Advances in Electronics, Communication and Computing

In today’s world which is full of technological and unseen errors security is one of the major issues that should not be over seen. The technological and modern perspective has to be used to resolve the modern day problems. This project here in is based on Open CV face detection module to feature a face recognition system to identify and recognize the face of a person using certain facial feature which is again integrated with the traditional keypad pin input for ensuring the overall security throughout. It can be used as an access control system that is by registering the staffs, students, employee or officials of an organization with their faces, and later which can be used to recognize the people by capturing the images of the faces, the system show cases the accurate recognition which happens to become more accurate as time passes due to machine learning algorithm which enables constant learning of the system. The system is implemented on desktop using web camera or mounted camera; it first captures the image using the web camera or mounted camera and then applies machine learning algorithm to chalk out the features that could be used for the prediction at the time of implementation.

The firefly algorithm in the field of swarm intelligence is recognized as one of the widely demanded algorithm for machine learning applications. This algorithm is inspired by natural leaving pattern of firefly. This algorithm is modified by several researchers for getting better exploration of solution space for various applications. In all these modifications the variable step size firefly algorithm is gathering popularity in the field of machine learning because of its simplicity in modification. In this paper, this modified version is further enhanced by the addition of t-distribution function. This newly proposed version helps in the improvement of exploration along with the exploitation of the searched space to generate better solutions. Simulation of the novel projected version is done with standard benchmarking functions to prove the enhancement in the solution. The analysis of results proves the betterment of the solution in a variety of cases. This approach of modification can be utilized for applications in machine learning.

The number of Web attacks is increasing gradually, mainly the popularity of Web application in organization, school, and colleges. For this reason, the security of their sensitive information against attacker becomes very important for all organization and companies. In this paper, we describe different type of Web application attack like SQL injection, XSS attack, CSRF attack, and Buffer overflow. Besides, we discuss about different types of penetration tools for Web applications. Penetration testing try to find the vulnerabilities of Web application so that we can build a defense mechanism to deal with Web attack. Finally, we build attack trees and defense trees to represent the attacks and to prevent those attack.

Credit card sharp practice detection is one of the most important issues which must be motivated to save the financial institution from huge losses. Several machine larning models such as random forest, logistic regression, Naive Bayes, and XGBoost have been used to verify whether the transaction is fraudulent or genuine one. The data sets which is used in the research has been taken out from European Cardholder consisting of 284,807 transactions. As the data sets are highly imbalanced, so, SMOTE oversampling technique has been used. This experiment is carried out in three phases. First with individual standard model next with soft voting and finally with AdaBoost to know which model gives better results. F1 and MCC have been used for evaluation of the model as accuracy might leads to misclassification problem.

The method of data hiding has become one of the most trustworthy techniques to combat the malicious attacks on communication networks. Among various data hiding methods, watermark has turned one of its foremost technique, in which an image is veiled in a cover image. The paper proposes a Differential Evolution based watermark technique that eases to locate the covert region for data hiding where the unwanted attacks can almost be nullified. Least Significant Bits are used for veiling the watermark. The proposed scheme is carried out in the spatial domain. The experimental results show the competent quality of images are restored even after various attacks, as well as the watermark itself, is still in a recognizable state.

With the significance of embedded system in conjunction with the Internet, this paper aims at designing a baby monitoring system that maintains the baby in the absence of any manual surveillance. The prototype includes an external heat supply for the baby in correspondence to the increase or decrease in temperature along with a constant notification to the admin—a fan, which starts instantly over a heat-set as the threshold. For a safety measure, firstly, a baby’s cry is detected, which is indicated by a buzzer. Secondly, if the baby’s surrounding is encapsulated with smoke, then a led glows. The entire data is transferred to the cloud server (here, Blynk Platform) via ESP8266 with an Internet connection. This overall system will work for monitoring a baby without any manual hindrance.

In this paper, a copyright protection scheme for images has been developed through dual watermarking method. Here, the watermark, i.e., the copyright information has been embedded visibly and invisibly into the cover image. The visible watermark enhances robustness against signal processing attacks, whereas the invisible watermark is used to prevent different types of malpractices. The proposed scheme has completely been developed in spatial domain, and thus, the system complexity is very low. The system proficiency has been evaluated in terms of the three major qualities: imperceptibility, robustness, and data hiding capacity. The output results, being compared with some other existing methods, confirm the efficacy of the proposed scheme.

The paper describes an Edge of Things-based vehicle speed monitoring system that could be helpful in realizing the smart city. The IoT is a conceptual framework that includes several systems that collect information from the various environments of the traffic system. These collected data further send to the central control room by means of various access points. This whole process is controlled through a microcontroller unit (MCU). The proposed model also incorporated with a radar system that gauges vehicle speed through the installed surveillance system. Upon exceeding the speed, the vehicles can be traced with its properties like vehicle types, registration number of vehicles, and speed of the vehicle through the video or imagery data which is captured through installed cameras. These captured data further forwarded to the edge nodes for further processing and decision making. The computation or analysis work is done through the edge computing (EC). Then it stores in the cloud environment and necessary action can take by the authority. The model consists of different sensors and controllers to capture the required information.

Speech is the standard vocalized communication media. Speech is one of the comfortable way for humans to communicate with each other. Similarly, speech recognition system is eagerly necessary to communicate with computer through voice. Speech recognition in English language already helps us to operate English voice command-based applications. But in rural and semi-urban areas, due to lack of knowledge in English in India, it is necessary to implement automatic speech recognition in regional languages. Here, we have built a Gaussian Mixture Model (GMM)-based Bengali (also called Bangla) isolated spoken numerals recognition system where mel frequency cepstral coefficients denoted as MFCC is taken for feature extraction. The proposed system achieved 91.7% correct prediction for the Bangla numeral data set of 1000 audio samples for 10 classes which is satisfactory for previous Bangla spoken digit recognition.

Nowadays, many people are suffering from chronic kidney disease worldwide. Factors responsible for such conditions are food, living standards, and the environment. Detection and identification of chronic kidney disease are costly, time-consuming, and often risky. Therefore, the early detection of such disease is very important. In this research study, we have tried to reduce the clinical effort by automating the process of detection. We have classified whether the person is suffering from chronic kidney disease or not. We have used different classification algorithms and regression algorithms like KNN, SVM, Naive Bayes, and logistic regression. We have got some good results in all the algorithms but KNN performed very well.

