher than that of a single approach model. The proposed hybrid model of CNN-LSTM achieved an accuracy of 92%, 88%, 86%, and 94% for 4 classes of severity. The overall accuracy of the model is 92%.

The Internet of things is playing a vital role in human life as well as society. Many mandatory services are provided by IoT devices like GPS, cab services, healthcare, weather forecast and the list is too long. Our life depends a lot on IoT devices. But these devices use sensor technology and data streams for acquiring the different types of information. With the growth of IoT devices based on IoT DDoS attacks is increased. Machine Learning techniques are already in use for detecting malicious network traffic. In this paper, we present the Convolutional neural network technique to detect botnet traffic in IoT devices and networks. This model shows enhanced accuracy results thus causing low loss. CNN is a capable automatic hierarchy of learning the most relevant features. This paper achieved a higher accuracy rate of 99.98% result applying Convolutional neural network (CNN). It presents the feasibility of DDoS attack detection in IoT networks.

MANET wireless network has proven to be a valuable tool in various communication applications. It has risen in popularity in recent years due to considerable advancements and the MANET network design presents the maximum difficulty. Recently, concerns about latency and availability have been worsened because of continual form and characteristic modifications, which have resulted in terms of performance and service quality difficulties. Self-organizations are the ones who have de-centralized MANET and without centralization, participating nodes are free to migrate. So, this paper undertakes a study to understand IoT for MANET network security where nodes can transit from being a host to become a router at any moment. This complicates the transmission of data packets among nodes in a de-centralized mobile ad-hoc network. Due to the nodes’ proclivity for self-organization, MANET networks offer both advantages and disadvantages. This along with simplifying network maintenance also changes the topology, but data transit must be authorized. MANET may also be used to connect to bigger networks like the internet. However, there are no intelligent devices that can transmit data among machines.

Tomatoes are very essential staple crop that is consumed by millions of people from all corner of the world. But, unfortunately, a huge part of the total tomato crop production is lost annually due to various plant ailments, and manual identification of this diseases is tedious and may need the assistance of trained expert. To overcome these issues, we have concentrated in relation to the usage of a deep learning algorithm according to a convolutional neural network to build a classification system to accurately classify leaf images and identify the disease. The tomato leaf ailment pictures in this paper is obtained from Kaggle database. If the plant is infected with a disease, the disease name is mentioned in the dataset. The dataset contains around 7928 images which are categorized into ten different classes. The dataset was divided into train, test, and validation sets in ratio 8:1:1. The first instance the dataset is iterated over, its elements are cached in memory. Subsequent iterations use the cached data. During a particular iteration, the dataset for the next iteration is also prefetched. The neural network is built by adding the rescaling and data augmentation layers first. Six convolutional and pooling layers were alternatively later applied for feature extraction. Dense layers were used for classification of the data from convolutional layers to the correct class name. The model gave an accuracy of 96.26% over the dataset which is much better than the traditional model.

To identify the variables of academic professionals on job satisfaction and retention. There is little research on academic professionals in private schools, especially in the north region. This study is examined to understand the short format used by the organization to engage its employees. Several factors such as remuneration and leadership will also focus on the pursuit of loyalty and job satisfaction for important employees. The structured questionnaire was received from the faculty members of private teachers. Good Compensation and good leadership give more satisfaction to the employee as well their long time with the organization. With the help, the Job Satisfaction Survey (JSS) questionnaire was used to evaluate the overall satisfaction of the teachers. There is a positive relationship between these two factors. The reliability and validity of the questionnaire were checked on Statistical Package For The Social Sciences (SPSS) and the value is 0.85. Random sampling is used to collect the data from 100 employees of private schools. The result indicates that a positive relationship between compensation and leadership plays an important role to maximize the job satisfaction of employees with the help of AI tools. This paper also contributes to society as well by encouraging the employees to contribute more to the growth of both.

Data Mining (DM) is a process which assists in mining the significant data from the irregular data. The current information is employed to predict the futuristic results in the prediction analysis. The heart disease prediction techniques have various phases such as dataset input, to pre-process the image, extract the attributes and classify the data. The various types of techniques are proposed in the previous decades for predicting the cardiac disorder. These techniques are broadly classified as machine learning, deep learning and clustering. In this paper, all techniques are reviewed in terms of methodology and results. It is analyzed that deep learning techniques are more popular now to predict the heart disease.

