Second International Conference on Computer Networks and Communication Technologies

IOT is an emerging area of network which acts as a future technology trend of sensing, computing, and communication. Internet of things integrates several techniques as wireless sensor network, Radio frequency identification and embedded devices with existing internet. IOT extends the concept of internet from network of homogeneous devices to network of heterogeneous devices. Air pollution is a major issue for providing healthy environment to the mankind. Healthy environment keeps mankind healthy. Hazardous gases detection becomes important. Hence a system for measuring and monitoring of air pollutants is to be provided. Wireless sensor network is found to be an efficient method of measuring and monitoring air pollutants.

Data Hiding is one of the frequently used security approaches for safeguarding the sensitive information of the available data as well as to transmit secret information among different ends in a vulnerable network. However, majority of data hiding scheme evolved till date is focused on its embedding capacity or else focused on introducing the distinct parameters of encryption. However, all these approaches will not only make the embedded file bulky but also, they will lose its imperceptibility characteristics. Therefore, the proposed paper introduces a simple and robust reversible data hiding process where a secret image is embedded within a video as a cover image. Motion prediction and histogram shifting approach is also utilized for obtaining highly secured bit-streams. The outcome of the study shows that the proposed system offers a better signal quality and retains maximum imperceptibility irrespective of the size of the secret image.

The central issue of any IoT device is its security in sharing the sensitive data. Different methods have been proposed for sharing of data from any IoT device. The ranges of security in these methods are different in various IoT architectures. This paper is a comparative study of these security schemes to determine which scheme allows the fastest and most accurate output. Our findings indicate that the attribute matching functions decreases the usage of keys and leads to an efficient key management technique. It helps in the easy addition and searching of the attributes and solves the trouble of complete re-initialization of attributes during updation. Attribute matching functions also reduces the need of large number of keys and is based on hashing of attributes towards a specified position that enhances the security. The authenticated people whose attributes matches with the specified condition can upload and retrieve medical files whereas non-matching attribute holders may be able to request and they cannot download the medical files nor have access to its contents. Since medical world is booming and the associated technology is advancing formerly and protection of the data from tampering and its sharing to various terminals need security enhancing methods and procedures which adds on the relevance of this work.

Casualties and huge losses could result from fire disasters in buildings. On the event of fires, evacuation from an isolated place of course becomes difficult and complicated due to factors such as fire spread, panic in evacuee movement and consequent congestion, failure in communication due to errors and delays, etc. This paper aims to compare competitive algorithms used for the calculation of shortest paths or the safest paths to mitigate the problem of evacuation. It also reviews corresponding evacuation models implemented in the past. By comparison with all existing methods, challenging issues are discussed, which should be met to enable basic requirements of an evacuation routing system. We conclude by underlining future directions towards enhancing the abilities for fire evacuation.

Nowadays most of the people are moving from rural to urban areas and they would rather prefer the advanced healthcare applications in their daily life. So in this century, the studies behind this idea are growing fast. In urban areas most of the homes are being equipped with smart devices and therefore the scope of the healthcare applications in this area can be done without any awkwardness. In this paper, the data collected from the smart devices can be used as the source data for this purpose. The human activities are monitored by the proposed system. From this huge data the patterns are being recognized by the use of Frequent Pattern mining and the appliance to appliance and appliance to time associations are built using incremental k-means clustering algorithm. The activity prediction will be done by frequent pattern based KNN algorithm. This system can predict the human activity pattern along with greater accuracy for the purpose of healthcare applications.

Healthcare analytics is a rapidly growing industry. Healthcare analytics have the potential to reduce cost of treatment, avoid preventable disease and improve the quality of life. This paper is an application of big data analytics in healthcare. A co-relation analysis on clinical big data from clinical reports and doctor’s notes are performed. Doctors consider similarity between health parameters to take better decisions. The co-relation analysis of health parameter is being used to cluster the patients based on similarity. Finally random model is designed to predict future health condition of most co-related patients based on the current health status. The future health prediction helps the monitoring of patients in diagnosis process. The system used modified future health prediction algorithm which is capable of predicting one or more diseases, which increase the possibilities of algorithm in health care. The performance evaluation gives about 97% of accuracy.

The term data science has been floating around as a popular terminology among social media applications globally. The associated device called IoT generates more than 2.5 quintillion bytes of statistics step by step, which could basically impact the business shapes. There is no doubt that the rising technology of IoE (Internet of Everything) is dependent on Data Science concept. The Industrial Internet of Things (IIoT) which makes up a good proportion of IoT tries to analyze the data they record and turn the data into meaningful information. In customary Data Science, the investigation is static and confined being used. The information that is got may not be refreshed so the outcomes accomplished in the wake of preparing may not be shrewd or usable. Then again, since IoT information is being got continuously, the investigation supplement the most recent market designs which permits making this investigation more significant and wise when contrasted with customary ones. Additionally, as more innovation layers are included or incorporated with IoT, it turns out to be harder to structure and process the huge numbers of approaching information. So truly, Data Scientists do need to up their aptitude with the end goal to grasp IoT-created information. As the engaging quality of IoT expands a flood of information lies later on. It is bound to the change the manner in which has seen Data Science for quite a while. The blast in information isn’t just going to require better foundation however more astute Data Scientists. Information Science for IoT can help overcome some wide-reaching difficulties in order to make more precise choices. This paper initiates to fulfill the readers to let identify the effective utilization of data science in IOT Platform in upcoming Era as IoT Opportunities for Data science as secured manner.