In today’s world, the frequency of occurrence of road accidents has increased exponentially. The news channels broadcast at least one case of road accident every day. So, it is very much required to find out the root cause of accidents so as to prevent them. This report will give a brief idea about analyzing the reasons behind the occurrence of road accidents and its analysis. It will also brief about the systems that has been created to provide safety measures after the occurrence of accidents. Based on our analysis, we discussed about some of the advantages and disadvantages of the existing accident alert system and shared our opinion that will have an impact in preventing the occurrence of road accidents. Finally, the future work is to include IOT techniques and sensors to analyze the cause of accidents as well as prevent the same.

Brain tumor accounts for 80–90% of all primary CNS tumors. Brain cancer is tenth leading cause of death. Automatic detection and classification can help in early diagnosis and deliver efficient treatment. Recent developments in medical imaging modalities such as MRI provide insightful image of our brain. Due to large amount data and variability of data, diagnosing in faster and proper manner is not humanly possible. In this paper, we focus on segmentation of brain tumor of 3D MRI images using 3D CNN in cascading format. Whole tumor is extracted from the images through the first CNN model, then the output is fed to the next CNN model to extract the core tumor and finally, it is fed to the last CNN of cascading network to segment the enhanced tumor core. To improve the brain tumor segmentation, three neural networks belonging to the class of convoluted neural networks (CNNs) were connected having 20 interconnected kernel slices with four downsampling slices to trade off simplicity with feature extraction. The proposed method was evaluated by considering Brats 2015 3D dataset consisting of 274 MRI images with their ground truth having different four modalities. The activation function used was Relu and the results obtained were calculated over DICE coefficient (F-measure) which was found to be 0.78 for core tumor in flair modalities.

Deductions are predominantly the short payments done for a generated invoice usually by the customer as a compensation or for the lack of products or services. Possible reasons for deductions to happen include shortage, damage, late delivery, and other-related factors. The machine learning approach has a huge impact on the deduction domain as eliminates the manual effort of a deduction analyst without compromising much on the accuracy. A deduction analyst can save so much on time as now he/she does not have to go through the complex procedure of deduction validity or invalidity. Also this solution will help in speeding up the business process which will lead to customer satisfaction due to on-time delivery. In this research, various machine learning techniques like LGBM and random forest are used for the analysis. It was observed that LGBM model provided optimum result thereby helping business analysts to take decision with respect to invoice payments.

The wide application of machine learning in dominant domains such as marketing, telecommunication, agriculture, and other industries has made an impact on its use in several other time critical applications. Health care is one of the vital sector where machine learning is finding acceptance in disease diagnosis. Though the medical zone is rich in raw information, but somehow not all information are successfully extracted that is needed to disclose uncertain trends & efficient decision making. Extraction of these uncertain patterns and associations usually turns unexploited. Modern optimization methodologies may be helpful in dealing with this scenario. In this research work, it is intended to use classification-based modelling algorithms which include Naïve Bayes, decision trees, artificial neural network (ANN), and support vector machine (SVM) with the use of health-related attributes like age, gender, level of blood pressure, and blood sugar, it can be used in predicting the probability of patients inheriting various disorders related to heart. Eventually, genetic algorithm is used as a feature optimizer which extracts the relevant attributes for classification. It is observed that with the use of genetic algorithm, the classification performance is enhanced with the implementation of the above classifiers.

Samples of data may consist of numerous attributes and variables which are irrelevant and redundant. Some of those attributes may not be of any vital use in classification and the irrelevant attributes can decrease the efficiency. Thus, the feature reduction process can be considered as a problem in machine learning which selects less quantity of vital attributes to obtain higher accuracy rate. This process minimizes the attributes count by eliminating less relevant and noisy samples from the data set to achieve better classification accuracy. This work uses particle swarm optimization (PSO) search algorithm for feature reduction in cervical cancer data set. The experimental result shows that the irrelevant features are removed and only 17 useful features are selected, out of which 36 in the cervical cancer data set.

The cardiac disease plays a major cause of death worldwide. The medical experts are facing the difficulties to foresee the heart attacks, because it seems to be a complicated job and also requires huge skill and knowledge. Today's health division consists of some crucial information which becomes significant to make decisions. To predict heart attacks disease, algorithms like J48, Naïve Bayes, REPTREE, CART, and Bayes Net of data mining get used and also applied in this research. The study result shows 99 percent predictive accuracy. Data mining allows trends in the data set to be predicted by the health sector.

Evaluation of healthcare policies and taking decisions with regard to complex problems require assessments at many levels and by a group of experts. This paper studies the selection criteria and their weights for the five drugs for metastatic colorectal cancer treatment. A comparative value assessment of the drugs was conducted with the analytic hierarchy process (AHP) and the fuzzy analytic hierarchy process (Fuzzy AHP) techniques of multi-criteria decision making. The ranking scores of all the alternative drugs have been examined and the implications of the vagueness in the decision making have been scrutinized for both the AHP and Fuzzy AHP.

Wireless Medical Sensors Networks (WMSNs) are mainly used to track and collect health data from patients using body sensors which are implanted on the patient body. This helps to reduce the cost of treatment, provide reliable resources and services, provide accurate diagnostic results, and quicker diagnosis. Security and privacy are the most significant challenges in such networks while the healthcare information is transferred via wireless communication to the medical server. Security requirements such as confidentiality, honesty, and authenticity should be maintained throughout the entire process. This paper proposes a robust privacy preservation framework based on homomorphic encryption. The randomness in encrypted data provides different ciphertexts which enhances the privacy of the processed data. The network and encryption performance of the proposed monitoring system are evaluated using an OPNET-based simulation model. It demonstrated that the proposed framework achieves better throughput and less delay as compared to previous works.

In current scenario, security is one of the major factors. People wants to be more secured by applying the current technology in the day-to-day appliances. In this paper, we are proposing an improvised IOT-based doormat which senses when anybody comes at our doorsteps, and it sends a notification in our mobile. With the built-in force sensing resistor, the doormat is able to monitor changes in pressure whenever someone steps over it. Smart mat is capable of providing a wide range of applications by sensing a user’s identity and activity in the house. This paper shows the various advantages of Arduino Uno and the various fields it can be applied.