Cocoa is worldwide most significant crop grown across the world. The crop generates revenues and employment for countries producing cocoa. Ghana’s cocoa industries play important role in the global cocoa market. Ghana is the second-largest producer and exporter of cocoa beans of worldwide production. In Ghana cocoa industries have large impact on social and economic services. In this paper, propose a mechanism to perform crop testing process for Coco Beans dataset. In cocoa beans dataset six image classes are used such a Bean_Fraction_Cocoa, Broken_Beans_Cocoa, Fermented_Cocoa, Moldy_Cocoa, Unfermented_Cocoa, and Whole_Beans_Cocoa. However, cocoa beans produced are tested for its quality before selling. Quality cocoa testing defines cocoa that has been properly dried, fermented, and is disease free, other physical flaws, and contamination. To perform empirical analysis, we applied three different ML algorithms such as CNN, DNN, VGG16 trained model. The performance of these techniques is analyzed using performance metrics such as data loss and accuracy. The results depict that in VGG 16 model the accuracy is 65.6% in perspective of other two techniques.

Satellite image photography with very high resolution (VHR) presents a significant problem in identifying water bodies. In this work, correlations between extracted features at each scale, which extract the whole target. Using data from several sources, including the immediate environment, a broader geographic area, and the relationships that exist between the various channels, display features. In addition, to better anticipate water bodies’ delicate contours, use Fusion of many scales of prediction. In addition, feature semantic inconsistency is resolved. Encoder-decoder semantic fusion allows us to combine the encoding and decoding processes module for promoting the fusion of features. The outcome demonstrates that our approach is cutting-edge superior performance in the segmentation process compared to the most contemporary and traditional approaches. In addition, have offered methods that are reliable even when used in the most difficult water body extraction situations.

Image-based detection and classification using CNN, its aim to detecting and classification of the plant diseases. Farmers are the back bone of the nation to help them, we present these plant disease detection. Nowadays, many of the farmers are facing these plant diseases for their crops and paddies, and it is the crucial issue to be addressed in the world. In order to detect the plant leaf disease, we address the present prospects and issues. Plant diseases are one of the most serious threats to food safety. Some plant diseases are infectious diseases and parasites that can spread throughout the entire field, affecting the almost all of the yields. It is not only the process of contextual but it is also time-consuming, labor intensive, and un-reliable. Some farmers were use the pesticides for the crops with far less experience. Food insecurity will worsen if plant diseases are not discovered in time. To tackle these issues, researchers are investigating the use of image processing techniques for plant disease recognition.

Air pollution is a major concern nowadays as it affects all living organisms. Air quality is dependent on the pollutants present in the air which include oxides, ozone, carbon monoxide, particulate matter etc. Air pollution is now acknowledged as a major public health problem, causing a growing number of health effects that have been extensively documented by the findings of numerous research conducted throughout the world. The air quality index allows us to rate different sites according to the amount of pollution they have, showing the more contaminated areas as well as the frequency of potential risks. The AQI aids in determining changes in air quality over time, allowing for prediction of pollution and it's mitigation. Prediction of air quality helps people, and organizations in planning; managing various activities. In case of poor air quality, people can take precautions and look into the methods to reduce its adverse effects of it. It helps in protecting public health by predicting air pollutants. This literature review focuses on the various techniques used by different researchers in the prediction of air quality and air pollutants using machine learning. Machine learning techniques have been applied to different areas with various pollutants, and the performance of various machine learning algorithms is compared using performance metrics.

An easy-to-use programming environment, open access to satellite data, and access to high-end consumer computation power has made it very easy to align remote sensing and machine learning during the new era. A variety of remote sensing applications have utilized publicly available data. Land use (LU) image classification has become vitally important in the natural environment because of the expansion of some global changes relating to the temperament of the earth. Therefore, researchers should investigate this area more deeply. This paper presents a complete review to help out the researchers to carry on with the land use/land cover for image classification process, as there are limited numbers of review articles to assist them. The purpose of this paper is to discuss the classification of satellite images using mainly employed machine learning algorithms. We discuss the general process of LU/LC based on multi-image classification, as well as the challenges and issues faced by researchers. Only a few studies evaluate machine learning algorithms for image classification using openly available data, however.

