Computer Networks, Big Data and IoT

A wireless network is a cluster of specific transducers with statement transportation intend to study it frequently operate in an unpredictable wireless background with vigour constriction. Several types of research are mainly interest in vigour consciousness and statement dependability of a wireless sensor network to maximize network lifetime. In this article, a greedy algorithm and ACO algorithm aims at obtaining the best clarification that satisfies the given set of Greedy algorithm. The aim of the Greedy algorithm obtains a most favorable explanation that satisfies the given set of constraints and also maximizes the given objective function. Ant colony algorithm has been practical to the traveling salesman problem to find the optimal solution in a short time. However, the performance of the ACO algorithm is considered for both high energy efficiency and good power balancing, and maximal energy utilization throughout the network. ACO algorithm it is understandable that the network time increases and extends the life cycle of the wireless sensor network since it manages the energy and power management. The Greedy Algorithm creates the problem of selecting a communication path using the traveling salesman and cracks the logic by using this algorithm. The algorithm uses the single source to all destination technique to find through path for optimum network connectivity. The simulation results were demonstrated with the help of the algorithm and it outperforms the shortest path length concerning the network lifetime.

ResearchesMoholkar, K. P. Patil, S. H. on question answering systems has been attracting significant research attention in recent years with the explosive data growth and breakthroughs in machine learning paradigm. Answer selection in question answering segment is always considered as a challenging task in natural language processing domain. The major difficulty detected here is that it not only needs the consideration of semantic matching between question answer pairs but also requires a serious modeling of contextual factors. The system aims to use deep learning technique to generate the expected answer. Sequential ensemble approach is deployed in the proposed model, where it categorically boosts the prediction of LSTM and memory network to increase the system accuracy. The proposed model shows a 50% increase in accuracy when compared to individual systems with a few number of epochs. The proposed system reduces the training time and boosts the system-level accuracy.

Instagram and Twitter are popular social media in India. This exploration plans to comprehend the client's behavior and strong/weak relationship inside the social media and to find the range of interaction in the social network. UCINET is an analytic tool for social media networks which examines the client highlights and floating inspiration of these four environments. To measure the centrality three markers such as degree centrality, closeness centrality and betweenness centrality are widely used. Clients obtained the most astounding estimation of centralities are mostly used in social media by people, and the least centrality is used the least. The recurrence to utilize four web-based lives is comparable as per the results of three centralities.

With the boom of malware, the area of malware detection and the use of gadget assist to gain knowledge in research drastically with the aid of researchers. The conventional methods of malware detection are incompetent to detect new and generic malware. In this article, a generic malware detection process is proposed using machine learning named AndroHealthCheck. The malware detection process is divided into four phases, namely android file collection, decompilation, feature mining and machine learning. The overall contributions made in AndroHealthCheck are as follows: (1) designing and implementing a crawler for automating the process of benign files download, (2) collection of unstructured data from the downloaded APK files through the decompilation process, (3) defining a proper mechanism for the feature selection process by performing a static analysis process, (4) designing and implementing a feature mining script for extracting the features from unstructured data collection from APK files, (5) generating a rich homemade data set for machine learning with a huge variety and different flavours of malware files from different families and (6) evaluating the performance of the generated data set by using different types of supervised machine learning classifiers. In this article, the overall architecture and deployment flow of AndroHealthCheck are also discussed.

Machine learning services are the comprehensive description of integrated and semiautomated web devices covering most facilities problems such as preprocessing information, design preparation, and design assessment, with the further forecast. REST APIs can bridge the outcomes of predictions with one’s inner IT infrastructure. Like the original SaaS, IaaS, and PaaS cloud delivery models, ML and AI fields cover high-level services to provide infrastructure and platform, exposed as APIs. This article identifying the most used Cloud Technologies for Machine Learning as a Service (MLaaS): Google Cloud AI, Amazon, and Microsoft Azure.

Mobile ad hoc network (MANET) is a network that permits mobile servers and customers to convey without fixed infrastructure. MANET is a promptly developing region of investigation as it employs a group of uses. To encourage effective information access and update data sets are conveyed on MANET. Since information openness is influenced by the portability and power imperatives of the servers and customers, the information in MANET is reproduced. As in ad hoc network, since portable host moves unreservedly network panel happens habitually in this way information accessibility is decreased bringing about execution debasement. The majority of the replication strategies expect that each node advances each packet provided to it and combine completely as far as sharing the memory space. A portion of the nodes may go about as selfish nodes which may selfishly conclude and participate incompletely with all the different nodes. In this manner, selfish conduct of nodes could lessen the general information openness in the network. At first, a design model of a MANET is built and the correspondence between the versatile is started. The packet drop can occur in MANET because of the selfish node or network clog. In this paper, MSD-SNDT and MBD-SNDT method is proposed to recognize the selfish nodes effectively in MANET. The reproduction study shows that the proposed MSD-SNDT and MBD-SNDT strategy improves the selfish node detection ratio, packet delivery proportion (PDP), and normal packet drop proportion.

IoT is defined as a pervasive and global network that aids and provides the system for monitoring and controlling the physical world through the processing and analysis of generated data by IoT sensor devices. Wireless sensor networks (WSNs) are comprised of a large number of nodes distributed in a vast region. Routing protocols are responsible for the development and the management of network routes. This paper intends to propose an optimized routing model for selecting the optimal shortest path in IoT-based WSN. More particularly, a dragonfly algorithm with Brownian motion (DABR) model is introduced to select the optimal route by taking into consideration of certain constraints such as (i) delay (ii) distance (iii) packet drop rate (PDR) and (iv) energy. Finally, the performance of the proposed work is compared with the conventional models to demonstrate the superior performance.

Edge computingZivkovic, Miodrag Bezdan, Timea Strumberger, Ivana Bacanin, Nebojsa Venkatachalam, K. is a relatively novel technology, which is closely related to the concepts of the Internet of things and cloud computing. The main purpose of edge computing is to bring the resources as close as possible to the clients, to the very edge of the cloud. By doing so, it is possible to achieve smaller response times and lower network bandwidth utilization. Workflow scheduling in such an edge–cloud environment is considered to be an NP-hard problem, which has to be solved by a stochastic approach, especially in the scenario of multiple optimization goals. In the research presented in this paper, a modified Harris hawks optimization algorithm is proposed and adjusted to target cloud–edge workflow scheduling problem. Simulations are carried out with two main objectives—cost and makespan. The proposed experiments have used real workflow models and evaluated the proposed algorithm by comparing it to the other approaches available in the recent literature which were tested in the same simulation environment and experimental conditions. Based on the results from conducted experiments, the proposed improved Harris hawks optimization algorithm outperformed other state-of-the-art approaches by reducing cost and makespan performance metrics.

The security of a communication model is defined by the strength of the encryption and key generation algorithm. To ensure security, in block cipher techniques, a complex computation process is used for encryption. But, in stream cipher techniques, complex key generation techniques are used. This paper proposes a novel and complex key generation algorithm for stream cipher techniques that generate 15,72,864-bit key from a synthetic color image using a pattern-based bit extraction technique. A segmented form of a generated key can be used as a dynamic key in block cipher techniques. The proposed algorithm randomly uses eight patterns to extract bits from a synthetic color image to generate the key. The generated key’s randomness is tested using the NIST statistical analysis tool and compared with the keys generated from existing techniques. The key length and keyspace are compared with existing methods.