The most commonly used platform in technical field is artificial intelligence, and machine learning is a subset of it. It is very well known about its model training and testing. Nowadays, it is used in every platform, and surprisingly, it is giving unexpected result. Like in recent studies, it is found that in mall the most buying product with beer is diaper, but as a person, we never even thought of this combination. The most cautious part of human life is heart. The moment it will stop working, we will be dead. So by looking at this concern, we have done our survey in the Internet, and all the sites says that after retirement, 90% people are being affected by heart. The cause of its affected may be directly or by any other disease affecting it. We have shown our concerned toward the heart disease and chose the machine learning platform for the analysis. In this paper, we are going to predict the heart disease by providing a brief introduction about all the machine learning algorithms their advantages and disadvantages.

In the current scenario, energy conservation is the need of the hour. Conserving energy not only includes making limited resources last for long but also reducing demand on a limited supply. People need far greater and more detailed feedback regarding energy use. The proposed model is an improvised IOT-based smart energy meter which can record the power consumption and then deliver the same data to the concerned user for tracking the daily power usage and accordingly switch off appliances to keep the energy usage under control. Smart energy meters can be used for a variety of applications by identifying the energy consumption of individual devices.

For a growing population, an innovation of the hybrid system is there to produce demandable electric energy. At the same time, the environmental condition is also changing rigorously and concentration of GHG also increasing due to burning fossil fuels for heavy energy demand. And the effect of climate change directly depends on renewable energy sources which are now the fastest-growing sources in the world as solar PV technology is one of the demandable renewable energy sources so in this paper, some experimental analysis has done on solar PV panel. But this system is affected by so many environmental factors by which the efficiency of the PV system is disturbed. The module efficiency is mainly affected by ambient temperature, module temperature, incoming solar irradiation, and photovoltaic module material composition. In this paper, an experimental result has carried out, which shows the effect of an increase in greenhouse gas on solar photovoltaic performance.

Many factors including noise are there for degradation of the voice signal that needs to be enhanced. In this paper, we have taken an attempt to improve the speech quality. The averaging spectral subtraction (SS) method is modified with the Welch method. In first phase, the Bartlett method is used and in the second phase, the Welch method is used as the novelty of the work. The method is based on the spectrum dependent gain function and the noise is estimated using Welch method. The overlapping windowed frames and the adaptive averaging of the gain function result better enhancement. The results obtained using the proposed method are shown in the result section. The comparison results prove the proposed algorithm is better for tidy noisy signals. For quality measure of the signals, signal-to-noise ratio (SNR), segmental SNR (SegSNR), and mean opinion score (MOS) are evaluated.

Optical communication is an indispensable technology for current society requirements. It satisfies the short-distance and long-distance communication with the help of different mode analysis of variety type of fibers. In this paper, an analytical study has been presented that can help the next-generation high-speed secure communication. The advantage is that, it can act as the filter in terms of signal processing and channel in terms of communication. The waveguide can satisfy in both the way. For revolutionary communication and signal processing, it is found that optical waveguide research can satisfy future generation communication. The suitable design can be used for all optical signal processing. This article can provide insight to the researchers working in this direction.

Electroencephalogram (EEG) signal processing is an emerging research area out of many biomedical applications. Brain waves study and characterization can be done for various faults. However, artifacts within the signal are like inherent property. The removal of artifacts is the major challenge that has been considered in this work by authors. The removal process is transformed based application within independent components. In first stage, the independent components are derived from the raw data. Further, the artifact channels are identified using statistical approach. Scaled entropy and kurtosis are used for it and fixed the threshold level. Finally, the application of discrete cosine transform (DCT) provides the clean signal that is used for analysis and diagnosis. The results are exhibited in the result section and compared with earlier methods.

Character recognition in handwritten data is a challenging work as the writing style varies from person to person. Choosing a method for handwritten numeral recognition is also having importance as the result depends on the method used in the recognition model. Deep learning-based recurrent neural network (RNN) is being used for prediction in time series data, generation of text lines, and other sequential data processing. But in this work, the pixel values of image are used as time stamp dependent input to the recurrent networks. We have applied simple RNN, long short term memory (LSTM), and gated recurrent Unit (GRU) for recognizing Odia handwritten numerals. A comparison study is also provided to understand the effect of vanishing gradient in RNN and how this drawback of RNN has been overcome by LSTM and GRU cells. The Adam optimizer is used in each proposed method. The accuracy values obtained in RNN, LSTM, and GRU are 50.04%, 88.81%, and 86.24%, respectively.

The physical signals are inherent in nature. The noise may be due to external environment or physiological change. These signals are non-stationary. Though it is a difficult task to suppress noise from non-stationary signals, enormous works have been done using different algorithms including LMS. In this work, consider two cases: (a) Speech signal (b) cardiac signal. In variant of least mean square is used for verification including LMS, NLMS, and DLMS. Gaussian noise is considered for speech signal due to environmental effect, whereas impulsive noise is taken for ECG signal, due to either artificial (cardiac) or spikes occurrence at the time of data acquisition. For current scenario of pipelining and parallel processing, authors have suggested delayed LMS for both the types of noises and found better performance in terms of SNR, stability, and convergence as compared to NLMS and LMS.

In this paper, a novel circular patch antenna with annular rings fed by a microstrip line is proposed. The antenna has been designed for ultra-wideband applications. The proposed design exhibits an ultra-wideband response centered around 24 GHz. The antenna covers the entire K-band and may be used in satellite communications. The antenna occupies a very low volume of 1317.7 mm3. FR4-epoxy is used as both substrate and superstrate in this design. The simulated performance of the antenna has also been presented.

The aim of event detection is to identify interested events in a user-generated content using multiple modalities automatically. However, it is a challenging task particularly when videos are captured in a restricted environment by nonprofessionals. Such videos suffer from poor quality, deprived lighting, blurring, complex camera motion chaotic background clutter, and obstructions. However, with the rise of social media, there is rising popularity of user-generated videos on the Web day-by-day. Each minute, 300 hours of user-generated video are uploaded on you tube due to which people find difficult to search the appropriate content among a large number of videos. Therefore, solutions to this problem are in great demands. In this paper, we study existing technologies for event detection in user-generated videos using multiple modalities. This paper provides key points about feature representations across different modalities, classification techniques.