The invention of cloud computing offers elastic storage services in pay-per-use policies and hence organizations store their data in cloud servers. A cybercriminal can hack the sensitive documents in the cloud and sell it for mere profit or a cloud server may be curious to learn the contents of documents. Though many works for securely distributing the keys to the mobile cloud users have been proposed, they are vulnerable to attacks and incur high computational and communication overheads. This research work addresses the security issues pertaining to secure group communication during the file upload and download over public clouds. This research is a pioneering attempt to securely transfer the keys, upload and download the documents from the cloud. The security analysis ascertains the effectiveness of the protocol and the experimental results confirm the efficiency of the proposed protocol.

Object detection is one among the major sub-domains of the computer vision which deals with the identification of objects of a pre-defined class. Recognition of object is imminent to identify numerous pertinent objects from an image or video. Several deep neural learning, machine learning based techniques are used for object detection in digital images and videos. This paper discusses a comparative study of some deep learning based object detection frameworks, and analysed on the benchmark mean Average Precision (mAP) and selected models are evaluated using PASCAL VOC 2007 dataset which is the standard image data set for object class identification and recognition. Among the selected detection models, PVANet has the highest mAP (84.9) with FPS 21.7 and is considered as the best object detection method.

In recent years, mobile ad-hoc networks have rooted their pillars for emergency communication owing to reasonable cost, diversity, and easiness of mobile devices. The mobile ad-hoc networks is a self-coordinated, distributed and infrastructure-less network of mobiles nodes. These characteristics of MANET enhanced the applicability of MANET in the field of emergency communication such as military and police operations, flood control and fire disaster management, etc. In MANET, a broadcast storm causes network problems as there are redundant broadcasts and packet collisions. Classical broadcast methods have motivated on evading broadcast storms by preventing some rebroadcasts. The further problem is the link breakages induced by node instability and their power exhaustion. In this research, we propose an adaptive neighbor knowledge-based hybrid broadcasting method to address these network problems. This method refines the counter threshold based on neighbourhood, mobility and energy of the node and makes use of the refined thresholds to make the broadcasting decision. The proposed method perform best as compared to AMECBB and TCBB by decreasing delay, packet dropping, and routing overhead and energy consumption.

Nowadays, due to the advancement in engineering and technology, the number of vehicles has been increased drastically. So, there is a need for proper management of traffic in order to maintain the smooth functioning of the cities and nation as a whole. Though various techniques are evolved for traffic object detection, it has been used for managing the traffic. The project focuses on developing an efficient algorithm for controlling the traffic signal lights. It uses a hierarchical decision-making model, providing local decisions based on statistics, and global decisions based on pattern learnt at a higher level. Situations like emergency arrival and accidents would be handled by the global nodes’ network. The decision taken would be communicated to the self-automated cars for their future decision.

As the education becomes more competitive in nature, students need to learn many other subjects or courses by interest or need. We identified the multidisciplinary knowledge of the students and created interested groups based on the grades or points obtained through questionnaire. But this learning happens through online, in order to maintain security of the student information, we have proposed a blockchain model SIBC(Student Interest Blockchain). This model is useful to identify the student interests and creates a platform to fulfill the student interest along with his regular course. It was proposed to maintain separate blocks for each specialization area, and the student can join any one of the block based on the points got in the questionnaire. Each block consist of students, faculty experts, industry experts, alumni of that institution and all other resources available are helping the students to enhance his knowledge in that particular area.

This paper has illustrated the comparison of various modulation formats with and without FBG. It is seen that internet service providers are looking for new ways to satisfy customer’s requirements related to quality, bandwidth, speed, and security. Optical access network proves to be the crucial broadband network to satisfy the demands such as video communication, cloud computing, data mining. The most common optical access network is the PON and the most promising is GPON (Gigabit Passive Optical Network). In this paper, a general study on GPON is done with FBG and without FBG. For the distance of 100 km and bitrate 2.5 Gbps analysis of GPON has been done. Role of FBG as dispersion compensator and multiplexer has also been analyzed. The outcome shows that NRZ with uniform FBG shows better results.

Weather monitoring is a systematic method that gives us detailed information about the weather of our surrounding environment. This device basically monitors the different weather parameters such as temperature, humidity, and rain. The heart of the device is Node MCU (12E). In this paper, three sensors are used namely temperature and humidity sensor (DHT11) and rain sensors have been used, which is connected to Node MCU. When the code is uploaded on Arduino IDE, the board is connected and the sensors start working. All monitored data are stored in thingspeak platform. We use another approach to calculate rain value by the use of IoT & its platform and fuzzy logic to set the weather condition easily for their users. Finally, we compare and analyze both rain value by the use of a different method and the performance of rain value is analyzed on the basis of a different method.

India is an agricultural country with over 58% of income is earned through agro-based environment. Out of all, the major crop produced in India is Black pepper. The production of black pepper has played a very important role in enhancing the economic growth of our country. To help in enhancing the production and the export of black pepper from India, this study proposes a method to detect the ripening stages of peppercorns. An android application is developed to predict the maturity and ripeness of the peppercorns. This will help the cultivators to produce peppercorns as per the global market requirement.The pepper images, representing various stages of maturity are collected from various agricultural resources. These images are trained and classified by extracting, the color features like RGB value. The training and classification is done using Support Vector Machine Algorithm. SVM classifies the images into three classes. The images in these classes are used as the dataset to further identify the maturity or ripen stage of the peppercorns in the input image captured by the android application.

In recent times, Heterogeneous Wireless Sensor Networks (HWSNs) are gaining increasing interest from various research communities due to their potential applications in large number of domains such as environmental applications, area surveillance and health monitoring. However, the performance of a HWSN primarily depends on a set of critical decision parameters including robustness, security, connectivity and efficient usage of energy and so on. Specifically, one of the most important research challenges is, enhancing the network lifetime of HWSNs, so that the object monitoring activities can be conducted without any interruptions. More specifically, Energy usage in a network is an essential factor, which directly influences the lifetime of the given network. In contrast, efficient energy utilization within a network can be achieved by means of grouping the nodes into clusters using effective clustering algorithms. In this paper, an overview of HWSNs is presented along with the different state-of-the-art techniques used for implementing energy efficient clustering strategies. Furthermore, this article also explores the future key directions and highlights the research challenges that need to be addressed in achieving energy efficient clustering in HWSNs.