Blockchain technologies offer an innovative approach to data storage, transaction execution, process automation, and confidence-building in an open and distributed environment. A wide range of blockchain-based applications have been developed and deployed since its inception, and the convergence of blockchain with other state-of-the-art technologies has been on the rise. However, the issues and challenges associated with it are also rising with the advancement of blockchain technologies and their applications. This paper provides an overview of blockchain and reviews the typical domains of blockchain application discussing the open issues and challenges. Besides, this article reviews the recent advances in addressing different blockchain issues and points out the research directions that help to develop technological innovations for future blockchain systems.

MANET is basically used for quick transmission of data without creating infrastructure-based network. In order to do transmission in mobile ad hoc network, it is required to find out efficient path between source and destination. MANET is developed using different type of topologies which frequently get changed on the movement of node that affects network performance result in selection of inappropriate path. So to find out best path between nodes, routing protocol is required. There are number of routing protocols available based on the requirement of network it get selected and implemented. This paper provides the information of various routing protocol and optimization algorithm developed by different researcher which is applied to improve performance of MANET.

In today’s world, with the innovations of various technologies and the subsequent growth and competition in various sectors, technologies such as IoT are playing a key role in amalgamating technology and industries. This vision to use IoT in the industry is to complement the needs of industries, increase automation, receive feedback, timely response, and most importantly increase production by many folds. This drastic change and implementation of technology helped to generate more revenue but the threat of cyberattacks is still prevalent and is largely overlooked. Industrial IoT which does not follow any standardized security protocols is heavily vulnerable to cyberthreats and critical sectors like power plants, smart meters, etc., which uses IoT to connect are at high risk. In this paper, a basic overview of the cybersecurity issues in industrial IoT and various cyberattacks related to different tiers of IoT architecture is reviewed. Also, research work has discussed various mechanisms that can be used to thwart and mitigate cyberattacks on the industrial IoT systems.

Frequent pattern mining is one of the key research areas in the Data Mining (DM) paradigm. There are many algorithms in the literature to identify the frequent itemsets whereas research on rare pattern mining is in the burgeoning stage. Rare items are the infrequent items, where few applications like medical diagnosis, telecommunications, and false alarm detection in industries demand for rare patterns and rare associations with frequent or infrequent items sets in the database. The algorithms that are used to identify frequent items can also be used to identify rare patterns. However, such algorithms suffer from RareItemProblem. Rare Pattern Mining algorithms that are based on Apriori and FP-Growth were designed but Eclat-based rare pattern mining algorithms have not been explored. This paper proposes an Eclat-RPGrowth, algorithm to find rare patterns and the support of itemset is calculated by using intersection of BitSets for corresponding k − 1 itemsets. Also, this research work proposes a variant of Eclat_RPGrowth as Eclat_PRPgrowth. Both the algorithms are outperformed in execution time, and with the number of rare items generated.

The study analyzes the scholarly information transfering through Social Media and Networks by the respondents in the selected State universities of Tamil Nadu. The study examined that sharing scholarly communication on Web sites that facilitate relationship caters of needs of the researchers in any disciplines varied. This study discusses the total number of 501 respondents have reported from selected state Universities in Tamil Nadu. It determines the results of that male 260 (5.19%) and female 241 (48.1%) of the respondents from the selected eight State Universities. There are eight Universities have been chosen for collecting data from the respondents. The study also attempts to find out that Facebook users are predominant factors in terms of sharing and interacting with peer groups. The respondents highly prefer group sites (Yahoo, Google, and Whatsapp). The research analyses that social media tools for research the majority of the respondents highly preferred Facebook wall for shared the research information by the respondents in the eight Universities in Tamil Nadu.

Social media is an essential part of life for most people around. No wonder even during emergencies like flood or cyclone, more and more people look up to Twitter, Facebook, WhatsApp groups, etc., for immediate assistance. This helps to get data from even remote places and from small groups which will be difficult to reach. This sheer amount of data generated during a short span of time is also the challenge in this approach. Even when there are resources available for help, many requests could go unnoticed. This paper addresses above-mentioned problem by collecting the generated requests for help and resource availability and plot the location in the map. Request data shall be analysed using three machine learning algorithms called linear ridge regression, SGD classifier and Naive Bayes algorithm for the initial filtering and will be passed through natural language processing to match needs and offers within a given geographic boundary. The system is working with 96% accuracy for linear ridge regression and Naive Bayes classifier and 95% accuracy for SGD classifier. The report shall be published to provide a centralized status of requests. This brings more efficient management of disaster situations.

The arrival of social media has initiated the platform for the public to express their views and to share their emotions. In addition to this, smart phones and mobile communication technologies have emerged and become as a tool to spend more time than earlier on social media to be in touch with their friends. The way of expressing the thoughts, attitude, and feelings is changed drastically due to that short messages and emoticons utilization has been increased through social media. There have been no constraints in using the micro blogging services like Facebook, Twitter, etc. Thus, these messages are the views of people used to describe their behavior. Nowadays political party workers and leaders also started spending more time on Twitter, Facebook, and blogs to be in touch with the public. Hence, political parties or social media campaigners started analyzing for the solutions to utilize the public opinions or views about their own political party. A program is developed which collects the reviews or text posts of people and transfer it to the next module called pre-processing or cleansing. This module is to remove URLs, special symbols, Junk text, stop words, and tokenize the sentences and to perform the stemming process. Word2vec model is used to perform the word embedding to convert the text data into numerical vector format to transfer it to Recurrent Neural Network. The sigmoid function is used as an activation function to classify and polarity of positive and negative sentiments have been evaluated. The effectiveness of the proposed system has been proved through experimental analysis in this paper. This proposed system will be used to understand the winning chances of political party and to analyze the response of the public on particular political decision during election campaign.

Internet of things (IoT) is a primary computational paradigm to develop a smart environment in every area of health, city, factory, and home in our daily lives. It incorporates wireless transmissions to all sensor devices through the internet. Equipping a smart environment to the society, IoT as the primary source provides alternative diversified communicating characteristics. Its ecosystem is the solution to all communication technologies as well as designed architectures. This paper deals with distinct core requirements to generate reusable features and technologies to develop a smart environment. Technological architectures like Radio Frequency Identification (RFID) and Constrained Node Network (CNN) are identified to enhance the Internet of things. This paper also describes the necessity of having smart environment sensors with the Internet of Things (IoT). This shows the involvement of a smart environment crossing all communicative disputes from the technical and informative perspective that desires to fulfill the efforts of the people in the coming years.

Internet of Things (IoT) is rapidly advancing as a most recent exploration point in various academic and industrial controls, especially in medical care, as the flow transformation of the Internet. Due to the rapid proliferation of wearable devices and smartphones, innovation facilitated by the Internet of Things is evolving medical treatment from the conventional center-based setting to a more Personalized Healthcare System (PHS). Nevertheless, making the use of cutting-edge IoT innovation in PHS remains essentially testing within the area thinking about various problems, such as lack of functional and accurate clinical sensors, unstandardized IoT system models, heterogeneity of related wearable gadgets, multi-dimensionality of generated knowledge, and interoperability popularity. This paper offers a description of the IoT of various diseases in healthcare systems.

Social media plays an indispensable role in the ever-expanding human population, which will in turn result in generating a huge volume of data.Srivastava, Sonam Singh, Yogendra Narain Social media analytics is the technique used for acquiring and analysing the data from social networks. The big data associated with social media finds state-of-the-art applications in the distinctive socio-economic domains. A considerable research literature evidence is available on the challenges that are inherent in particular applications of social media or big data separately, but there is hardly any exclusive study that has been performed on the social media big data. To address this gap, this paper presents the state-of-the-art social media big data applications in business, healthcare, education and crisis management with which the challenges associated with them are also critically evaluated. Also, different frameworks of big social media analytics are presented, and their potential in addressing the application-specific challenges is also described. Henceforth, this research article leverages significant advantages to researchers and experts, who wants to gather and analyse the social media data.