Application of machine learning in healthcare sector is increasing day-by-day. It can be very useful for automated and early diagnosis of different diseases. In the proposed work, authors have compared the classification performance of three different classifiers for cardiac signal classification. The ECG data is collected form Physionet database. Relevant features are extracted from the original signal by applying dual tree complex wavelet transform (DTCWT). Multi-layer perceptron (MLP), radial basis function (RBFN), and support vector machine (SVM) classifiers are considered for classifying the cardiac signal. From the result, it can be observed that, SVM is performing better as compare to other two types of classifiers.

A priority analysis and prediction is required to study about consistent growth in Science and Technology Sector (S&T). It is by developing a artificial intelligence tool to bolster technological growth by studying and comparing behavioral factors of temporal and current data. Obtain the epidemiological, temporal, and cohort study about Technological Growth (TG) on the sectors such as Environment Engineering Science (EES), Health Engineering (HE), Energy and Resource Engineering (EE), and Agricultural Engineering (AE) to find deviated habitat behavioral data for observation and analysis to strengthening Technological Growth (TG). Knowledge of deep learning in hybrid intelligent system tool confers an alarm and pavement to smart, healthy, and quality innovation to overcome the interdependent pitfall.

Internet is a powerful platform for students to develop the areas of interest and improve existing skills. Students in the age group of 18–25 are most frequent users. People in general, especially students, use the Internet mainly for research and educational purposes. The way in which people use the Internet meaning the websites browsed indicates the behavior seeking perspective. The outburst of Internet has aided students in many ways, but it also brings negative impact in academic performance. Study on people has also told that the balance maintained between students with study and Internet has a good impact in their academic performance too. The proposed work uses machine learning algorithms to quantify the relationship between performance in academics and behavior perspectives of a student in the usage of Internet and to bring out novel features that have a generalized value. Students are grouped according to academic performance and grades obtained with further processing by decision tree, support vector machines (SVM), and neural networks algorithms. Students’ Internet logs could be obtained and expose affluent information on students’ behavior. The proposed work has strong practical value for improving management skills of students’ in education with the university and college sector.

Egocentric vision, an offshoot of computer vision, offers unique insight into the visual world by examining images and videos taken by a digital camera. This research, with the aid of ecocentered view, describes the actions of the different dog breed in different situations. It will improve predicting the breed behavior of the dog and gaining more knowledge of the breed of dog. To understand the breed behavior of the dog is to evaluate the historical breed data of the dog with current data. The comparison between dog’s current situation and dog’s activities will predict behavior in indoor and outdoor situations.

Recognizing body movements are vital for fields like virtual reality, human computer interaction, and safety monitoring. The research seeks to overcome the limitations of existing body movement recognition systems, which include individual human body recognition rather than multiple recognition and also fails when the body is partially visible. Since most of the real-world applications, like accident monitoring requires recognition of multiple people; at the same time, the existing methodology has reached an impasse. Classification algorithms are predominantly used to recognize various body movements and actions. Even Kinect, which is well known for recognizing human actions and movements, uses a random forest algorithm to track the movements. With recent technological advancements, researchers are keen to increase the accuracy of tracking the movements. New research methodologies seek to overcome the above limitations by using deep neural networks and many other machine learning algorithms. With the inspiration of the success of convolution neural networks (CNNs), the implementation explores usage of the mentioned deep learning algorithm in estimating human poses and recognizes corresponding actions. The idea is to build a scalable model that can be used for many applications and get accurate results as much as possible.

In the era of wireless communication, internet devices such as smart phones, hotspots, and Wi-Fi zone are important player of rapid growth of data usage. Internet connection devices are building new challenges for internet service providers such as higher bandwidth and indomitable increasing users day to day. This article gives an overview of existing different technologies of wireless communication as well as future enhancement of wireless services is explained. Different techniques during last decade are introduced for handling next generation techniques for ISPs. We provided a comparative analysis of different generations such as 3G, 4G, and 5G, and we focus on the next generation communication technologies 6G and 7G. Current research works are focused on 5G and next era generation of communication technologies. The development of internet infrastructure is essential for better mobile broadband experiences. In this work, we compared all the available generation communication systems, either wireless or wired, that are being deployed as well as future directions.

Green energy has led to an increase in biogas plants and consequently to an increased use of organic waste. Kitchen wastes are reliable because of their potential to yield methane. Biogas is a combustible mixture of bio-methane typically 50–75%, carbon (IV) oxide and also other trace gases. By monitoring the system with an indicator of digester behaviour and various parameter sensors, the plant can be operated successfully. For the increased performance, temperature is an important factor. Solar assisted biogas generation system offers several benefits. Thus, by increasing digester temperature using solar aided biogas is recommended to obtain higher amount of biogas yield. Reasonable pH moisture is vital for a fruitful anaerobic digestion course. Small particle sizes, suitable mixing of waste and water ratio are important factors too. This paper presents aspects of biogas plant with a smart system to monitor the system and enhance biogas production. It emphases on effective robust indicators, applicable monitoring methods. The details and facts presented here intend to boost and to foment biogas via anaerobic digestion. This is also beneficial for the development in research of biogas monitoring and management system.

In this paper, the effect of exogenous electromagnetic fields is used to understand the behavior of biological entities. However, in some cases, endogenous signals could be preferred as it does not require external excitation. On the other hand, exogenous signal can be used for both internal as well as external electromagnetic excitation for studying the effects or presence of various biological entities. This study will lead to the understanding and development of non-invasive diagnosis techniques and non-medicated disease treatments. The method will be further extended by applying various feature extraction and machine learning algorithms to make the detection real time, without the need for additional processing of samples or chemicals.

The wireless sensor network (WSN) consists of a large number of low-cost sensor nodes with limited resources. The limitations of the wireless sensor node are its features which include reduced memory, low computing power, are implemented in a hostile area and are left unattended, a small range of communication capabilities and low-power consumption capabilities. The basis of these characteristics makes this network vulnerable to numerous attacks, such as the sinkhole attack. The sinkhole attack is a type of attack in which the compromised node tries to attract network traffic by announcing its fake routing update. One of the impacts of the sinkhole attack is that it can be used to launch other attacks such as the selective forwarding attack, to recognize the phishing attack and the spills or modified routing information. It can also be used to send false information to the base station. This survey paper focuses on exploring and analyzing the existing solutions used to detect and identify the sinkhole attack in the wireless sensor network.