The contribution of Wireless Sensor Network (WSN) towards commercial sensing application is tremendously progressing day-by-day. However, it is still shrouded by security problems owing to less practical applicability of existing research solutions as well as inherent nature of resource constrained nodes. Key management in encryption technique is one of the most frequently exercised techniques; however, it lacks the robustness against various attacks due to flawed in design principle of dependable parameters in construction security solution. Therefore, the proposed system introduced a novel temporal key-based mechanism that is meant for strengthening the encryption operation. The study also offers higher flexibility to include different encryption algorithm for scaling the security feature based on different vulnerability over the challenging environment of WSN.

Now-a-days the networks which are wireless are characterized through a constant range policy assignment. A big part of the allocated range is used occasionally and based on the physical area differences inside usage of allocated ranges degree is fifteen percent-eighty five percent with an excessive differences in time. A confined attainable range and incapability within the ranges utilization lead to a brand-new designed named as Next Generation Network (xG network) and also cognitive radio network.

Wikipedia is recognized as one of the largest repositories in the Web. The term knowledge base was in connection with the expert systems as it is the part of Artificial Intelligence. A knowledge base can be created for any entity. The existing system like YAGO, MediaWiki tries to convert Wikipedia into a structured database to provide a vast knowledge base across the domains. It is very difficult to get the information which we want across the domains. So, the solution would be to get a systematic automated approach to build a knowledge base using Wikipedia on entity which we are interested in. The proposed system provides a knowledge base built upon the location as its entity. The system is feeded with seed data, by using these seed data it traverse through the Wikipedia graph and builds knowledge base using similarity measurement between seed data and traversed upcoming pages of wiki graph. Any expert AI systems uses gold standard knowledge base to take any decisions.

Internet of Things (IoT) is expanding exponentially due to the manufacturing of cheaper electronics and high-speed communication. Blockchain technology has paved way for the introduction of digital cryptocurrency called bitcoin. The blockchain is one of the four main technologies that have enabled cryptocurrency. Conceptually blockchain is a distributed digital ledger that stores data in blocks, which are cryptographically linked with the previous blocks. This makes blockchain immutable. Blockchain has gained its popularity in the scientific community with the success of bitcoin and ever since researchers are trying to adopt this technology to other areas in order to solve problems beyond the smart sectors like Internet of Things (IoT), supply chain and so on. In our research work, we adopt blockchain in IoT and implement ethereum based IoT architecture for securing the IoT network. We have implemented a two-node ethereum network where transactions are made at the expense of ether and each transaction is verified by the miner using the smart contract written in the ethereum network.

Globally, there is an upward trend in the use of two-wheelers, both for transport and recreational purposes. This has led to an increase in the demand for a safety device that can act more than a cushioned wall between the wearer and the point of impact. The proposed solution is an Advanced Helmet with features like a Head-Up Display which would provide Distraction free Navigation and a Fall Detection system that could potentially reduce the number of deaths in road accidents by allowing faster Emergency Response through Accurate Location information being provided within minutes of the incident.

In the world various methods are being adopted to create a systematic identification for their citizens. In year 2009, UIDAI a government body of India initiated a 12-digit number called Aadhar number, which is generated out of biometric and demographic fusion of an individual and their identity. In order to ensure a highest level of data protection as well as privacy preservation of the Aadhar card propagation through a network, it requires an efficient model of security that becomes synchronous with the cloud infrastructure. This paper initially investigates the existing approaches and their limitations towards the Aadhar card security proposes a security model namely ECrypto-AaDhaar based on the cryptography approach synchronous to the cloud architecture. The privacy preservation of the AAdhar card associative demographic and biometric information is performed considering a statistical crypto-blocking operation prior propagating it through the network in the context of cloud infrastructure. The study later also presented an experimental analysis to demonstrate the performance of ECrypto-AaDhaar technique from a time complexity perspective.

Social media is an emerging platform where people share their opinion about any national or global issue. This opinion attract marketing and research team to analysis people sentiment over social media for their marketing of their product. The reviews of end user over social media play an important role to show the quality of products. This paper discus the analysis and the evaluation of the sentiment based on the customer review s that are available in the online for presenting the sentiment based review for the products by using point wise mutual information of opinion word. The Results shows that using modified approach gives improved efficiency on feature-based sentiment analysis.

Today’s Digital World is evolving rapidly and the smartphone usage has became mandatory. People use smartphones to get services like email, education, business, social communication, etc. For each service category, there are plenty of applications (Apps) available in the market. Along with Apps usage, it is obvious that a user need knowledge on securing devices and private data. We analyzed Naive Android Smartphone Users (ASUs) on different Security Areas where awareness is in need to secure the device as well as the data. The responses indicates that majority of the participants have a fear of Malicious Attacks on their private data, business information and financial transactions.

Sensing devices like camera, satellite, earthquake monitoring, video etc., are producing large number of data. Big data techniques paves the way for the handling the more number of data along with streaming data. Cloud computing technology make it easy to store, access and manage the data with low cost. The data compression techniques helps to minimize the data in the cloud and store the data effectively. The aims of the study is to provide a systematic review of the data compression on big sensing processing. The image compression is used to minimize the size effectively and useful for the cloud environment. The deduplication technique is another method is used to compress the data in the cloud and helps in minimize the size. The clustering based compression technique process the cluster for similar data. The three kinds of compression technique in the cloud are investigated in this study. The investigation of this methods shows that the compression technique is still need to be increased in the manner of scalability and flexibility.