Day-by-day India’s population is consistently increasing, and it is expected to reach beyond 1.3 billion.Darshini, P. Mohana Kumar, S. Prasad, Krishna Jagadeesha, S. N. In fact after 25 years, serious problems will be predicted regarding food. Hence, it is necessary to develop an innovative system for benefiting the agricultural sector as early as possible. The farmers are suffering due to the reduced rainfall and water scarcity. The traditional farmland irrigation techniques require manual intervention. To overcome this challenge, the proposed research work has developed an automatic irrigation system, which is powered by a smart solar system for saving time, money and work of the farmer. Solar panels will continuously track the position of the sun to ensure the maximum energy production. In addition, an intruder detection system [IDS] is installed into the irrigation system with the help of passive infrared sensor to track where the birds and some other animals are repelled from entering into the field. A Wi-Fi module has been used to establish a communication link between the farmer and the field. The farmer can access the information about the field condition anytime.

Intelligent computing is a novel means of delivering computing services and resources. Due to massive demand of healthcare services that are transforming the face of health information systems (HISs), investing in healthcare research becomes more fundamental. Nonetheless, just like any form of healthcare advancement, intelligent computing needs to be investigated and evaluated before health practitioners decide to apply it globally. The application of innovations as a result of intelligent computing can be one of the best means of transforming healthcare provisions especially during the process of sharing patient data between medical practitioners during the urgent actual-time cases. However, before initiating the complete transformation of the healthcare sector, there should be a specific strategy that requires thorough evaluation. For instance, a feasible intelligent strategy that is applicable in the healthcare facility must use a public domain cloud infrastructure to permit public accessibility to fundamental engineering healthcare data during the process of retrieving medical resources.

This work aims to detect the signs of the Coronavirus, also known as Covid-19. A dry cough, sore throat, and fever are the most common symptoms of Covid-19. For Covid-19, it is important to find fast and precise results at the time to stop it in the early stages and to avoid it from the vast spread. To interpret and identify the symptoms from X-rays and CT scan images, the machine learning and computer vision principles were applied. The works are usually performed with the CSV datasets. However, the analysis is performed to compare the images of patients with Covid and Non-Covid. To enhance the classification performance, it is feasible to use feature extraction techniques such as CNN, local directional pattern (LDP), gray-level run length matrix (GLRLM), gray-level scale zone matrix (GLSZM), and discrete wavelet transform (DWT) algorithms (Barstugan in Coronavirus (Covid-19) classification using CT images by machine learning methods [1]). In this paper, the convolution neural network model is selected as the classifier. Softmax is used during the classification process to classify the images given, whether they belong to Covid or Non-Covid. This implementation is carried out on both the X-ray images dataset and the dataset of CT scan images which are obtained from the repository that is publicly accessible.

Multi-access/mobile edge computing (MEC) is a structural design for enabling cloud computing platform at the edge of mobile network, so as to reduce network congestion, improves fast response, and optimization of mobile resources to compute complex applications. MEC provides a distributed computing environment for its applications. In MEC, all the computational reserves will be brought to the base location of the mobile networks. It has the ability to store and process the contents at physical proximity to the mobile subscribers to assist delay sensitivity and contextaware applications. Generally, the packet forwarding and filtering will be operated by the traditional edge network. At present, applications’ online computations and storage are done at remote servers, and those servers are placed far away from the users. New technologies are emerged to shift all the available resources to a edge from a cloud. Radio access network (RAN) could able to do the computational off-loading to the edge of the mobile network. So, it is mandatory to have a load balancer at the edge to distribute all the job to all the available processors without any congestion or delay. A routing specifies how to route the traffic between each origin-destination pair across a network. The traffic sharing is applied in a routing and allocating process to enhance the survivability of a network. The experimental results show the proposed algorithm works better than the existing dynamic load balancing algorithms.

Internet service providers and techno enterprises evolved with CISCO IOS called MPLS built a strong and diligent network for succeeding generations to utilize all sorts of full-fledged amnesties over a solitary structure. MPLS is considered as a routing method, and it is not a facility or a service. MPLS can be encased with any prevailing infrastructures, namely digital subscriber line, asynchronous transfer mode, frame relay, and IP. MPLS is not platform dependent. It can work seamlessly without making any change in the current environment of these technologies. But implementing MPLS is quite expensive, so with the support of SD-WAN, ISPs and enterprises are attempted to enhance its usages using inexpensive Internet connection. This survey articled aimed to compare the pros and cons of both the technologies MPLS and SD-WAN.

Telehealth uses wireless communications and Internet services to offer healthcare services remotely. In fact, during pandemics such as COVID-19, telemedicine and e-health services play a crucial role. Security and privacy are of prime importance in e-health systems/services. This paper describes common security issues in telehealth services, such as attacks on end users, medical equipment, and access networks. Further, it also discusses security solutions are employed in telemedicine, including encryption techniques, watermarking, message authentication code (MAC), digital signatures, and the zero trust model. This paper concludes that the use of e-Health services will continue to increase, and security solutions/architectures for these services are of research interest.

In this fast-paced world, people are busy chasing their lives. The stress, burden, and carelessness have caused different accidents on roads. Over speeding is one of the reasons for accidents which have caused a huge loss of lives due to the lack of immediate treatment. This research work provides an optimum solution to this problem. According to the proposed work, turbulence on the vehicle will be detected by the accelerometer and can be deactivated within the initial 20 s period using false switch alarm in case it is a false emergency. The accident location and other details of the driver will be sent to the rescue team. The paper proposes a system where immediate alert can be given to authorities and thus save a person injured in the accident within minimum time.

Cloud Computing provides on-demand, flexible, ubiquitous resources for clients in a virtualized environment using huge number of virtual machines (VMs). Cloud data centers don’t utilize their resources fully which leads into a underutilization of resources. Virtualization offers a few exceptional highlights for cloud suppliers like saving of power consumption, load adjusting, and adaptation to internal failure, resource multiplexing. However, for improving energy proficiency and resource utilization, various strategies have been introduced such as server consolidation and different resource structuring. Among all, Virtual Machine Placement (VMP) is the most vital strides in server consolidation. Virtual Machine Placement (VMP) is an efficient mapping of VMs to Physical Machines (PMs). VMP issues go about as a non-deterministic polynomial-time hard (NP-difficult) issue and metaheuristics strategies are widely used to solve these issues with enhancing boundaries of power utilization, QoS, resource usage, etc. This paper presents an extensive review of Metaheuristics models to deal with VMP in the cloud environment.

Image Segmentation plays an indispensable role in Computer Aided Diagnosis system to extract the region of interest for diagnosis. Transrectal Ultrasonography (TRUS) is the best imaging technique used mostly by physicians to separate the prostate region from tissues around it and the same used for abnormality detection. The extraction of a region of interest from TRUS images is complex as it contains speckle noise, tracker, low dissimilarity, the fuzzy region between object and background, and also irregular form and size. To resolve these problems, a novel approach Ant Colony Optimization with Boundary Complete Recurrent Neural Network (BCRNN) is aimed to take out an accurate prostate region from the TRUS images. This method comprises two stages such as pre-processing and segmentation. In the beginning stage, Ant Colony Optimization (ACO) method is adopted to eradicate speckle noise from the TRUS prostate image. And, in the second stage Boundary Complete Recurrent Neural Network (BCRNN) method along with shape prioritize and multi-view fusion is employed to draw out the rigorous shape without boundary loss of the prostate. BCRNN comprises of three modules. Firstly, images are serialized into dynamic sequences and Recurrent Neural Network (RNNs) is applied to obtain shape prior to an image. Secondly, multi-view fusion approach is embedded to merge the shape predictions attained from various perspectives. Finally, to refine the details of the shape prediction map, the RNN core method is inserted into a multiscale auto-context. The proposed method is assessed using statistical performance metrics such as Dice Similarity Coefficient (DICE), Jaccard index (JI), Precision, Recall, and Accuracy. From the result analysis, it is found that ACO with BCRNN successfully extracts Region of Interest (ROI) from the image with an accurate boundary that is used for diagnosis. The experiment results clearly revealed that the performance of ACO with BCRNN is superior to BCRNN.