This paper presents a signal processing and machine learning-based approach to classify different types of arcs due to the interaction of a medium voltage distribution line and different surfaces. Different kind of arcing surfaces, i.e., concrete, wet-sand, grass, and leaning tree, are considered in a real-time environment to create different arcs. The similarity found in various arcing events is the low (in mA) current flowing during the arc. The voltage signals are taken as the basis of the whole analysis. The signal processing technique used in this study is empirical mode decomposition (EMD). The results obtained by the application of EMD along with different support vector machine (SVM) techniques on voltage signals successfully classifies various high impedance arc faults (HIAFs) for various arcing surfaces based on their harmonic footprints.

In this study, an analysis of static and dynamic arc in sphere gaps and leaning tree is carried out. A Stockwell Transform (ST) or S-Transform-based approach is used for the detection and classification of the high-impedance arc signals. Conductance variations in this scenario are also observed for various voltage levels and sphere gap length for different arc. The findings obtained by applying the ST technique to arc voltage signals effectively identify and distinguish the high-impedance arc based on their frequency signatures due to leaning tree from sphere gap arc.

The level to which a driver worry affects the driving by causing numerous accident. In some cases, numerous lives are lost due to driver anxiety. Stress is something we can’t ready to truly observe and recognize. Hence, various kinds of sensors like heartbeat, ECG, pupil dilation, pulse, breath rate, skin reaction are used to anticipate feeling of anxiety of the driver and driver behavior while driving. Also, in this system, we detect vehicle accidents using flux sensor. Flux sensor would be able to detect car hits and accidents by analyzing the pressure. If the accident is detected an alert message or call is sent to the respective bus authorities, police control room, and hospitals.

All around, air sullying recommends the proximity of deadly substances into the air that is hindering human thriving and the planet considering. It is worth everything considered to be portrayed as one of the most dangerous perils that mankind at whatever point standing up to. It makes hurt creatures, harvests, timberlands, and so forth. To frustrate this issue in transport regions need to imagine air quality from harms utilizing AI structures. In this way, air quality appraisal and need have become an enormous research locale. The fact of the matter is to investigate AI-based frameworks for air quality assessing by need achieve the best precision. The appraisal of the dataset by directed AI strategy to get a couple of data resembles a variable explicit check, uni-variate assessment, bi-variate, and multi-variate examination, missing worth game plans and separate the information support, information cleaning/getting ready and information assertion will be finished everything considered given dataset. Our evaluation gives an extensive manual for the affectability examination of model parameters as to execution in line of air quality sullying by precision estimation. To propose an AI-based framework to effectively foresee the air quality index, the central purpose by need accomplishes the sort of best accuracy from looking at controlling solicitation AI calculations. In addition to that, different AI estimations are calculated from the given vehicle traffic office dataset with an appraisal of GUI-based UI air quality check properties.

Sign language is popular with hearing-impaired individuals around the globe. There are a pair of languages that employ predefined steps as well as motions to express a personal message. These languages are largely created to help deaf and verbally inhibited individuals. They normally use a precise and simultaneous mixture of motion of hands and wrists, the orientation of hands, hands styles, etc. Various areas have varying indication languages such as American Sign Language, Indian Sign Language. In this paper, we have concentrated on Indian sign words. This particular research work is designed to exhibit a simple strategy toward sign dialect to textual content transformation through personalized region of interest (ROI) segmentation as well as convolutional neural network (CNN). Multiple hint gestures are educated utilizing a customized picture dataset as well as applied by Python dialect. Making use of the ROI choice strategy, the task displays much better results compared to traditional methods within the terminology of real-time detection and accuracy level from video clip streaming by webcam.

Brilliant innovations within reach have encouraged age and assortment of immense volumes of information, on regular schedule. It includes exceptionally touchy and differing information like individual, authoritative, condition, vitality, report, and personal information. Information analytics give answer for different issues being looked by brilliant hospital information like emergency reaction, debacle versatility, rise the executives, doctors, the executives framework, and so on. It requires conveyance of delicate information among different elements inside or outside the hospital. Sharing of touchy information makes a requirement for effective utilization of brilliant hospital information to give savvy applications and useful to the end users in which are liable and experimental method. This mutual delicate information if gets pilled as an outcome can make harm and extreme hazard the hospital assets. Strong hold of basic information from informal exposure is greatest issue for accomplishment of any task. Information leakage perception gives a lot of instruments and innovation that can productively settle the worries identified with patient basic information. The principle goal of this task is to identify the guilty operator/individual who are indented to hack the information. We additionally send blockchain idea over this undertaking for high security. We give fake/copy record to those liable individual.

The primary point of this venture is to give a solid and secure cloud support and furthermore increment dependability of confirmations by continuous auditing. Cloud service certifications (CSC) are a decent way to address a more prominent degree of security. Keeping in view that, cloud administrations are a piece of persistently evolving condition; question unwavering quality of such confirmations may happen in multi-year legitimacy periods. To expand dependability of accreditations, it is must to guarantee consistently solid and secure cloud administrations. CA of cloud administrations is still in its juvenile state. In this manner, we directed an exhaustive workshops, interviews writing survey with professionals to finish engineering for nonstop cloud administration inspecting. However, outsider evaluating techniques are not accessible in existing strategies. In this way, we propose a potential methods for execution that demonstrates different advantages and changes that must be figured out how to diffuse the idea of persistent cloud administration inspecting. Evaluators and suppliers who are connected together in a calculated design that are getting advantage over pertinent inward and outsider reviewing approaches. The security examination of the proposed open inspecting system demonstrates the accomplishment of wanted properties, and execution has been assessed with the itemized try set. The outcomes show that the proposed secure cloud reviewing system which is exceptionally secure and productive away, correspondence and calculation costs.