Today’s healthcare system deals with a huge data that needs a larger storage space and proper updating phenomenon. It is so obvious that big data is a buzz word in the field of healthcare. Not only specific to engineering field it ranges from institutional to organizational domain. Also its techniques not only helpful for doing research also in storing, manipulating with the observational or stored data. Many online based companies adopt big data analytical techniques for developing assets or medical equipment, so that respective domain can be benefited while providing treatment to patients. There are such companies’ deals with genomic data and helps in finding the specific reason to particular disease. Applying the big data techniques in medical data helps in finding best result and provide proper treatment to patients. Now it is possible to avoid such difficulties found in past time on giving treatment to patients through traditional methods.

Accidents are undesirable, incidental and unexpected events that can be prevented if they are recognized and acted upon, prior to their occurrence. Driver drowsiness is one of the major causes of road accidents. A solution to this problem is inclusion of a drowsiness warning system in vehicles to warn the driver of drowsiness. The main objective of this paper is to develop a non-intrusive real-time drowsiness warning system by monitoring eye state of the driver. The proposed system uses eye-aspect-ratio and is implemented in Raspberry Pi. This system is optimized using three techniques, by re-sizing the frame to be processed, using a single eye for drowsy detection and by not processing every frame. The selection of frames is based on analyzing the redundancy nature of video frames. This system is designed by considering several factors to minimize frame processing latency to match with Raspberry Pi’s processing speed.

A future-oriented, blockchain real estate market could help with making new plans of action of allying potential buyers and sellers. It is difficult to assume how properties of numerous sorts could be liquified, tokenized and exchanged. This blockchain paper will considerately influence transaction times as property sellers find entrusted buyers and the other way around, instead of hunting down a certifiable buyer and seller. It is more than likely that the entire real estate industry will gain advantage by this move, by bringing down the obstacle of real estate speculation and likewise eradication of center men. New meanings of property ownership and rental agreement will emerge from this move in the real estate business. The ownership of property need to be clearly stated and historically defined by documentation. There are many conflicts in ownership which is challenged in judicial/legal systems. The provision of maintaining the clarity of ownership is the responsibility of identified advocates and provided as Certificate of Title. There is no system taking care systematically of the overall transactions. The proposed will provide the secured transaction maintenance system for land/property transactions to support the timely verification of titles of said property. Also the implementation would be based on Blockchain Technology making the system transfer proof in storage of transactions.

Various organizations use plain text to store data related to day-to-day computations. Data is stored in the form of plain text documents without any structure and specifications. Retrieval and searching from structured data will be easier with various existing database systems. Querying and searching on unstructured content is difficult. In general, searching on unstructured content can be implemented using a similarity between input keywords and documents. Organizations are moving towards cloud to store the data because of high availability, lower maintenance cost, reliability, and portability. In a cloud system, sensitive data like personal records are to be protected to avoid malicious access from intruders. But searching data from encrypted content is difficult. In this paper, we are proposing a scheme called Multi Keyword Search on Encrypted text (MKSE) which enables searching on encrypted unstructured text without decryption in the cloud using Cosine Similarity. To store the documents, we are using CryptDB database where documents are stored in encrypted form. Thus the multi-keyword search is done on encrypted data in the cloud using cryptDB for providing data security.

Traffic congestion has become a major problem in the world wide. So we need efficient system which monitors the traffic and updates the time setting in traffic signal. The cameras installed in the road junction will be used to capture the real time traffic and these images will be processed to count the number of vehicles in each lane. MATLAB Platform is used where it develops the various object detection algorithms for the combination of many image processing algorithms. The real time object detection and tracking will be generated by control signals where Arduino programming will provide an interfacing hardware prototype. The centroid value will be calculated in each lane. Based on the centroid values obtained from the system, the signals will be sent for the traffic pole as the output.

Data Mining Techniques easily handle and solve the problem of handling the massive amount of data due to heterogeneous data, missing data, inconsistent data. HealthCare is one of the most important applications of Big Data. Diagnosis of diseases like cancer at an early stage is also very crucial. This paper focuses on the prediction model analysis for the breast cancer diagnosis either benign or malignant at an early stage as it increases the chances for successful treatment So predicting breast cancer at benign increases the survival rate of women. Data mining classification algorithm like SVM, Naive Bayes, k-NN, Decision Tree compares a variety of statistical techniques like accuracy, sensitivity, specification, positive prediction value, negative predictive value, area under curve and plotted ROC curve in R analytical tool which is promising independent tool for handling huge datasets is proven better in a prediction of the breast cancer diagnosis.

The design and analysis of Coplanar waveguide fed antenna for ultra-wideband application is proposed in this paper. The designed antenna has a compact size of 28 mm × 32 mm and provides impedance matching over a bandwidth of 3.07–11.58 GHz. The characteristic parameters such as return loss, voltage standing wave ratio, current distribution and gain are simulated and these parameters compliance with existing UWB standards. The results are simulated and analyzed using Ansys HFSS Electromagnetic Solver Version-16.2.

Text summarization is the process of shortening a text document in order to create a summary covering important points, aspects of the original document. Text summarization methods are based on extractive model and abstractive model. Two-level text summarization is used to form summary of different news articles. In the first level, multiple news articles are read and first level summary is generated. These multiple summaries are then analyzed and a single summary concerning the news topic is generated in second-level. TextRank with TF-IDF algorithm is used which is an extractive summarization technique to create news summary. The performance of the summary is evaluated using ROUGE matrix.