Depending on the seriousness of the offence, any deliberate act that causes harm to oneself or another, as well as damage to or loss of property, qualifies as a crime for the purposes of criminal law. The number and diversity of unlawful activities are increasing at an alarming rate, necessitating the creation of efficient enforcement tools by law enforcement agencies. Due to their slowness and inefficiency, traditional crime-solving methods are no longer effective in today’s high-crime climate. As a result, we may be able to reduce the workload of police personnel and contribute to crime prevention if we develop ways for reliably anticipating crime in advance of its occurrence. In order to do this, we suggest using ML and computer vision technologies and methodologies. Throughout this study, we describe the results of several cases in which such strategies were used, which sparked our interest in further study. The fundamental reason for the change in crime detection and prevention strategies is based on the prior and subsequent statistical observations made by the authorities. Machine learning and computer vision may help law enforcement and other authorities detect, prevent and solve crimes more quickly and accurately, and this is the primary goal of this research effort. With the use of artificial intelligence and computer vision, law enforcement agencies might be transformed.

In this article, we describe our early efforts with sentiment analysis on tweets. This project is meant to extract sentiment from tweets depending on their topic matter. It utilises natural language processing methods to determine the emotion associated with a certain issue. We used three different approaches to identify emotions in our study: classification based on subjectivity, semantic association and classification based on polarity. The experiment makes advantage of emotion lexicons by establishing the grammatical relationship between them and the subject. Due to the unique structure of tweets, the proposed method outperforms current text sentiment analysis methods.

This paper proposes the implementation of BCH Encoder in the ZYNQ-7000 AP SOC to guarantee the sensitive data that is acquired from capsule endoscopy is transmitted without any errors though a wireless medium. The BCH Encoder and Decoder is designed, Synthesized and Simulated using Xilinx Vivado. On successful Simulation, the code is dumped to the ZYNQ-7000 which features a single-core ARM Cortex™-A9 processor mated with 28 nm Artix-7-based programmable logic. ZYNQ-7000 is proposed because it is the most flexible & scalable platform for maximum reuse and best TTM, also Industry-leading design tools, C/C++, and Open CL design abstractions with Largest portfolio of SW & HW design tools, SoMs, design kits, and reference designs. To ensure the safety and security of sensitive encoded data, Encryption is done using the Triple DES algorithm. Triple DES is the sophisticated version of DES. It is a block cipher technique based on Feistel Structure and it uses symmetric-key block cipher which applies the DES algorithm thrice to each data block. TDES comprises a key bundle of three keys. By doing this sensitive data’s are secured, and provides no chances for stealing or modifying the data. The abstract should summarize the contents of the paper in short terms, i.e., 150–250 words.

In the actual world, images are extremely significant and these are used for describe environment changes. The existence of smog, fog, mist, and haze in the climate corrupt the quality of image acquired by photographic device. Based on the investigation that a hazy picture displays low disparity, we recover the hazy picture by improving its clarity. However, the over damages of the depraved contrast can trim pixel elements values and cause data loss in picture. In this paper using color, attenuation method with histogram equalization, and it is helpful to clear the hazy images. In this paper, we remove haze from a foggy image, improve image quality, and finally, restore and enhance a haze-free image with clear clarity. The proposed technique has been developed and tested in MATLAB2012b, and experimental results show that proposed method completely discards the haze.

In two bus systems, the UPFC is a common device for improving voltage of weak buses. UPFC has advantages in terms of power system static and dynamic operation. The goal of this research is to model and simulate a closed loop controlled three-phase voltage source inverter-based UPFC in a three-phase system. The performance of closed loop PI and FOPID-controlled UPFC systems is explored and compared. The analysis compares time domain response metrics such as steady-state error settling time and THD. The FOPID-controlled UPFC is shown to be quicker than the PI-controlled system.

This research is to investigate momentum innovation for constantly detecting street signs from a moving vehicle. The most encouraging innovation for perceptive vehicle frameworks is vision sensors and image preparation, therefore this is examined the most thoroughly. Various handling calculations and study the world over concerned with sign acknowledgment are investigated. A functional framework has also been implemented using a regular web-camera installed in a testing car. This framework is limited to speed signs and achieves outstanding displays due to rapid but hearty computations. The division is based on shading data, and the recognition is based on a model coordinating computation. The human–computer interface is a voice that announces which sign has been discovered. Driver carelessness is a key element that commonly results in distortion of surroundings, for example, erroneous recognition or disregard of street and traffic signs. Although significant progress has been made in developing a vehicle capable of autonomous guided driving, development has been slow because to concerns of speed, safety, and the ever-changing complexity of the on-road situation. Street plays a key role for the motorist in gathering data from the sign and then acting in a similar fashion.