In software engineering, the single point is to deliver top notch yield while enhancing the expense and the time expected to finish the application development. To accomplish this objective, software groups will play out the test on their application before live creation. For test automation documentation assumes a critical job. This paper center around mechanize experiment age dependent on accessible test assets, challenges which could be conveyed by methods for orderly practice improvement and characterizing test system for any software application. One noteworthy motivation is to ensure perceptibility among necessities and system experiments. Along these lines, the significance of experiments is dull and testing, especially under time objectives and when there are progressive changes to necessities. Right now, customized test age diminishes the cost of testing just as guarantees that experiments properly spread all necessities, a huge objective in prosperity essential systems and for the measures they need to come. This proposed application guarantees proficient test inclusion of software surrenders, where key usefulness won't be missed in the automatic test absconds expectation. Right now, experiments can be produced after application development completes a component or a lot of highlights.

Right now, present a novel machine learning-based ship detection for pre-explained transport database, which is intended for preparing and assessing transport object recognition calculations. Programmed object identification in the oceanic condition has gotten significant, with a wide cluster of utilizations in regions, for example, maritime fighting, vessel traffic administrations and fishery the board. The manual detection of items is not extremely effective due to the intense climatic conditions, for example, mist, downpour, storm and so forth. It might prompt flawed expectations and it is absolutely reliant on the ability of the individual. This adds to the usage of programmed object recognition component. The acknowledgment of inshore and toward the ocean ships is a major task for an enormous combination of employments in both military and non-military faculty fields.

Measure structure framework attempt to apply based directory listing allows search questions and supports encoded information at a decent pace in the fine-grained cloud location. In any event, earlier CP-ABKS plans were used to support in reality, unexpressed tri-owner frameworks cannot be legally related to the tri-owner setting), in the absence of acknowledging higher computing cost and point of confinement value. What is more, because of security weights on find a workable pace, existing plans are vulnerable against isolated keyword guessing ambushes if the keyword range is of larger space. Likewise, it will be hard to see malevolent clients that releases riddle keys when many information client has a relative subset of chrematistics. At this moment, present a security saving CP-ABKS framework with camouflaged access approach in shared multi-owner setting (essential ABKS-SM structure), and show how it is improved to help hurtful client following (adjusted ABKS-SM framework). We can show that by ABKS-SM structures accomplish explicit certainty, negate isolated keyword hypothesizing snare in the standard bilinear party model. We moreover assess their introduction utilizing authentic world datasets.

Parkinson disease (PD) is a progressive neuro degenerative disorder that impacts more than 6 Mio. People around the world. Nonetheless, non-specialist physicians still do not have a definitive test for PD, similarly in the early stage of the diseased person where the signs may be intermittent and badly characterized. It resulted in a high rate of misdiagnosis (up to 25% among non-specialists) and many years before treatment, patients can have the disorder. A more accurate, unbiased means of early detection is required, preferably one that individuals can use in their home setting. The proposed system for predictive analytics is a mixture of clustering of K-means and a decision tree used to gain insights from patients. The problem can be addressed with reduced error rate with the application of machine learning techniques. Our proposed system also produces accurate results by combining the spiral drawing inputs of patients impacted by common and Parkinson’s. From these drawings, the principal component analysis algorithm (PCA) for extraction of the feature from the spiral drawings and support vector machine is used for classification. UCI machine learning platform voice data collection in Parkinson's disease is used as feedback. Thus, our study results will show early detection of the disorder can promote the therapeutic care of the elderly and increase the chances of their life span and healthier lifestyle living peaceful life.

In the current framework, the appropriated, dynamic attributes, and the cooperation prerequisite make it face numerous new security and protection give that can’t be unraveled by the conventional open key or symmetric cryptosystem. The objective of the paper is to build a numerical model of fog registering and survey its relevance with regards to IoT, where it is significant to fulfill the needs of the idleness touchy applications running on the system board. The work further plays out a near presentation assessment of distributed computing with that of fog figuring for a situation with a high number of Internet-associated gadgets requesting ongoing administrations. In the proposed framework, the mist processing worldview to serve the requests of the inertness touchy applications with regards to IoT. By moving a data sensor, the IoT relies on a distributed calculation. This is a decentralized condition method to collect data from any place in the city. The system will test the vitality of every server and its region. Since at whatever point server transfers the subtleties of the sensor which will eventually weaken its vitality. We therefore need to transfer the information by allocating another server with the vitality to submit the information.

Cloud computing is broadly received gratitude to the high business spryness; it assurances to its shoppers. Cloud administrations arrangement in the commercial center is an essential undertaking for both cloud customer and cloud supplier principally if the cloud buyer asks explicit properties for its applications. In this manner, this undertaking is designated to a third part which is the cloud intermediary. The proposed framework a cloud asset dealer is suggested that will administer the task of suppliers’ assets to purchaser progressively. The proposed representative uses different prerequisites and imperatives determined by the shopper in the necessity depiction format as contribution, to figure amassed necessities, utilizing an accumulation calculation. Further, the administration planning calculation is characterized to discover a streamlined match between the accumulated necessities with the suppliers contributions. From that point, this calculation is executed regularly, in view of a procedure for dynamic planning to profit customers by virtue of presentation of new supplier or some great contributions. Results demonstrate that the arrangement gave by agent ends up being a success win circumstance for the customer concerning cost just as execution. We propose a monetarily roused pay way to deal with increment the granularity and utility of saved calculation and capacity administrations. The result is a profoundly virtualization of cloud asset intermediary. It comprises of a guardian asset administration commercial center and a shopper configurable virtual machine for asset sharing. The framework bolsters progressively settled virtualization with powerfully customizable asset limits for fine-grained auxiliary, worldly, and vertical–spatial versatility.

Advanced advertising is viewed as the favored strategy contrasting with conventional showcasing. It is valuable to the two professionals and scholastics of web-based social networking promoting and buy expectation. The exploration gives some underlying bits of knowledge into shopper points of view of web-based life advertisements and online buy conduct. Business, academician, specialists all are share their notices, data on web so they can be associated with individuals quick and effectively to study on accessible item sites by web scrap. Web scratching is a robotized technique used to remove a lot of information from sites, and the information on the sites are unstructured. To forestall this issue, web scratching helps gather these unstructured information and store it in an organized structure. Consequently, client cost and rating of item assessment and forecast have become a significant research region. The point is to examine given dataset utilizing AI-based procedures for item appraising determining by expectation brings about best precision. The investigation of dataset by support vector classifier (SVM) to catch a few data resembles variable recognizable proof, univariable examination, bivariable and multivariable investigation, missing worth medicines, and dissects the information approval, information cleaning/getting ready, and information perception will be done on the whole given dataset. Our examination gives a far reaching manual for affectability investigation of model parameters concerning execution in forecast of item appraisals with value subtleties by discovering precision estimation.