Authentication is a critical step in initializing a bank transaction. Core banking and net banking systems use ID cards, passwords/PIN, OTP, etc. for authentication. However, these methods are still prone to forgery and hacking, due to which unauthorized person could gain access to user’s account. We propose a methodology which uses an image share as authentication key. This has been created using visual cryptography and other image processing techniques to encrypt authentication parameters. It is encrypted such that the share becomes incomprehensible to human eyes and non-decryptable by hacker, thus protecting the data from forgery or hacking. This paper proposes a double authentication system where user is identified by decrypted image and PSNR value. Hence this system provides a very high level of security.

By integrating various unmanned vehicles with ground station on a cloud infrastructure for data streaming, monitoring and visualization processes in real-time has been considered as a challenging task. In a traditional approach, data collection method for unmanned sensors is not flexible and it has various limitations when it is dealt with the humongous volume of data. To address this, we propose a cloud data framework for sensing, securing and maintaining the sensors from cloud for surveillance and take quick decision based on the input. The another main intention of this framework is to enhance the system throughput, reduce the man power and cost by leveraging unmanned sensors and all integrated sensors. In order to meet the above, we create a platform to support in all levels, such as infrastructure, network and software to simplify the processes and services. In addition, all integrated sensors tend to have multiple roles, all should be monitored, data should be captured and captured data will be analyzed and visualized.

In today’s word, high-speed data streams are continuously generated via a variety of sources like social media and organizational business related data. We have listed the basic characteristics of big data and challenges in handling big data and data streams. This paper shows present work on processing and analyzing big data and data streams, real-time data analytics, decision making, and business intelligence. Our aim is to research different trends in distributed data analysis, a study on security of big data, applications of big data and processing of data streams. Even though there is vast research happening in the field of big data across the globe, still there is a scope of improvement in this field.

Sensor cloud is the integration of cloud computing and wireless sensor networks (WSN). This integration is beneficial for both WSN providers and cloud service providers. WSN data can be kept over cloud which can be efficiently used by several applications and cloud provider can provide sensor- as a- service through this integration. In sensor cloud, there are multiple physical sensor networks which are mapped with the virtual sensor networks using a cloud to provide services to the users. End users are facilitated to execute multiple applications of WSNs through such VSNs. Some applications can also demand more than one WSN at a time. Virtualization supports to achieve this goal in sensor cloud. Although sensor cloud has several advantages, still it has many research issues like energy efficiency, security, data transmission, QoS, etc. This paper presents a survey on energy efficient data transmission in sensor cloud that discusses and compares the existing techniques of this field.

In real world, visually challenged people face the great challenge of understanding nearby objects and movements going on in their vicinity. They mainly depend upon their auditory or physical abilities of touch to recognize things that are happening around them. Being able to describe the surrounding environment and objects that are present around them using properly formed sentences could only be done if a normal person is present and can describe it to them. We plan on creating an application that can solve this very challenging task by generating description of a real-time video captured from a mobile phone camera, which will aid the visually challenged in their day to day activities. In this paper, we are using concepts of Object detection and Caption generation and present our approach for the same, this which will enable us to run the model on smart phone devices in real time. The description pertaining to the objects, as seen in real time video generated will be converted to audio as the output. We train our proposed model on various datasets so that the generated descriptions are correct and up to the mark. Using the combinations of Convolutional Neural-Network and Recurrent Neural-Network and our own modifications, we tend to create a new model. We also are implementing an Android application for the visually challenged people to show the real-life applicability and usefulness of the Neural Network.

“The next Industrial Revolution”, Internet of Things (IoT) is the technology giant which is rapidly gaining ground in the scenario of modern Wireless Sensor Networks (WSNs) where physical objects (or “Things”) are connected through the common platform i.e. Internet. The Internet of Things paradigm is paving the way towards success, providing services in different domains. Single architecture is not enough to provide essential services in all domains. The Reference model can be made to provide a starting point for developers looking further for developing strong IoT solutions. In order to facilitate future research and to help the product builder to choose from the different architectures, This paper does an analysis of Layer-Specific, Do-main-Specific and Industry defined IoT architectures. Intel, IBM, and CISCO have also made a push towards developing new models that aims to standardize the complex and fragmented IoT Industry by releasing IoT platform reference architectures. This paper contributes the IoT, basic elements of IoT, systematic classification of different IoT architectures and comparison of industry-defined architectures.

E-mail is one of the leading and most reliable mode of communications even after the emergence of many new methods of electronic communication systems. E-mail messages are transmitted from the senders’ system to the recipients’ system over the Internet with the help of some intermediary nodes including servers, switches, gateways, etc. E-mail communication relies on some of the oldest communication protocols which have not been modified ever enough since. Hence e-mail communication is vulnerable to some security risks including eavesdropping, spoofing, tampering, phishing, etc. The e-mail communication system makes use of some security protocols, techniques and data encryption methods in many ways to make this communication secure and reliable. This paper reviews the security of e-mail protocols and data encryption techniques that are in use in the e-mail system. It reports the results of a study of the roles and advantages of e-mail security techniques, protocols, and algorithms. This paper also highlights security vulnerabilities in an e-mail communication system and the possibility for improvement in the control measures.

Visual Secret Sharing (VSS) schemes address the security concerns of digital images. Rotational VSS schemes are unique in that the shares stacked with the specific rotated shares reveal the secret images. However, these schemes iteratively process huge pixels in sequence. We propose a GPGPU based rotational (2, 3) VSS scheme which leverages concurrency inherent in the traditional rotational schemes. The proposed scheme also makes use of contemporary hardware resources efficiently. The reduction in the memory band-width demand and increased data transfer speed further accelerates the speedup of this scheme. A significant improvement in performance with an efficient resource utilization is observed.