Data security and GC concepts are the two most promising areas with which the world is most concerned nowadays. As everyday technological progress is witnessed, hackers also develop new techniques for penetrating security. People are migrating toward GC since the power consumption is also a big problem. Using the Zynq 7000 FPGA, we have tried to optimize the total power dissipation (TPD) for the ECC algorithm in our proposed work. The implementation is implanted on VIVADO ISE. In the proposed work, the TPD for ECC design on Zynq 7000 is analyzed for various clock (clk) pulses. From the power calculation, it is observed that the TPD gets decreased as the time clk pulse increases.

Blockchain has many applications in the digital world one of which is non-fungible token (NFT). The NFT originated from Ethereum and it is gaining the attention of researchers over the years. Increasing researchers’ attention toward NFTs has played a most important role in growing the digital market size. NFT is a creative new art type for artists, manufacturers, and musicians. The digital items can be exhibited, sold, stored, and purchased in virtual galleries through NFT. Blockchain provides permanent evidence of trust, rights, and acquaintances to these digital items. In this paper, the NFT market framework is designed and developed. Smart contracts for the NFT market are written in solidity language.

Communication is vital for humans. It helps to share knowledge and information and develop relations with others. We know that it is hard for people to communicate with people speaking sign language. Therefore, the purpose of this work is to implement a machine learning model for classifying and identifying hand gestures like sign language for translating interactions into written and oral forms. We have developed a machine learning model which will detect hand gestures and translate it to words which will help people to understand sign language.

This research work proposes an IoT-based system for stress detection. For stress detection, physiological parameters are acquired and analyzed. Electromyography (EMG) and galvanic skin response are acquired and are used to detect stress. The IoT-based system was designed to transfer data through the Arduino microcontroller. The Arduino Uno processor was employed for the processing of data, biosignal, and physiological parameters are acquired with the aid of dedicated sensors. The statistical analysis of the sensor data was carried out in this research work. The portable system gains its prominence in its usage in industrial and domestic sectors. The system allows for continuous monitoring of stress levels and data transferred through the cloud facilitates disease diagnosis and prediction.

Car diagnostic tools use sophisticated software to quickly and accurately identify problem areas in a car’s engine or elsewhere. The adaptation from mechanical systems to electrical paved the way for much research in the area of vehicle monitoring through a server. Although many studies focus on the tracking of vehicles, fault detection has also gained considerable attention. From our background study, we noticed that most of the systems currently available display the data from OBD and do not perform any functions using that data. We propose a machine learning-based driving pattern analysis and sensor damage prediction system using the data collected from an Onboard Diagnostics (OBD) port. Our system consists of an ELM327 adapter to read data from an OBD port, a mobile application, and a cloud backend. Using the time series data collected, we perform driving pattern analysis and sensor damage detection from the backend server. The smartphone application makes the findings of these analyses visible. The findings can be accessed by the drivers directly via the smartphone app. Drivers are notified of alerts created in the backend as a result of unfavorable conditions.

Many quantum algorithms rely on quantum walks, which are the quantum equivalent of a conventional Markov chain. This paper will present the review of already existing classical Markov model and exiting quantum walk algorithms such as coined quantum walks and Szegedy quantum walk. Then we show the quantum walk search algorithm (QWSA) formulation to solve the problem of detecting elements in graphs. We implemented the algorithm which will detect the element and as well as marked every detected vertex in any given graph and measurement phase in quantum algorithm done in the computational form. Finally, we implemented it on several specific graphs, i.e. four-dimensional hypercube.

Cloud computing is one of the popular topics of the current internet world. Cloud computing technology no doubt makes life simple and easier by offering computing, storage space, and many more services on demand over the internet but associated threats cannot be ignored while accessing these services online. Various efforts have been made by both user's end and cloud services provider to cope up with security issues. Various researches provide advanced defense mechanisms like cryptosystem authentication and intrusion detection system for the last 15 years. This study provides a critical analysis of the security strategies along with there’s role in circumventing threats. The main contribution of this paper is to make society, cloud users, and research scholars aware of the current scenario of widely used technologies. The critical analysis of prevailing approaches reveals the various used prospects which act as the solution essence against emerging attacks on the cloud, at a glance.