Diseases in cows are an influential point for human concern. There are some diseases in animals identified in the early phases that can be diagnosed and cured in the early phases of the disease itself. The effect of lumpy skin disease can cause large capital losses in the farm animal industry if it is not taken care of properly. The main reason for this disease is the lumpy skin virus, and this virus is a part of the Poxviridae family. The major symptom of lumpy skin disease is the Neethling strain, and other symptoms are a few mild forms of circumscribed skin nodules. These symptoms also include mucous membranes of internal organs like respiratory organs and reproductive organs. By the infection of such disease, animals like cattle get their skin permanently damaged. Some of the detrimental outcomes of this disease in cows are reduction in milk projection, infertility, poor growth, abortion and sometimes death. In this research work, an architecture using machine learning techniques to detect the disease is proposed. This architecture employs the pre-trained models like VGG-16, VGG-19 and Inception-v3 for feature extraction and then followed by multiple classifiers. The work is tested on our manually collected dataset, and the extracted features were further classified using the classifiers like kNN, SVM, NB, ANN and LR. Using this methodology, the state-of-the-art solution obtaining a classification accuracy of 92.5% over the test dataset.

The social networking sites are currently assisting in delivering faster communication and they are also very useful to know about the different people’s opinions, views, and their sentiments. Twitter is one of the social networking sites, which can help to predict many health-related problems. In this work, sentiment analysis has been performed on tweets to predict the possible number of cases with H1N1 disease. The data will be collected country wise, where the tweets lie between four ranges on which the further analysis will be done. The results show the position of India based on the frequency of occurrence in the tweets as compared to the other countries. This type of disease prediction can help to take a quick decision in order to overcome the damage. The results predicted by sentiment analysis of Twitter data will then compared with the data obtained from the ‘Ministry of Health and Family Welfare-Government of India’ site. The data present at this site gives the actual number of cases occurred and collected by Indian Governments “Integrated Disease Survellience Program”. Comparison with this data will help in calculating the accuracy of the sentiment analysis approach proposed in this work.

An image is a distributed amplitude of colors on a plane. An image may be in the form of two-dimensional image or three-dimensional image. Such images are compiled using optical sensors like camera and are processed using various image processing tools for better visualization. Purpose of the image processing is not limited for better visualization, but it is extended to remove noise from the captured image. Noise is a random variation of brightness, contrast and color pallets in an image. In the present discussion through review of denoising framework for efficient removal of noise from 3D images, different filters which are used so far for removal of noise are discussed. The research work is further extended by designing novel denoising framework for efficient removal of noise from the 3D image.

Machine learning models are becoming progressively predominant in the algorithmic trading paradigm. It is known that a helpful data is taking cover behind the noisy and enormous information that can give us better understanding on the capital markets. There are multiple issues, which are prevalent as of now by including the overfitting model, irrelevant/noisy data used for training models due to which the efficiency of the existing models fail. In addition to these problems, the existing authors are facing issues with the dispersion of daily data, poor presentation, and the problems faced with too much or too little data information. The main objective in this undertaking is to discover a technique that choose gainful stocks ordinarily by mining the public information. To accomplish this, various models are assembled to foresee the everyday return of a stock from a lot of features. These features are built to be more dependent on the cited and outside information that is accessible before the forecast date. Numerous sorts of calculations are utilized for anticipating/forecasting. When considering machine learning, regression model is implemented. Neural networks, as a wise information mining strategy and profound learning methods are valuable in learning complex types of information by utilizing the models of regulated learning, where it progressively learns through datasets and experience. Because of high volumes of information produced in capital markets, machines would master different designs, which in turn makes sensibly great predictions. In the current proposed venture, LSTM model is utilized through RMS prop optimization for anticipating future stock estimations. Also, feed forward multi-layer perceptron (MLP) is utilized along with recurrent network to foresee an organization’s stock worth by depending on its stock share price history. The cycle in the financial exchange is clearly with a ton of vulnerability, so it is profoundly influenced by a great deal of numerous elements. Nonetheless, the outcomes acquired show that the neural networks will outperform the existing one-dimensional models.

Intrusion detection system [IDS] is a significant base for the network defence. A huge amount of data is generated with the latest technologies like cloud computing and social media networks. As the data generation keeps increasing, there are chances that different forms of intrusion attacks are also possible. This paper mainly focuses on the machine learning (ML) techniques for cyber security in support of intrusion detection. It uses three different algorithms, namely Naïve Bayes classifier, Hoeffding tree classifier and ensemble classifier. The study is performed on emerging methods and is compared with streaming and non-streaming environment. The discussion on using the emerging methods and challenges is presented in this paper with the well-known NSL_KDD datasets. The concept of drift is induced in the static stream by using the SEA generator. Finally, it is found that the ensemble classifier is more suitable for both the environments with and without concept drift.

Due to the advancements of Internet, majority of people will prefer to buy and sell the commodities through e-commerce websites. In general, people mostly trust on reviews before taking the decisions. Fraudulent reviewers will consider this as an opportunity to write fake reviews for misleading both the customers and producers. There is a necessity to identify fake reviews before making it available for decision making. This research focuses on fake review detection by using content-related features, which includes linguistic features, POS features, and sentiment analysis features. Ontology-based method is used for performing the aspect-wise sentiment analysis. All the features of reviews are calculated and incorporated into the ontology, and fake review detection is also accelerated through the rule-based classifier by inferencing and querying the ontology. Due to the issues related with labeled dataset, the outliers from an unlabeled dataset were selected as fake reviews. The performance measure of the rule-based classifier outperforms by incorporating all the content-related features.

Wireless ad hoc network [WANET] systems are creating multihop communication structure between the mobiles to transfer the data groups. The remarkable qualities of remote frameworks cause the correspondence to interface among the conflicting mobiles. To manage high convey ability and biological blocks, most physical directing shows will not believe the stable associations during pack communication, which prompts the elevated delay and bundle reducing in the mastermind. The proposed research work recommends a way perception support physical steering convention that specifies POPR for WANET. The anticipated guiding show merges the relative partition, course and mid-expand forwarder center point with transfer thickness to propel data toward the objective in order to recover physical sending among the connection point. Multiplication results illustrate the projected directing convention, which performs superior to the existing arrangements.

Problem importance: Predictive analytics seems to be an exceptionally complex and vital concern in domains like computer science, biology, agriculture, business, and national security. When big data applications were indeed accessible, highly efficient cooperation processes are often meaningful. Simultaneously time, new subjective norms originate when the high quantities of data will conveniently assert confidential data. This paper has reviewed two complementary huge issues: data integration and privacy, the ER “pay-as-you-go” approach (Whang et al. in IEEE Trans Knowl Data Eng 25(5):1111–1124 (2012) [1]) in which it explores how the developments of ER is maximized to short-term work. Stepwise ER problem (Whang and Molina in PVLDB 3(1):1326–1337 (2010) [2]) is not even a unique process; it is done concurrently by the better usage of information, schemes, and applications. Joint ER problem with multiple independent datasets are fixed in collaboration (Whang and Molina in ICDE (2012) [3]) and the problem of ER with inconsistencies (Whang et al. in VLDB J 18(6):1261–1277 (2009) [4]). To overcome the research gap in the existing system, the proposed research work addresses an entity resolution (ER) problem that tends to address the records in databases referring to a certain complex real-time entity.