Support Vector Machine (SVM) is a learning model which can be used as data analyzer for classification by its associated algorithms. SVM classifies the data by finding the hyper-plane that maximizes the gap between two classes. The structure of decision tree consists of root, branches and leaf nodes and the tests performed on an attribute and leaf nodes were represented by internal nodes which denote the result of the test. In this paper, a model to classify data using an ensemble of decision tree and support vector machine is proposed on a dataset collected on the topic of ‘Engineering Degree’. Combining the decision tree and support vector machine can be an effective method for classifying the data as it reduces the testing and training time of the data collected. The analysis of the result has been performed on a sample set of data taken from a large collection stored in the cloud using Hadoop.

The Controller Area Network is a simple, well designed, highly efficient and reliable in-vehicle bus standard widely used since its development in 1983. Controller Area Network is a serial communication protocol that supports distributed real-time control in Automotive Electronics. CAN is a thrust area for the past two decades in multi-disciplinary research, encompassing various tools and concepts for solving real-time problems. In this paper steering based headlight control system using CAN bus is implemented. The model consists of a steering control unit which determines the direction of the headlight according to the changes in the steering position. The Headlight control unit is varied from −16° to +16°. The proposed model is efficient in comparison with the conventional gaze controlled headlight system in terms of reduced transmission time, the speed of transmission and provides safety to the drivers during night time.

Wireless Sensor Networks (WSN) are rapidly developing technological platform with tremendous applications in several domains. Body Sensor Network (BSN) plays a major role in the fields of social welfare, medical treatment and sports. The major problems identified in health care based sensor networks are energy consumption and lifetime. These two factors directly depend on clustering and routing and hence this research resolves in enhanced uniform clustering of the sensor nodes. Initially, routing with better load balancing is determined by tree clustering technique. The K-Nearest Neighbor (K-NN) algorithm is used to enhance the uniform clustering of sensor nodes. Secondly, the biometric iris fusion based trusted anonymous secured routing protocol is proposed to preserve two factors namely anonymity and unlinkabilty of the wireless body area network. Finally, secure routing technique based on retina with DNA coding is made with the help of onion routing to avoid computational overheads.

Now-a-days, more services depend on Information Technology (IT) systems. Some of these services such as health care service and financial service are very crucial to the customers. Even a very small amount of data loss or a short downtime could lead to huge economic crisis or social issues. So most of the important business and public services use disaster recovery technique for protecting important data and reduce the downtime caused by catastrophic system errors. In this paper, we proposed an efficient replica management based disaster recover using Elephant Herding Optimization Algorithm (EHO).in this paper, initially, the data are uploaded into could with the help of EHO algorithm. Then, to avoid data loss, we create the replicas for each data. Finally, the request based data are back up and retrieved.

Utility mining is a prospering trend in data mining area. High utility itemsets are those ones that yield high profit when sold together or alone, in other words groups of items that generate high returns/profits in the customer transaction database. Regard of finding high profitable products, this is necessary to understand the recent preference of the customer. Detecting of HUP in a single phase with customer purchase behavior for a time period is addressed in this paper. In this paper not only consider the individual profit and quantity of each item in a transaction but also consider the customer preference of an item for a given time period. We propose an algorithm and a list structure for finding high utility patterns over large volume of data on the basis of sliding window technique and time cube. The objective is to discover groups of items periodically purchased by customers and generate large profits to the sales revenue.

India is an agrarian economy, with three-quarters of its rural population relying on agriculture as their primary means of livelihood. Agriculture shares 17% of the country’s GDP. With rising demands due to the ever-growing population, increasing farm productivity is the need of the hour. However, annually more than one-third of the crop yield is affected by the diseases in India. Thus, identification of the disease in the early stage is essential to provide proper treatment. Traditionally, the disease identification is done by visual examination, which often is done after major damage has already been done to the crop. With the help of state of art technologies like deep learning and cloud computing, the same can be achieved on a real-time basis. With the help of Convolutional Neural Network architectures, researchers propose a system that focuses on detection and identification of the plant disease with a mere click of leaf picture and provides solutions. Furthermore, the system also generates heat maps, which provides insights about disease spread in a region, thereby easing out the data analysis process.

In the recent past, acoustic detection and ranging techniques has become more advanced and sophisticated. In general, the sonar systems consist of distributed data acquisition systems. For accurate ranging these systems needs to be synchronized in time. This can be achieved by custom synchronization techniques or through standard mechanisms like IEEE 1588 PTP (precision time protocol). This project outlines the implementation of a synchronization mechanism with minimal hardware support. The main objective is that it will minimize custom hardware requirements. Using a standard hardware interface like Ethernet will rule out any signal integrity related issues at the physical layer. The accuracy of synchronization across the systems has to be measured and tuned. The implementation will be done on the Microzed board using the Vivado 2018.1.

DDoS (Distributed Denali of Service) attack majorly issues on vehicular Ad hoc network (VANET). When two or more vehicle exchanges information to each other either direct or indirect, DDoS attack effects on VANET networks. VANET networks can manage communication by using trust management model for predication and detection of DDoS attacks. Trust model based on Hash encryption key and Identity based technique will work ageist DDoS attacks node. Trust Management System is divided in to three Steps. Step-I: System Initialization, Step-II Trust-Value Updating and Step-III: Vehicle to Vehicle Authentication and Trust Evaluation. Its method applies scenario 20 nodes, 40 nodes, 60 nodes, 80 nodes, and 100 nodes and totally taken 15% malicious nodes on a given scenario. With the help of NS2 simulation tool through which calculate PDR, E2E and Throughput parameters during Simulation. Analysed result based on three protocol AODV, DDoSAODV and TAODV that’s by TODV perform better result compare to another protocols on the bases of parameters.