In the present work, a fuzzy bi-criteria multi-index bulk transportation problem (FBCMIBTP) with all the parameters to be fuzzy is considered. Bulk transportation problem (BTP) extends to form a multi-index BTP (MIBTP). A bi-criteria BTP focuses on minimizing of cost and time with the solutions being a trade-off between the two. Transportation problem with fuzzy parameters is alike the classic transportation problem, the only difference is the objective function minimizes the total fuzzy instead of a crisp transportation cost. BTP comes into existence when the production of commodity at the origin is of more than one type or when the distribution of commodities occurs through varied transportation modes, but any destination’s requirement can be satisfied only through one origin although one origin can satisfy the requirements of more than one destination. In actual life, the costs and the values for demands and supply are fuzzy numbers. The optimal solutions that are expected to determine the commodities’ value for transportation from an origin to a destination is attained to be fuzzy. A fuzzy cost-time efficient relations on trade-off are considered in the FBCMIBTP, and an algorithm is put forward to obtain the fuzzy cost-time efficient pairs of trade-offs. At first, the algorithm put forward determines the minimum fuzzy cost and corresponding fuzzy time in the considered fuzzy MIBTP, and later the successive fuzzy cost-time efficient pairs of trade-offs are deduced. For clarification of the algorithm put forward, an illustration is formulated by considering triangular fuzzy number.

Finding the right person for the job is a daunting task, especially when we have a large number of job applicants. This will reduce the performance of the team looking for a suitable candidate, the process sometimes takes a long time, and the default “Restart Arrangement and Alignment” program can really help the short course of a short shortlist, which can undoubtedly help to select a competitor and a flexible cycle. The proposed model is capable of dealing with a large number of restatements, first combining classes using a different separator and then the expected set of tasks. Now, candidates can be rescheduled using content-based recommendation, using cosine intimacy and using KNN to split resumes close to a set of expected commitments. In this study and proposed work done on the linear support vector classifier with natural language processing and text classification with the help of algorithm and obtained 79.53% accuracy.

With the great rise in Crypto, the future of blockchain has become susceptible to security threats. These security vulnerabilities allow attackers to compromise sensitive data and manipulate legitimate websites for malware attacks. In this paper, we observe threats and vulnerabilities in blockchain-based websites. By the use of security testing skills and software, we discover severe categories of security vulnerabilities in web applications. To carry out the security testing, we used tools such as, Nmap, Whatweb, and Uniscan, whereas the scans were carried in the OWASP’s Zed attack proxy and Nikto tools. After the scanning process, these tools identify the low, medium, high-level risks, and vulnerabilities and configure the overall security analysis for these web applications. This research puts forth the most common vulnerabilities faced by blockchain-based applications and contributes in reducing the attacks and risk in future.

In simple words, it is mainly used to solve a problem in which the best option has to be selected from the available options at each step. An algorithm is designed to find the best solution for a problem. It is used to find the shortest path from the starting point to the target in a weighted graph. A graph in which the distance or value from one point to another is known, that is, the length of the line joining all the points is known in it. Let us say we are trying to find the shortest route from your house to your friend’s house and you know the distances between different paths across the city. Now if you consider the different locations as a vertices (point) and the path between them as edges (the line joining the points), we can design a weighted graph. Given a set of cities and the distance between every pair of cities, the problem is to find the shortest possible route that visits every city exactly once and returns to the starting point. It does a blind search, so a lot of time is wasted in processing. It cannot handle negative edges. This leads to the acyclic graph and mostly fails to find the correct shortest path. We need to keep an eye on those vertices which are visited once. Note the difference between the Hamiltonian cycle and the TSP. The problem with the Hamiltonian cycle is if there exists a tour that visits every city once. Here, we know that there exists a Hamiltonian tour (since the graph is complete) and that in fact many such tours exist, the problem is to find a minimum weight Hamiltonian cycle.