Cloud computing servers gives a stage for the clients to remotely store information and offer the information to everyone. Healthcare record management system (HRMS) has been developed as a tolerant-driven model for health-related data exchange. Classification of the common information, i.e., visit details, report details, etc., remains as a serious issue when patients utilize public cloud servers since it might be viewed by everyone. To guarantee the patient’s control over admittance to their own uncommon well-being records, it is a promising technique to encode the reports before redistributing and pick up the gain power to that data. Security introduction, adaptability in key association, and flexible access experience have remained the most essential difficulties toward accomplishing fine-grained, cryptographically endorsed data get the opportunity to control. This research work proposes an arrangement of systems for information get to control to healthcare record management system (HRMS) put away in outsider servers with classification of data into public and private data. In HRMS for accomplishing smooth and adaptable information get the chance to control, attribute-based encryption (ABE) methodologies are utilized to encode each patient’s HRMS record for getting to. While providing secure data outsourcing, the main concentration is the multiple owners of data. On account of this framework significantly diminishes the key administration unpredictability for information owners and clients, HRMS ensured a high level of patient security. With respect to emergency situations, a break-glass extraction method is introduced here to access the data in emergency situations with authorized keys.

With the development of Web technology, the Internet has become a stage for wide scope of criminal scheme including spam promotion, budgetary fraud, Web page defacement, etc. Because of the fast development of the Internet, Web sites have become the interloper’s fundamental objective. As the quantity of Web pages expands, the malicious Web pages are moreover extending and the attack is dynamically gotten progressed. The existing approaches like blacking listing and dynamic analysis approaches are time and resources intensive, hence these methodologies are not adequate to classify the Web sites as malignant or benign. The proposed research work develops a lightweight classification framework to recognize malicious Web pages effectively using distinctive supervised machine learning classifiers like Naïve Bayes, K-nearest neighbors, random forest, AdaBoost, and some distinct relevant URL features. So here, some URL-based features are extracted for the detection of malicious and benign Web sites. We have experimented the test results using Python environment and found the random forest achieves highest classification accuracy of 98.7%.

Parkinson’s disease is one of the common chronic and progressive neurodegenerative diseases across the globe. Speech parameters are the most important indicators that can be used to detect the disease at its early stage. In this article, an efficient approach using Convolutional Neural Network (CNN) is used to predict Parkinson’s disease by using speech parameters that are extracted from the voice recordings. CNN is the most emerging technology that is used for many computer vision tasks. The performance of the approach is discussed and evaluated with the dataset available in the UCI machine learning repository. The dataset will contain the attributes of voice recordings from 80 individuals out of which 40 individuals are Parkinson affected. Three recordings of each individual are used and 44 features are extracted from each recording of a subject. The experimental setup of the proposed approach on the benchmark data has achieved the best testing accuracy of 87.76% when compared with the available ground truth of the dataset.

This paper examines the relevance of a fast approaching highly secure and fast data transmission technique using Li-Fi. It describes the upcoming technology Li-Fi and its applications as well as the developments made in it so far. It enlightens on the new era that will soon be used in almost all domains like health sector, school, bank and so on. An application framework design has been studied to analyze the role of Li-Fi in the process of communication.

People with visual impairment face a lot of challenges in their daily lives, be it small or big. This is mainly due to the lack of assistance provided by the modern assistive devices in terms of providing self-independence and the cost matching it. The main aim of this article is to research and explore the existing assistive technologies in the domain of visual impairment aid. The main objective of the assistive technology is to provide assistance in the day-to-day tasks, with a simple and wearable design to deliver a better user experience. This paper focuses on different approaches that will help the visually impaired through technology and learn those technologies that leverage a comfortable experience to the user. The primary objective of this survey is to navigate through the different approaches to find out the best suite for the authors in developing their technology.

To preserve the confidentiality of important data stored in third-party managed platforms like public cloud databases is continuously raising security concerns.Mallaiah, Kurra Gandhi, Rishi Kumar Ramachandram, S. Such databases need to be shielded from malicious administrators or malicious software attacks, so as to increase the trust of the potential customers of such security. Therefore, it is very important that the data must be protected at rest, in transition and also while in operation to achieve full confidentiality for the data stored in databases. Shielding the confidentiality of data at rest and in transition is adequately addressed but shielding the confidentiality of data while in operation is still a big challenge. CryptDB provides confidentiality for relational databases by supporting the computations on encrypted data. The keys which are used in CryptDB are stored in the proxy. This paper proposes a stateless key management scheme, inversely the state-full key management scheme of CryptDB. The security proof is presented along with cryptographic security definitions and that the security of our key management under the random oracle model security assumption. The proposed key management scheme is the first of its kind that is a stateless key management. The scheme eliminates the storing of user cryptographic keys in the proxy and, hence, avoids possible attacks and analysis on keys in the proxy and also various other problems such as key backup, key loss, and key audit. Our proposed solution also satisfies regulatory compliance by not storing the cryptographic keys anywhere.

Internet is the technology that is highly used for communication as well as data transmission in the current scenario. Because of the rapid growth in such technology, the issues related to security are also increasing. Blockchain is emerging as a revolutionizing technology in this field. Basically, blockchain relies on the four components, namely decentralization, anonymity, persistence, and audibility. It focuses more on protecting data from unauthorized parties in performing any modification to the existing data. Blockchain technology is also defined as a digital ledger. It is not only limited to cryptocurrencies but with the potential to be transparent and fair, which opens many doors to various technologies like IoT, big data, and many more. This paper includes blockchain locution, and it will encounter some typical consensus algorithms that are ought to analyze blockchain applications and some technical challenges. The proposed research work will also concentrate more on the recent advances to tackle those challenges.

FBMC has arisen the most multicarrier (MC) procedure since it satisfies the need of the next-generation 5G standards. Optimized generalized Gaussian pulse (OGGP) filter of the FBMC systems have been proposed in this paper. A PTF strategy based on optimized generalized Gaussian pulse is proposed along with the lowest time domain side lobe levels and tail energy. The proposed system has been used to OQAM modulation and different Tx and Rx antenna. From the numerical perspective, the transmission execution can be improved by considering the greatest signal-to-noise ratio and reduced ISI. The proposed system is simulated by using MATLAB software and also by using the BER and magnitude response.

Globalization and increasing distribution of IC supply chain have resulted in various third parties having a key to precious intellectual property or the physical integrated circuit and therefore information can be exploited. Information security is the practice of safeguarding information by minimizing information risks. It is required to scale down the danger of unauthorized information disclosure, modification, and destruction. Hardware security threats have been observed at several levels of the IC supply chain. To protect the hardware from potential attacks, there are various design-for-security (DFS) techniques. Interleaved memory with logic locking techniques for information security is proposed in this paper. Interleaved memory is a solution for random arrangement and logic locking is needed to interrupt the chain and to protect the data by restricting its access to authorized users. It is a versatile and easy-to-integrate solution that needs only trusted designers. The approach proposed in this paper compares the Hamming distance (HD) and Levenshtein distance (LD) and BER obtained for random logic locking (RLL) and weighted logic locking (WLL) on the interleaved memory. From the results obtained, it can be concluded that weighted logic locking on interleaved memory provides better security.