Publicizing extortion, especially click misrepresentation, is a developing worry for the Web-based promoting industries. The utilization of snapbots, malware that naturally taps on advertisements to create fake traffic, has relentlessly expanded in the course of the most recent years. While the security business has concentrated on distinguishing and evacuating noxious doubles related with click bots, a superior comprehension of how fraudsters work inside the promotion biological system is should have been ready to disturb it productively. The demonstration of tapping on an advertisement, not due to enthusiasm for this advertisement, but instead as an approach to produce unlawful incomes for the application distributor.

Proposed approach illustrates Detection touch-less heartbeat and a cardiopulmonary sign. A microwave frame is attempted to find a good way off of 1 m from the pulse sound of a person by using a vector organize analyzer. The proposed method requires the ability to distinguish pulse patterns with both recurrence and the plausibleness of the force tuning. Calculations are made at 2.4, 5.8, 10, 16, and 60 GHz, and also at varying force rates between 0 and −27 dBm. A classification that applies cardiopulmonary activity is defined in terms of the air and heart beatings statistics. Using wavelet and wide channels, heartbeat rate and heart rate changeability are isolated from the test signal for SNRs between 0 and −20 dB.

Android and Mac OS apps have become an important asset of our daily lives of mobile device users by which translates into an increase in mobile applications. Now, a one-day user can access an ample amount of applications through different platforms like play store, apple store, etc. Due to a certain amount of vulnerabilities, hackers are developing mobile malware, which in turn threatens the system and can lead to remote control, loss of privacy, etc. Therefore, it is necessary to detect the threat level of a certain application installed on mobile devices. In this next module, we present an approach in which we allow the user to select any application from the play store, where the user has the possibility to select a specific authorization, and the privacy policy extracts a list of relevant phrases and presents them, together a proper describing of the permission from the user . This interface allows the user to quickly evaluate the private owned risks of android apps that are highlighting the relevant sections of the private policies that are owned by them and providing useful information on sensitive permissions, a particular application has authority on it.

In the prefecture of this project is to conjure an accurate but yet cost-effective, viable solutions used in detecting number plate region by implementing the system and technologies needed to process the image locally and convolution neural network (CNN). In the world of smart cities, the object detection algorithm has become a crucial element in it’s working. In urban surveillance applications, the image sensor/camera acts as a crucial aspect in the process of virtualizing the scene in which the scenario is monitored. A singular united deep convolutional neural network has been proposed that enables the ability to detect car/bike license plates to be obtained from a certain captured image and aids in recognizing the labels that have been captured. The whole structural integrity involves the need of any heuristic or tiring processes, such as the use of any provisional plate types or distinct character differences and also eliminates any intermediate procedures.

Ad hoc networks for cars (VANET) otherwise referred to as systems for wise transportation. In order to improve road well-being and maximize highway efficiency, VANET guarantees auspicious and exact match between vehicle to vehicle (V2V) and infrastructural vehicle (V2I). VANET is ineffective against malicious networks that can get into the structures and trigger real Medium Access Controls (MACs). It involves denial of service attacks, modification of knowledge, pantomime attacks, Sybil attacks, and replay attacks. Great security issues emerge with the proliferation of Blockchain. At the same time, a sample remains open for protection and discernibility. Steps were made to address the problems, thus revealing them to specific events. We are considering an echo statement in CreditCoin automobile declaration convention. This guarantees flexibility and security for use in the submission of comments. In CreditCoin, we design a stimulus portion centered on the Blockchain. When they gain or invest coins when driving powers, they track legendary focuses. In the meantime, CreditCoin does jam protection and keeps black. However, CreditCoin foreshadows a number of security threats and allows dependent defense in view of Blockchain, as trace manager approaches malignant hubs in an awful case.

With the precinct of e-mail message discharge events, for example, the Hillary Clinton’s electronic mail argument, certification besides safety of subtle e-mail info consume changed into customer’s chief concern. Mixed e-mail is unmistakably a sensible response for giving security; regardless, it will outright tie their exercise. Open encryption with catchphrase search plan is a standard headway to solidify safekeeping shield in addition inconceivable feasibility works organized that can recognize a huge development in inspecting mixed electronic message in a cloud server. Within the paper, we recommend a applied PEKS plan termed by means of open key multi-catchphrase available encryption through covered edifices (PMSEHS). That may perhaps attract e-mail recipients toward the multi-watchword also Boolean requesting from the monstrous mixed electronic post database as savvy would be reasonable, deprived of skimpy extra data to the cloud server. In same manner, we provide relative tests, whichever show that our strategy takes a sophisticated capability in multi-catchphrase break down used for mixed messages.

The idea of the Internet of things (IoT) has become a popular research subject in many fields in recent years, including business, trade and education. To build sustainable urban life, smart cities employ IoT-based infrastructure and applications. IoT enables smart cities to make community residents more aware, responsive and effective by using information and communication technologies. As the number of IoT-based smart city applications increased, the amount of data produced by this application increased tremendously. Effective steps are taken by governments and community stakeholders to process these data and forecast potential consequences to ensure sustainable growth the field of forecasts; recurrent neural network methods were used in big data for many predictive problems. This inspires us to make use of recurrent neural network to predict IoT results. A novel model of recurrent neural network is therefore proposed in this paper to analyze IoT smart city data. Through this research, we have gathered SO2 and NO2 gasses at various locations through Chennai.