Among all non-linearities, in particular, FWM (Four Wave Mixing) is a major issue in optical communication. In this paper, an investigation has been done on the impact of FWM on the performance of high-speed optical communication systems. The analysis has been done by comparing different modulation techniques; increasing the number of bits, channels, effective area, sample per bit, length of the fiber; and reducing channel spacing. On the transmitter side, the WDM (Wavelength Division Multiplexing) system has been designed with different techniques, such as RZ (Return to Zero), NRZ (Non-Return to Zero), and On–Off modulation, a range (2–11) of channels, a range (102–17 GHz) of channel spacing, and a range (10–140 bits/s) of bit rate. The channel has been designed with a range (16–112 µm2) of effective area, and a range (70–200 km) of fiber length. The designed WDM system obtains the best value of Q-factor (Quality-factor) and BER (Bit Error Rate) at a particular point in the abovementioned range of parameters. For a range of channels (2–11), the Q-factor varies from (69.5286–36.5789). The optimized values of all the above mentioned parameters are incorporated and the system attains a Q-factor of 178.849, which is found to be best as compared to other values in the given range of the input parameters. The designed system can be extensively tested in the future for a number of other input parameters so as to implement the system in very high-speed applications.

In this manuscript, Octuple RoF upconversion systems based on dual parallel Mach–Zehnder modulator (MZM) configuration (PMC) and Cascaded MZM configuration (CMC) have been generated and compared. Modulation index and bias points of both MZMs and phase of local oscillator (LO) Radio frequency (RF) are properly optimized for generation of 80 GHz upconverted electrical signal from a 10 GHz RF drive signal. The effect of RF drive signal frequency on optical sideband suppression ratio (OSSR) and RF spurious sideband suppression ratio (RFSSR) is evaluated. Mm-wave signal tunability from 8 to 80 GHz is investigated by applying RF drive signal from 1 to 10 GHz. It is worth to mention here that no optical filter is required in both schemes to obtain mm wave signal of high spectral purity, hence leading to reduction in the system cost and complexity.

The paper is an analysis of the impact of metareality on the human mind and its adverse and therapeutic effects on psychological well-being thereof. Metaverse is marked by a blurring of boundaries between the actual physical world and the virtual digital world—a metareality reflected in a three-dimensional, multisensory experience. Based on research evidence, the paper discusses how applications within virtual and augmented reality, such as, the gaming platforms present potential risks and significant safety challenges. There are ethical risks in the metaverse that are related to identity crisis, uninhibited behaviour, and unleashing of sexual and aggressive impulses. At the macro level, the real-time robust mechanisms of this simulated parallel universe impact various societal domains, such as, home, family, religion, and spirituality. The literature review presented highlights the need for digital safety and the prevention of online abuse by launching safe zones and safety tools. However, studies highlight that there are dialectical mechanisms inbuilt in metaverse and these are manifest in its innate therapeutic capacity. Researchers have reported that the imaginal, embodied, and connectivity aspects of this digital reality form the basis for treatment of psychological ailments. The conceptual framework developed will serve as a foundation for researchers interested in studying the link between augmented reality and its effect on the psyche of the human user. The paper has implications for mental health awareness and education which include debunking of associated myths, inculcating of active interest in designing interventions for patient recovery, and for individual growth at the societal level in a digitally transformed world.

In this work, a simple, precise, physical parameters-based model of Amorphous-InGaZnO thin-film transistor is proposed by employing dominant attributes of the device in the specific operating region. The model precisely replicates the measured current–voltage characteristics of any long channel length TFT. In addition, the scope of the model is augmented to replicate the electrical behavior of short channel length TFTs by employing a basic approach that requires a few least-square empirical parameters. For a fixed short channel length TFT, an average error between model outcome and measured device response is found to be less than 3%. The unified model is incorporated in cadence simulator to simulate a 15-stage ring oscillator with long and short channel length TFTs. The simulation response is well accorded with expected device behavior, which further emphasizes the use of the proposed unified model to design oxide TFT-based circuits for a wide application range.

Today, network communication holds a crucial importance of all other types of networks. Of all the advanced network structures, distributed networking system has huge demand. Smart dust network which is one of the categories of distributed networks is the future of distributed networks and has its applications in crucial fields. It is envisioned to blend the features such as ability to sense, compute, and communicate wirelessly. These micro devices can be sprinkled to form a dense network that can monitor real-life processes with high precision results. In this work, we present the characteristics, applications and possibility of using artificial technologies in Smart Dust networks.