In data analysis, clustering is one of the important tasks. In this context many clustering methods are proposed in literature for big data analysis. Density based clustering (DENCLUE) is one of the powerful unsupervised clustering methods for the huge volume of data sets. In denclue, hill climbing plays important role to find the density attractor. In this paper, we apply Differential evolutionary algorithm in the place of hill climbing to find the global optimum solution. In this model, we propose Gaussian based mutation function in DE to improve the accuracy and execution time on spark platform. We test this approach on big data sets presented in literature. Experimental results shows that the proposed approach outperforms other variants in terms of execution time.

Data mining is the process of identifying patterns in data sets by applying appropriate methods with cluster of machine learning techniques. In recent decades, high utility itemset (HUI) mining has become the emerging research area, which focuses on frequency and also on utility related with the itemsets. Each itemset has a value like profit or user’s interest, called as the utility of that itemset. HUIs are present in a customer transaction database which yield a high profit. The target of HUI is to discover the itemsets that have utility value higher than the threshold value. The issues faced in HUIs are dealing with negative item values and number of database scans, mining in XML database, candidate sets and distributed computing network. This paper presents a survey of various algorithms and their restrictions in mining HUIs and the performance analysis of the surveyed algorithms.

The Internet of Things helps in filling the space between the real world and the virtual world, which is one of the forthcoming networking regions. Our daily lives have been changed by the role of the mobile appliance. Devices can interact among themselves where protocol played a key role in IoT implementation. The simple protocol of IoT technology is designated here. We can classify the IoT protocols into four categories such as data link layer protocol, network layer protocol, transport layer protocol, application layer protocol. Depends on the disparate benchmark, comparison of data link protocols is done. After comparing all protocols, we can’t determine which integration of protocol is the best for the implementation of IoT.

The deaf & dumb (or the Mute community) find it a tedious task to converse with ordinary people through sign language. This stands as a hindrance in even the most basic of their activities. It affects their personal development, interpersonal relations and limits the contributions they could otherwise make to society. The prime motive of this project is to provide an easy to use platform for the hard of hearing people to express themselves despite the sign language barrier. We aim to achieve this motive through gesture recognition. Using gesture recognition, we compute the mathematical interpretation of human hand gestures to recognize the signs conveyed by American Sign Language. The system enables real-time hand gesture and speech recognition and provides an innovative and simpler mode of communication for the mute people.

A very efficient and well-known computing environment is Mobile Adhoc Network (MANET). However, the node capacity and delay are affected during the data transmission due to its mobility. To tackle this problem, Multi-criteria Enhanced Optimal Capacity-Delay Tradeoff with Efficient Fuzzy-based Multi-constraint Multicast Routing (MEOCDT-EFM2R) method has been proposed to solve the optimal route selection problem where the multiple network metrics are taken into consideration. However, it requires fault tolerance-based Cluster Head (CH) selection since it does not handle any transmission or link failure in the network. Therefore in this paper, Efficient Fuzzy-based Multi-constraint Multicast Fault Tolerance Routing (MEOCDT-EFM2FTR) method is proposed to efficiently handle the transmission failure in the network. In this method, the CH for each cluster is chosen depending upon on the maximum nodal weight which is stored in the routing table. Initially, each and every node computes its weight with different network metrics like residual energy, mobility speed, computational power and fault tolerance. Then, the node with the highest weight is elected as CH of a specific cluster and each CH coordinates itself as a ring to tolerate any transmission failure by choosing a new CH. So, the new CH takes the responsibility of initial CH to complete the transmission without degrading the network performance. Finally, the simulation gives the evidence for the performance effectiveness of the MEOCDT-EFM2FTR method compared to the MEOCDT-EFM2R for throughput, routing overhead, delay, etc.

Filtered orthogonal frequency division multiplexing is one of the most promising candidate waveforms for 5G and beyond technology for the upcoming wireless communication system. Filtered OFDM ensures all the advantages of the orthogonal frequency division multiplexing technique with additional advantages in terms of spectrum efficiency and robustness in high SNR system and also in some specific cases, ideal spectrum utilization can be achieved by F-OFDM which can completely eliminate the guard band. This paper aims at comparing the performance of various filters by different windowing techniques such as rectangular, triangular, bartlett and hanning window to complex window function such as Kaiser, Nuttall’s Blackman-Harris, root raised cosine window which can be applied to Filtered OFDM in terms of spectrum confinement, BER improvement and overall throughput betterment. In this research work, multiple windowing techniques have been simulated with Filtered OFDM and their performance has been evaluated. Based on the simulation results, Filtered OFDM with complex window function such as Kaiser, Nuttall’s Blackman-Harris, root raised cosine can be seen as the potential candidate for future generation wireless communication system in terms of betterment in spectrum efficiency and improved out of band interference.

Internet of Things (IoT) is one of key areas of research, where number of smart devices is connected to internet. In IoT network Users operate various devices which process and exchange data with other devices and end users. User Verification is a major aspect in web based applications. One of the most well known evidence solutions used right now are the alphanumeric-based plans. Graphic critical solutions are offered to control the way users are ready to examine graphic styles pictures greater than reviewing alphanumeric secrets.Lexicon Attacks are focused on clients who select weak passwords. Graphical passwords offer enough security for such assaults. In this paper we suggest a two step authentication approach combining graphical passwords and Internet of Things.

Nowadays, due to the exploding growth of vehicles which paves a way for the unusual population in the road traffic, accidents are becoming more and more common among people. There are various factors which causes accidents. This paper lists out the classification of accidents and various solutions provided to avoid accidents as well as to overcome from accidents when it is encountered. Classification of accidents includes accidents by pedestrians, cyclists, mass casualty and animal accidents. The various solutions to avoid as well as to overcome accidents include usage of smart phones, VANET and GSM along with GPS. The proposed solution detects the accident and sends ambulance to the accident area by the usage of sensors. In this paper a derived factor named Crash rate analysis is used to determine the number of ambulances required for the accident area to take up the recovery process. This work is intended to provide better and faster lifesaving solutions for the road accidents.