The research paper discloses a theoretical model of a “content personalization device” that functions over an operating system (OS) of any mobile handheld device. The device is in the form of a thin-layered transparent film that fits on the screen of the mobile device. The device operates as a layer over the OS as all the content mediated or allowed by the OS on the device passes through this personalization layer, where it gets filtered. This layer analyzes user interactions with the device and provides filtered content based on the user’s response to the content. The externally integrated device tracks the user’s device usage pattern and adapts to changing user choices and preferences over time for filtering and displaying relevant content to the user of the mobile device.

One of the major challenges faced by visually challenged people is self-navigation in unknown environments. They often tend to get hurt by objects that they cannot feel using their hands or a walking cane, as certain objects are hard for a blind person to detect by just using tapping their walking cane. To avoid obstacles and navigate through a new environment, a smart belt for the visually impaired is performed in which he/she can continuously detect the obstacles around the user with its sensors that span the entire 360° of his/her field of view. Whenever an obstacle is in a nearby range, sufficiently enough to cause a hindrance, the device will give sensory cues to the user about the location of the obstacle and family members are also tracked them using GSM and GPRM system.

A smart grid is an updated version of the traditional electrical grid that uses Information and Communication Technology (ICT) in an automated fashion for the production and distribution of electricity. Several attributes like efficiency, sustainability, and reliability are improved in the smart grid as compared to the traditional grid. Smart grid gives significant benefits for the entire community, but their dependence on computer networks make them vulnerable to various kinds of malicious attacks. This article focuses on identifying the various cyber-security issues of different areas of the smart grid which are prone to vulnerabilities. Finally, the possible solutions for resolving the cyber-security issues in the identified areas for making the smart grid more secure were analyzed.

The utilization of data streaming is becoming essential in mobile computing applications to reduce latency and increase bandwidth.Bandi, Ajay Vast amounts of data are generated continuously from the Web sites of stock markets and financial institutions. The data’s meta-analysis is critical for investors and needs to analyze in a short time. Traditionally, it requires several heterogeneous resources with high storage capacity to process and compute the data. Data streaming helps to capture, pipeline, and compute the data without storing it. This research aims to visualize the continuous updates to the cryptocurrency temporal data using aggregations and simple response functions. The cryptocurrency data is collected from multiple data sources. A macro-enabled Excel external live data from Web feature, C3.js, and Tableau tools are used to capture and pipeline the streamed data in real time to make better decisions. The results show that the visualizations are dynamically updating in the events of trades in cryptocurrencies over time. Data streaming researchers and practitioners benefit from extending the streaming architecture methodology and dataflow to other domains.

Load Balancing is a significant process to dispense the numerous types of traffic into various servers and clients. The key aim of load balancing is every single server must not be overloaded with tasks. Disseminate the network traffic through various routes would reduce the congestion in the network and it would reduce the latency of the servers. Actually, load balancing ensures the server’s well-being. There is a high possibility of a server fall through when there is no load balancers assigned to servers. Load balancers enhance the overall proficiency of the servers and minimizing the load of the server by efficient traffic management. The OSI reference model characterizes four types of load balancing namely application, transport, network, and channel. This review article intends to analyze the pros and cons of layer 4 and layer 7 load balancing in the OSI reference model.

Internet of Things is an upcoming technology,Cherian, Mimi Verma, Satishkumar where IoT devices are interacting with cloud over Internet. Large number of IoT devices generate exponential amount of data that creates a huge impact on storage, network elements and specifically on the security and analysis of data. Recent research indicates that there should be a change in the networking paradigm to inculcate the dynamic demands of IoT environment. The network security issue like distributed denial of service [DDoS] attacks are of major concern, and its mitigation at the earliest remains vital. In IoT-related environment, the security issues of traditional network have major impact in IoT application domain. The IoT-related data that are depending on domain of application can be time sensitive or highly confidential, and hence, it arises the need to change the paradigm of traditional network. The expectant network should be more secure and flexible to detect and mitigate the network attacks. IoT environment with software-defined network seems to be promising enough to reduce many security issues with respect to IoT in traditional network environment. The proposed research work has created a test bed that collects IoT live data and sends it through secure SDN into the cloud platform.

The number of connected devices is exponentially growing in the world today and they need to work without having any interruption. This scenario is very challenging to cybersecurity and needs proper attention of network administrators, service providers, and users. Implementing security frameworks in this scenario is very difficult because attackers are using very sophisticated easy to operate weapons to launch huge attacks such as Distributed Denial of Service. Intelligently detecting and mitigating the attacks in the network requires the use of machine learning algorithms. This work proposes a strategic way involving feature selection based machine learning for the detection of stealthy attacks. The detection system works by performing ınformation gain-based feature selection as a preprocessing step. This ensures case-based preprocessing of each attack vector present in the traffic and is proved to be effective empirically. The proposed method has been tested using two supervised machine learning classification algorithms, namely Random forest and J48. The evaluation results show that the Random forest algorithm gives a satisfactory True Positive rate of 99.6% in detecting stealthy layer 7 attacks. The overall accuracy obtained is 99.81%. This approach causes the algorithms to exhibit improved performance while doing classification.

In this prevailing situation of automation, private data related to the public needs to be stored for all businesses and transactions. Most often, the private data of the patient’s name, mobile number, diseases, laboratory report, and treatment undergoing is breached by intruders. Automation in the healthcare sector can be made in a highly secured and less complicated manner by integrating patient’s health records and health insurance agencies by adopting blockchain technology. In this article, a real-time patient health monitoring system has been built and implemented on an IoT platform which can allow healthcare agencies like hospitals and doctors to monitor critical data of the patient in real-time and respond to the needs of the patient accordingly. The proposed system will prove to be beneficial for patients as well as the doctors and nurses and will lead to the implementation of an intelligent system which can automate various overviews being conducted by healthcare agencies adopting blockchain technology.

Vehicle-to-vehicle (V2V) communication is an advance application and thrust area of research. In the current research, the authors highlighted the technologies which are used in V2V communication systems. Advantage of such technology is that it helps to detect live location and tolling. It plays an important role if there are huge amount of traffic. The current research work can obtain more information about Li-Fi, RFID, VANET, and LORAWAN technology. Li-Fi is known as VLC communication system that uses visible light for high data transmission and reception. RFID technology helps the emergency vehicle to reach destination quickly by avoiding any kind of traffic. LORAWAN is a large-scale network technology with a long range and VANET with low power that allows to obtain accurate traffic information on each route and this saves time. The comparison between the different technologies is reviewed in order to obtain the optimized technology as per the applications.

Medical re-admissions are expensive and indicate poor quality of hospitals. A remarkable re-admission rate has a huge financial impact on the patients and the hospital. Through the increasing use of medical health records, enormous data is available to us for review which can identify high-risk patients effectively and decrease the mortality rate. According to the 2010 Affordable Care Act, hospitals may be penalized for re-admitting patients within 30 days of discharge. However, hospitals claim that the root cause of re-admissions lies similar to the populations served. In this paper, various re-admission prediction techniques that have been used earlier to estimate the re-admission rate of patients after being discharged from the hospital have been discussed. This article will also give a summary of various socioeconomic and demographic factors that play a vital role in medical re-admissions. Moreover, it will cover some of the measures that can be taken to reduce hospital re-admissions.