Recently, there is an upsurge in tourism and hospitality sector using the VR technology with major players like Airbnb and Hilton using it as an instrument of advertisement. Blockchain based networks have opened newer avenues for electronic business and authentication and identity modeling. The article attempts to elucidate how blockchain and metaverse will transform the travel and tourism—and enlist the facilitators and inhibitors of this paradigm change. The article will also shed light on supporting technologies in metaverse and how it will extend the space of virtual tourism.

The undecimated discrete wavelet transform domain is used for fusing (UDWT). Medical image fusion aims in the direction of coalescing several images out of multifarious sources as an individual representation that may be utilized for better diagnosis. The discrete wavelet transform is used in the majority of state-of-the-art picture fusion algorithms (DWT). When DWT is employed for picture fusion, there is a small blurring. This blurring is greatly decreased when utilizing UDWT. In UDWT, there is no decimation step. As a result, wavelet coefficients for each site are generated, allowing for better recognition of dominating characteristics. It is a multi-resolution decomposition that is not orthogonal. The maximum selection rule employed en routes merging of low-frequency sub-bands in the UDWT domain, whereas modified spatial selection rule is utilized to combine high-frequency sub-bands. In this project, the recommended approach is implemented using the Python tool. The proposed method's superiority is displayed and justified. Various quality measures, include peak signal-to-noise ratio (PSNR), entropy, correlation coefficient, global consistency error (GCE) and mean square error (MSE), are employed to evaluate the performance of fused images.

Image fusion is the process of gathering all information from different multiple images and their inclusion into few new images. These single images are more informative than other single source images. In this paper, we have determined the various systematic literature reviews on the basis of author’s review and conclusion analysis. The images fusions with wavelet transform state that first input images to be fused decomposition by forward wavelet transform images. The wavelet transform decomposes images into low–high frequency sub-band images. The performances of two more images of image fusion on wavelet transform are briefly described for comparison. To evaluate the fusion result, matrix-based mutual information MI is presented for measuring fusion effect.

Pneumonia is a disease that can be caused by bacteria, viruses, and fungi. According to WHO, pneumonia is responsible for 22% of all deaths of children under the age of 1–5 years which is one of the main causes of increased mortality rate. Congestion, gray hepatization, red hepatization, and resolution are the stages of this disease. If the disease is not detected in time, it can progress to a fatal stage. The chest X-ray image is used to diagnose pneumonia, but it requires the presence of experienced radiologists. Pneumonia, COVID-19, cancer, and various other diseases can be identified using X-ray images. If the disease is incorrectly identified, severe difficulties may arise. A deep learning-based model called VGG19 is used to address this issue, which classifies pneumonia from normal lungs. A chest X-ray dataset containing 5856 images was used in this study to classify pneumonia from normal lungs. The outcomes have been demonstrated as accuracy, precision, recall, F1-score, and receiver operating characteristics with the values of 93%, 0.931, 0.93, 0.931, and 0.973, respectively. Furthermore, for validating the proposed model, the performance parameters are compared to the existing work, which results that the proposed model outperforms the other models. In future, this work could be used in hospitals and medical applications.

Wireless Sensor Networks (WSNs) are becoming more popular and are also used in a variety of mission-critical applications. Security in these applications plays a significant role. However, these networks are constrained by a number of factors including limited computation capabilities, energy and storage capacity, unreliable communication, vulnerability to physical capturing and unsupervised activities. The main challenge is to retain security in the network despite these constraints. For secure transmission, key management plays a vital role. But re-keying is necessary when the node is compromised or after a specified number of rounds. So key refreshing increases communication overheads in the network which degrades the network performance. To overcome this problem, a Keyless Data Security (KDS) scheme for WSNs is proposed which eliminates the requirement for key management in the network during data transmission. Simulation results prove that the proposed scheme provides better performance without increasing communication overheads in the network.

Scanners have a wide range of functions. When it comes to an embedded system that converts scanned documents into meaningful data, there is seemingly a void in the industry. Data that can be mined from prescriptions are invaluable and any approach to make meaningful sense of such data is always a beneficial. Such approaches not only push the bounds of how much creative programming can achieve, but also how much it means to the people who can benefit from such innovations. We implement one such innovative approach to convert medicine data within a prescription to make meaningful sense of the data that resides within the prescriptions. We propose a device for converting a medical prescription into a standardized format. To achieve this, we implemented a HP Ink Tank 410 scanner connected to a Raspberry Pi 4 running Ubuntu, we propose a revised algorithm for detecting and parsing related medicinal information from a prescription. This algorithm guarantees a reduction in processing time and improve improves performance.