The contemporary arrangement on cloud asset movement is regularly centered around considering the relationship among clients and cloud boss. Considering, the consistent improvement in the clients’ requesting and the rising of private cloud suppliers (CPs) attract the cloud supervisor to lease additional assets from the CPs in order to deal with their amassed assignments and pull in more clients. This likewise renders the investments between the cloud administrators and the CPs a basic issue to look at. In this paper, we analyze the two coordinated efforts through a two-organize auction segment. For the collaborations among clients and cloud supervisors, we embrace the alternatives based consecutive closeouts to plan the cloud asset portion worldview. When contrasted with existing works, our system can deal with clients with heterogeneous requests, give honesty as the overwhelming methodology, appreciate a straightforward victor assurance strategy, and block the postponed passage issue. We likewise give the presentation examination of the OBSAs, which is among the first recorded as a hard copy. Concerning facilitated endeavors equivalent markets for asset gathering. We get the radicalism of the CPs by their offered costs. We direct a wide evaluation of the two markets and perceive the commitment frameworks of the cloud boss.

Within the paper, we examine how to place mobile phones on behalf of permitting correspondences in a fiasco reconvert via connecting the holes amid different types of isolated system. We include structured as well as actualized android applications. In this application, we use this to convey to other colleagues simple to approach to impart utilizing distributed storage; in this module, client transfer the catastrophe area and other client will see the calamity area and get the specific area in map. It sees and gives cell phones the capacities of interchanges in misfortune recuperation such a assembling of mobile phones. This will be able to cooperatively stir up and about in addition to express disaster communication during a energy industrious method through their locale in addition to location information during sort in the direction of assist salvage activities. This application will likewise give the close by clinics and fire station and police headquarters. We group telephone numbers as a replica function so that it can take over the advanced android cells. Exploratory outcomes exhibit that group telephone can suitably assure association necessity besides extraordinarily support salvage activities during misfortune healing.

Organizations have known about the significance of Quality of Service (QoS) for seriousness for quite a while. It has been broadly perceived that work process frameworks are an appropriate answer for dealing with the QoS of procedures and work processes. The right administration of the QoS of work processes takes into consideration organizations to expand consumer loyalty, lessen interior expenses, and increment included worth administrations. The propose framework shows a novel strategy, made out of a few stages, portraying how organizations can apply information mining calculations to anticipate the QoS for their running work process occurrences. Our strategy has been approved utilizing experimentation by applying various information mining calculations to anticipate the QoS of work process. This framework proposes a conveyed work process mining approach that can rediscover ICN-based organized work process models through steadily amalgamating a progression of vertically or on a level plane divided transient work cases. The methodology has two fundamental calculations: One is a transient section disclosure calculation that can find a lot of worldly part models from the divided work process order occasion logs, and the other is a work process mining calculation that rediscovers an organized work process model from the found fleeting piece models.

One of the primary objectives of shrewd framework is to decrease power top burden and to adjust the hole between power market interest. Clients can participate in the activities of keen matrix, where the vitality cost can be decreased by vitality conservation also, load moving. Right now, estimating is a key pointer of clients exchanging load. For the most part, exact point cost anticipating is normal due to the necessity of economy and industry. As for clients, they are really anxious to know whether the power cost surpasses the particular client characterized edges, which they used to choose to turn the heap on or then again off. Under this situation, clients require the power value characterization. Subsequently, some particular limits dependent on point value estimating calculations are utilized to characterize the power cost. Capacity estimation methods are the essential of point value anticipating calculations, in which the essential procedure of value development is imitated by a value model. In addition, value grouping requires lower precision. Consequently, power cost arrangement turns into a key need in the value determining. In genuine world, the power costs are impacted by a number of components in which demand and supply are the two direct factors. Other than them, the power costs are impacted by physical attributes, for example, fuel value, power necessity, sustainable power source supply, and so forth also, it fluctuates hourly. Power cost anticipating is a huge piece of keen framework since it makes shrewd network cost productive. Since the power value changes often and a lot of keen meters screen the earth, for example, fuel age, wind age, and transmission, continuously, the measure of recorded information is very enormous.

The broad development of the radio frequency identification (RFID) in internet of things (IoT) application has provoked system interruption recognition, which turn into a basic part of intrusion detection system. Because of the open society of the IoT, the security of IoT frameworks and information is dependably in danger. The major objective of this research paper is to design an intrusion detection system framework using Anomaly-Based Detection technique. Optimization of interesting rules from a dense database is determined, using computation intelligent algorithm such as genetic algorithm (GA), genetic programming (GP), and swarm intelligence algorithm.

The mathematical modeling of conventional energy detector (CED) uses the squaring operation of signal amplitude to detect the random signals in Gaussian noise of cognitive radio (CR). In our approach, this operation is replaced by the arbitrary positive power operation. By choosing an optimum value for this operation with respect to the system settings, it has been observed that the non-linear modified energy detector (NMED) with better performances can be obtained. From the results, it is also observed that the improvement in the performance of current wireless systems. And we can conclude that the conventional energy system could not deliver the optimum performance in CR networks.

Internet of things (IoT) concept is related to a ubiquitous and pervasive connection of cyber-physical systems to the Internet. These cyber-physical systems can be identified and incorporate communication and sensing capabilities which turn them able to cooperate in a collective objective. IoT architectures have been successfully tested and valeted in several fields. Food safety is a relevant topic for public health and well-being. This paper presents a literature review on the application of IoT architectures for food monitoring in the past five years (2014–2019). The main contribution is to synthesize the existing body of knowledge, to identify common threads and gaps that would open up new challenging, relevant, and significant research directions. The review on real applications of IoT for food quality monitoring has conducted by analysed 13 studies which show that IoT implementation in this field is rare. The results state that most of the IoT implementations have been conducted in Asia, particularly by Indian authors. The most used sensors in these systems are temperature, humidity and gas sensors, and the most used communication technologies are ZigBee, Wi-Fi, radio-frequency identification (RFID), and Bluetooth low energy (BLE). Furthermore, the authors found exceptional potential in the implementation of IoT for food monitoring systems; however, some limitations are also found.

Optical communication is an indispensable technology for current society requirements. It satisfies the short distance and long-distance communication with the help of different mode analysis of variety type of fibers. In this paper, an analytical study has been presented that can help the next generation high-speed secure communication. The advantage is that it can act as the filter in terms of signal processing and channel in terms of communication. The waveguide can satisfy in both the way. For revolutionary communication and signal processing, it is found that optical waveguide research can satisfy future generation communication. The suitable design can be used for all optical signal processing. This article can provide insight to the researchers working in this direction.