The information technology era faces an inflection spot due to current changes in the network industry due to the surrogation of cloud application implementation instead of data centre-based application implementation. The fiery growth of portable devices, sensor, and IoT-based devices would cause a stern effect on cloud application development and deployment in the network industry. The edge computing model is developed to enhance cloud application deployment. Edge computing is defined as where the computed calculations and storage of data are located in close physical proximity to improve the response time, reduce the latency, and improve the network bandwidth. This article analyses the ways to achieve the goals of edge computing with the assistance of load balancing.

Network-based information transmission has brought huge convenience for users in terms of the ease of use. However, the increased transaction not only lures fraudsters but also makes attack detection with a complicated process and mandates scalable models that can handle big data. This paper presents a network intrusion detection model that uses a hybrid ensemble model to identify intrusions in network transmission process. This work proposes the cross-bagging-based stacked ensemble model, which is a two-layered prediction mechanism used for operating on the complex network data. The first layer that contains a modified bagging mechanism is called the cross-bagging, where the results are passed to the second layer for final prediction. Experiments were conducted by using the NSL-KDD dataset. The usage of ensemble modelling enables the model to be effectively parallelized and also ensures high scalability of the model. This ensures effective prediction even on data with large volumes and high velocity. Comparisons with recent models show high performance for the proposed model.

Network security is considered as a major task in network architecture.Arefin, Md. Taslim Uddin, Md. Raihan Evan, Nawshad Ahmad Alam, Md Raiyan A network administrator had to focus, and it is defined and demonstrated as the rules, plans, and procedures followed by a network administrator to protect the network devices from different threats, and simultaneously, the passive and active attacks are generated from various vulnerable sources. Further, the unauthorized users must be prevented from accessing the network. There are different types of threats that need to be identified, explored, and take a step for preventing it, wherein the attacks are like DoS and DDos attracts, Aurora attacks, malware attack, port scanning, password sniffer, IP spoofing, session hijacking, and man-in-the-middle attacks, and cyber-attacks. This could be done with the help of firewalls, which can secure the network from malicious attacks. This paper is more focused on strong policy and performs incredible directions for averting the mentioned attacks. Firewalls are one of the strongest hardware attachments to secure the zone of networking sectors like local large, multinational, or enterprise networks. The deployment of firewalls that enforce an organization’s security policy is network devices. For this kind of tiresomeness, the concern of this paper is to enhance and develop network security like IPsec VPN, strong masquerades, port forwarding, create a trusted zone on WAN and LAN side, etc., based on the firewall by the execution of various tasks and different policies.

Due to the current exponential rise in users of the Internet, a need for highly scalable and easily configurable network devices is required. To meet this growth in demand, software-defined networking (SDN) has become increasingly popular. However, despite their several advantages, there still exists a gap in the level of quality of service (QoS) required in the existing fault detection and recovery mechanisms to promote a wide-scale carrier-grade network (CGN) selection. In this study, the basics of SDN were reviewed and compared the five diverse failure recovery models as a function of their QoS.

Recently, the idea has reached toward considering the emotions in the learning procedure which prompts to design an innovative framework that empowers correlative analysis and classifications of various emotional variations of an individual. There is an absence of teaching pedagogues to distinguish the standard articulations of people. In the past decade, research articles have recorded the lacking properties and attempt to recognize the equivalent benefit to overcome the difficulties. This enhances the part of identifying emotions as a device to perceive the sentiments of understudies while learning. This article encompasses the record of all analytical examinations of student’s emotions by applying different strategies, models, calculations and devices. This article wraps the considered works and gives the examination of qualities, shortcomings, openings, whose parts are addressed and the future work to be accomplished.

In the Internet of Things (IoT) paradigm, devices involved will frequently face different types of attacks, for example flood attacks, eavesdropping attacks and so on. Once attackers have compromised the IoT device, the data materials of the IoT device will not remain confidential, and it will then be captured by the attacker; this will in turn threaten the entire network. Consequently, to safeguard IoTs, this paper proposes a public and secret key with token sharing (IoT-PSKTS) algorithm to avoid key leakages in IoT. Cryptography can be utilized for secure communication in the presence of attackers. In cryptography, a conventional public key cryptosystem will be suitable because they do not require the sender and receiver to deliver the same secret to contact without threat. But, they regularly depend on intricate mathematical calculations and are thus much more incompetent than comparable symmetric key cryptosystems. In a large number of applications, the high price of encrypting lengthy messages in public key cryptography can be prohibitive. A hybrid system tackles it by utilizing a mixture of both. In IoT network, admin generates a public key, private key, secret key and token. The public and secret key will be used for packet encryption in IoT devices and base station side, and the private key will be utilized for decryption in the admin side and token used for IoT devices access control. For encryption purpose, admin shares public and secret key with a token for IoT devices and base station. Therefore, PSKTS algorithm has been used to securely share the public and secret key with a token for IoT devices and base station in a distributed way. The experimental results show that the proposed PSKTS algorithm shares a public and secret key with token in a secured way.

Today, the taskforce is getting increasing and the streets are getting more hazardous by the impact of blockage and increment of crashes. Intelligent transportation systems (ITS) are utilized to incorporate data innovation in transportation. Vehicular ad hoc networks (VANETs) are a subset of MANET which provides communication between the mobile nodes. VANET is a collection of various dynamic nodes that can change and configure. In VANET, enormous routing protocols are implemented to route the packet reliably. One of the protocols is ad hoc on-demand distance vector (AODV) routing protocol, and this protocol can only be used when the nodes are in static movement. The other protocol is destination sequenced distance vector (DSDV) in which each node maintains a table of information about the presence of the other nodes. In traditional system, only one algorithm used for the communication strategy, but in our proposed system GAD protocol (a combination of Genetic, AODV and DSDV) in which the functionalities of entire algorithms can be used based on the communication range in the network. Our proposed idea, using GAD protocol (a combination of Genetic, AODV, and DSDV) in which the functionalities of both the algorithms can be used based on the communication link in the network. Here proposed a numerical model to calculate the path duration between source and destination using GAD protocol and solve the Sybil attack, its severe attack on vehicular ad hoc networks (VANETs). The numerical model is simulated, and the GAD protocol is developed using MATLAB. The result exposes that when the transmission range increased then the path duration and numbers of hops become decreased in VANET.

In this paper, the performance study of free space optical (FSO) systems is completely analyzed, when the atmospheric channel is affected by atmospheric conditions such as haze and fog. Comparison of the FSO system’s performance with semiconductor optical amplifier (SOA) and erbium doped fiber amplifier (EDFA) under the influence of haze and fog is studied. The received power, eye diagram, and quality factor (Q-factor) are the performance metrics taken into consideration in this paper. Through simulations, it is demonstrated that the fog has a detrimental effect on FSO system performance, when compared with the haze. Further, the distance over which the FSO system can work reliably will be improved by using EDFA in the place of SOA for performing pre-amplification and post-amplification.

Websites are software applications that allow us to connect and interact with the data located in the web servers. Websites allow the user to capture, store, process, and exchange sensitive data like banking details and personal details. Web pages are accessed by merely entering the required URL in the browser. To prevent sensitive information from users, the attackers/hackers make duplicate websites and send them to victims through phishing emails. In this article, the machine learning framework is used to find malicious URLs. Here, five different machine learning algorithms such as the logistic regression algorithm, K-nearest neighbor algorithm, decision tree algorithm, random forest algorithm, and support vector machine algorithm have been used. An ensemble modeling has been done using these algorithms, and the performance of each algorithm has been compared.