Water is the necessity of life; without water, human being not survives, but people are polluting the water. Water pollution is the major problem today and affects the groundwater quality. The main causes of water pollution are industries’ waste product disposal, urbanization, crowded population, wastewater, sewage waste, and harmful chemicals’ released by industries. There is an urgent need to resolve this issue in order for us to have safe drinking water. This article proposes a suitable classification model for classifying water quality that is based on machine learning algorithms and can be used to classify water quality. An evaluation and comparison of the performance of various classification models, visualizations, comparisons, and algorithms were carried out in order to identify the significant features that contributed to classifying the water quality of groundwater in Ambala, Haryana. Three models, each with its own set of algorithms, were tested, and their results were compared to each other as well. According to the results, the random forest algorithm was the best classification model out of the five models tested, with the highest accuracy of 82.67% compared to the other models. Overall, wastewater is hazardous to our health, and using scientific models to solve this problem is an absolute must.

The approximate computing is a viable way of lowering the amount of energy. This energy is wasted due to complex designs. This paper proposes an effective approximate multiplier by using an exact multiplier and speculative Han-Carlson parallel-prefix adder. This optimization cuts down on power consumption as well as hardware overhead. Even though this method lowers the precision to some degree, the multiplier is still much more precise than is required for use in practical applications like image processing. An exact error compensation module that is effective and efficient is also designed to enhance the precision of the proposed approximate multiplier. As a result, authors developed a satisfactory solution that strikes a balance between precision and hardware metrics. The design that is implemented offers a significant improvement over its competitors in terms of the trade-offs between the performance characteristics and the precision. The proposed approximate multiplier shows 74.76, 44.41, 34.24, and 53.80% PSNR improvement in comparison with different state-of-the-art work.

New emerging technologies in today’s world play a vital role in each and every sector. In health sector, also a lot of research work is going on to transform the traditional health sector into a digital health sector with the help of these new technologies. And the biggest step toward this transformation in health care is remote patient monitoring (RPM). Remote patient monitoring (RPM) is the most powerful tool to observe the patients effectively at any time. RPM allows the healthcare providers to get the real-time monitoring of the patients with the help of sensors, wearable devices, smart phones, etc. Data is a new fuel of twenty-first century, with the implementation of different new technologies in various fields the legitimate concern of data security and privacy of the users is also arise which has to be resolved for maintain the trust of the users. And this research acknowledges this concern seriously and efficiently. The objective of this research is to enhance the security of the data of patients with the help of RSA by preventing the malicious data infusion in wireless sensor networks (WSNs) used in remote patient monitoring architecture, to maintain the integrity and authenticity of data by en-route filtering using RSA and to enhance the efficiency of overall system by minimum use of energy with the help of AODV protocol. The proposed enhance security mechanism is simulated using NS2 simulator. This research work provides a path to move a step further in transforming the health sector when dealing with the legitimate concern of data protection simultaneously.

Cloud computing is gaining popularity in the domain of computer science and is commonly called as a new data hosting technology; thanks to the cost reimbursement inflicted upon companies, the technology is becoming very popular. Cloud computing plays an integral role in computer science and assists in the development of computer science in a very fast manner. Cloud computing can be defined as a method of sharing resources with clients in a more efficient way. It works with the idea of virtualization and there are several types of service providers, like SaaS, IaaS, and PaaS. Cloud computing has some problems in security and the data stored in the cloud server. The main concern in cloud computing is the security. Due to its distributed and open architecture, an intrusion detection system plays an important role to protect a computer system. Main goals of the IDS include monitoring access and identifying unusual access or attacks on the system. In machine learning, detecting anomalies in data is a fundamental task. These techniques used to identify known and unknown attacks in the cloud.

Correction to: Chapter “Smart Industrial Scanner for Implementation of Relevant Data Parsing from Prescriptions Using SSWF Algorithm” in: © The Author(s), under exclusive license to Springer Nature Singapore Pte Ltd. 2023 N. Marriwala et al. (eds.), Mobile Radio Communications and 5G Networks, Lecture Notes in Networks and Systems 588, https://doi.org/10.1007/978-981-19-7982-8_52