Recently, wireless sensor networks (WSNs) incorporate their prominent role in various applications like monitoring and tracking remote environments. WSN exhibits a distributed nature and dynamic topology which increases the challenge of designing an energy-efficient protocol for routing. Enhancement of energy efficiency in WSN is considered as the primary goal of the routing protocol. This review mainly intends to discuss the hierarchical-based energy-efficient routing protocols to maximize a lifetime of network and energy efficiency. The objectives, challenges, and issues of the WSN routing protocols are also discussed in this review. Finally, the performance analysis for each energy-efficient routing protocol is also summarized in this article. In this, the systematic literature survey from 2010 to 2020 for hierarchical-based energy-efficient routing protocol has been carried out. From these reviewed details, the researchers can obtain a valuable technical direction while emerging an energy-efficient routing protocol. The information available in this review is helpful for various researchers to acquire significant information about the current status of WSN's energy-efficient routing and the various potential concerns that need to be discussed. Finally, the future aspects and research gaps for the reviewed protocols are also discussed.

In this new era of information technology world, security in cloud computing has gained more importance because of the flexible nature of the cloud. In order to maintain security in cloud computing, the importance of developing an eminent intrusion detection system also increased. Researchers have already proposed intrusion detection schemes, but most of the traditional IDS are ineffective in detecting attacks. This can be attained by developing a new ML based algorithm for intrusion detection system for cloud. In the proposed methodology, a CIDS is incorporated that uses only selected features for the identification of the attack. The complex dataset will always make the observations difficult. Feature reduction plays a vital role in CIDS through time consumption. The current literature proposes a novel faster intelligent agent for data selection and feature reduction. The data selection agent selects only the data that promotes the attack. The selected data is passed through a feature reduction technique which reduces the features by deploying SVM and LR algorithms. The reduced features which in turn are subjected to the CIDS system. Thus, the overall time will be reduced to train the model. The performance of the system was evaluated with respect to accuracy and detection rate. Then, some existing IDS is analyzed based on these performance metrics, which in turn helps to predict the expected output. For analysis, UNSW-NB15 dataset is used which contains normal and abnormal data. The present work mainly ensures confidentiality and prevents unauthorized access.

A sensor network consists of the random deployment of large numbers of tiny sized sensor nodes in a region of interest to detect the physical/environmental events and transmit the relevant data to the sink node through multihop communication. The nodes have many physical constraints like energy, memory, processing power, and hence the result is the limited or insufficient network lifetime of a network. To solve this problem, data gathering in an energy-efficient manner is an important task in the wireless sensor network to enhance network lifetime. Data aggregation is one such energy-efficient data gathering technique that reduces the data traffic and thereby the energy consumption substantially. The prime idea of the data aggregation is to gather, combine, and compress the data from various sensor nodes during transmission to the sink node. Among the available aggregation methods, in-network processing plays a major role to reduce the amount of data to be transmitted in the network. This article analyzes the various in-network data aggregation algorithms in detail and provides an insight into the techniques utilized.

The different methods used for classifying traffic along with the prediction of congestion and performance in software-defined networks were discussed. Although congestion prediction has foreseen many challenges, the algorithms did not give very accurate results. But over a period of time, several methods have been discovered to identify and predict the performance and congestion in software-defined networks (SDN). In this article, various techniques of classification were compared and predicted through tables and graphs.

To forecast conditions of action or actions during physical activity, the issue of classifying body gestures and reactions is referred to as human activity recognition (HAR). As the main technique to determine the range of motion, speed, velocity, and magnetic field orientation during these physical exercises, inertial measurement units (IMUs) prevail. Inertial sensors on the body can be used to produce signals tracking body motion and vital signs that can develop models efficiently and identify physical activity correctly. Extreme gradient boosting, multilayer perceptron, convolutional neural network, and long short-term memory network methods are contrasted in this paper to distinguish human behaviors on the HEALTH datasets. The efficiency of machine learning models is often compared to studies that better fit the multisensory fusion analysis paradigm. The experimental findings of this article on the MHEALTH dataset are strongly promising and reliably outperform current baseline models, comprising of 12 physical activities obtained from four separate inertial sensors. The best efficiency metrics were obtained by MLP and XGBoost with accuracy (92.85%, 90.97%), precision (94.66%, 92.09%), recall (91.59%, 89.99%), and F1-score (92.7%, 90.78%), respectively.

Mobile networks are playing a tremendous role in our day-to-day activities. In the currently evolving networks such as 5G, satisfying quality of service (QoS) is remaining as a challenging problem due to the dense network deployment. Moreover, multiple technologies such as LTE, Wi-Fi, and 5G contending and cooperating make the resource allocation a complex problem. This paper attempts to optimize the radio resource allocation in heterogeneous wireless networks for a particular geographical region by finding the throughput by maintaining the QoS along with a combination of network parameters. Specifically, the proposed research work uses different parameters such as RSSI, RSRP, and RSRQ for calculating the throughput of user equipment in a specified area. The feedback-enabled method (FETE) is then compared with and evaluated for the MIMO system, where it is observed that an overall throughput gain can be obtained by using right optimization technique for different parameters.

Nowadays, vehicle monitoring is emerging as a very tedious job, which requires the maintenance of the record or by recalling the date again and again for the service, and one more problem is tracking the vehicle location for providing better security and safety measures during the travelling. In both cases, it requires more human effort. The proposed model uses the novel technologies like IoT, cloud computing and machine learning. IoT allows various devices to interact and collect data like distance travelled, lubricant level, tyre conditions, smoke emission, other hardware parts conditions and also global positioning system (GPS) to track vehicle location. This data will be collected from various sensors like IR (infrared), MQ 6 sensor, HC SR 04 ultrasonic sensor, light-dependent resistor (LDR) sensor and stored in the cloud storage system. The machine learning algorithm is used to train the proposed model by using the samples data which is collected from the real-time vehicle service stations for service monitoring and GPS data for vehicle tracking purpose. Then this trained model is used to predict the vehicle’s condition based on that it will suggest the next date of service. This will help us to condense the quantity of human effort required to predict the vehicle service date. By use of previously fed data and algorithms used to analyse, the model is capable of providing the efficient result. Finally, the collected data is stored in the cloud storage and used to forecast upcoming service date, and all these activities like vehicle service date and GPS location data are provided through an Android application for ease of use to the user and the service provider for their need.

Pepper leaf disease detection is one of the interesting challenges in the field of machine learning. In this paper, a machine learning-based approach is proposed to extract texture features and use dimensionality reduction techniques called principal component analysis (PCA) and create a composite feature descriptor. There are two different texture-based feature representations extracted by using HOG and LBP feature engineering techniques were used for the pepper leaf images, and PCA is applied to obtain reduced representations. These representations are fused and passed to machine learning models like logistic regression, naïve Bayes, decision tree, support vector machine, and HistGradientBoosting classifier for classification. HistGradientBoosting classifier achieved highest the accuracy of 89.11% and outperformed other models.

One of the simplest methods to understand people's thoughts using images or text is commonly given as sentiment analysis. Sentiment analysis is used mostly in products advertisement and promotion depends on the user’s opinion. The process is based on the aspect-based sentiment analysis and it is used to understand and find out what someone is speaking about, and likeness and dislikeness. One of the real-world models of the perfect realm of this subject is the huge number of available Indian welfare plans like Swachh Bharat Abhiyan and Jan Dhan Yojna. In this paper, labeled data is used on the basis of polarity. Tweets are preprocessed and unigram features are then extracted. In the initial steps, tokenization process, stop word removal process, and stemming process are performed as preprocessing to remove duplicate data. The unigram features and labels trained by support vector machine (SVM), K-nearest neighbor (KNN), and a combination of SVM, KNN, and random forest as a proposed model are used in the presented work. Implementation of experimental proposed approach demonstrates that better results in accuracy and precision than SVM and KNN.