Virtualization is a core technique used for running multiple applications and services at the same time. In addition to effectively increasing the utilization rate of bare metal hardware, virtual machines are also used for replicating test environments. However, the emergence of containers has been seen as an enhancement to virtualization. This paper discusses the concept of containers, how they function, and provides a theoretical analysis of containerization with reference to virtualization. In the end, the performance of container migration policies provided as part of CloudSim framework, namely First Fit, Least Full, Most Full and Random is evaluated using allocated number of hosts, containers, and virtual machines.

Smart grid being the major component of our power system is progressing fast to become more flexible and interactional. For the proper operation and planning in these systems, the load forecasting is of prime concern. In spite of the fact that many traditional methods are present for load forecasting but still a prediction of electrical load is needed to be explored further as the variation from time and surrounding weather conditions make it too complex. In this paper, we present deep learning based recurrent neural network approach for the short term load prediction using data captured from smart meters. This framework helps us to handle the over-fitting problem and uncertainty issues. The Tensor flow as a deep learning platform is used for our implementation. We compared our proposed model with the autoregressive integrated moving average model (ARIMA) and Fbprophet model and it is manifested that the proposed model outperforms in terms of RMSE, MSE, and MAE.

Twitter like microblogging site is used by millions of people to share their daily lives. During a natural disaster, the situational updates posted by users will get mixed with millions of other tweets and will be difficult to monitor manually in real time. Also, timely identification of situational updates, along with the location is very important for the rescue and relief operations during the disaster event. The tweets with contextual information posted during disaster provide information regarding the need or availability of resources and services, the number of casualties, infrastructures damage, and warnings or cautions. Some disaster-related tweet may not have any actionable information. This paper presents an alert generation framework, which will intake the tweets posted during the disaster, detects, classifies and geocodes the tweets belonging to each class, which provide actionable information, in order to alert the concerned authorities about the current situation in a timely manner.

The cheap availability of portable devices and processing elements has triggered the emergence of IOT. It is an integration of powerful computers with large number of tiny devices that support automation and smart decision making in all walks of life that may be agriculture, battle field, industry, office or home. Thus, IOT has very wide range of applications. However, the success of such applications heavily depends on efficient handling of numerous challenges related to computation and communication because huge volume of data needs to be aggregated, transported and processed at faster rate. We illustrate the potential of IOT and various technologies used to realize it and reap benefits. The paper also makes a mention of a futuristic application which is a wearable device that assists to protect victims from various threats.

The world of big data becomes a Business-critical component for Enterprise resource planning system and Business Intelligence. The ERP system runs big data longer and uses resource locks, which directly blocks the users from running queries on the database. Additionally, users will require updates on real-time data changes. More computational resources are required to reduce the loading cycle creating expensive processes with complete data loads. An ETL technique with CDC is used to resolve problems, through periodic updates of changed data. A process which identifies changed records to reduce the extract volume is knows as CDC. This paper proposes a structure capable of performing CDC by means of timestamps and replication tool designed for spontaneous synchronization between two databases. The overall performance of CDC technique to ERP system is compared. This approach is employed in a real-world project has noticed a transition to near real-time data ETL and performance improvement.

Social media has become an integral part of everyone’s life. Twitter is the most prevalent social networking service where millions of users share information astronomically everyday. Aggregation of these tweets provides a reflection of public sentiment which has a notable impact on the Stock Market. The objective of the proposed method is to predict the impact of Tweets and News Articles on Stock Market and provide insights to the investors to help them decide whether or not to invest in a company. The paper proposes a method to improve the efficiency of the existing methodologies by including news articles and adding specific weights to each tweet based on authenticity, followers of the twitterer and it’s retweet count for scrutinizing the public sentiment. It also considers sentiments expressed via emoticons and converts prolonged words into their normal form to increase the efficiency of sentiment analysis. Further, it uses N-gram representation for Feature Extraction and performs sentiment analysis on the collected tweets and news articles using Natural Language Processing and classifies the result into negative, positive or neutral using Naive Bayes classifier. The empirical results displayed in graphical format show that the proposed system can predict the fluctuations in the stock prices based on the sentiment analysis performed on previous day’s collected tweets and news articles.

Increasing volume of digital data in cloud storage demands more storage space and efficient technique to handle these data. Duplicate is unavoidable while handling huge volume of data. Data Deduplication is an efficient approach in cloud storage environment that utilizes different techniques to deal with duplicate data. Existing systems generate the hash value by using any kind of cryptographic hash algorithms such as MD5 or Secure hash algorithms to implement the De-duplication approach. These algorithms produce fixed length of 128 bit or 160 bit as output respectively in order to identify the presence of duplication. So, an additional memory space is used to store this hash value. In this paper, an efficient Distributed Storage Hash Algorithm (DSHA) has been proposed to lessen the memory space occupied by the hash value which is utilized to identify and discard redundant data in cloud. Experimental analysis shows that the proposed strategy, reduces memory utilization of hash value and improves data read/write performance.

Immense volume of data has been generated from unlike sources like health care, social media, business applications, manufacturing industries and many more. HealthCare plays a pivotal role in Big Data. Spotting and safeguarding of the diseases at a primitive stage are very much crucial. Heart disease specifically implies the condition of the heart that contracts or obstructs blood vessels which result in heart attack, chest pain or stroke. This paper emphasizes on the diagnosis of heart diseases at a primitive stage so that it will lead to a successful cure of the diseases. In this paper, diverse data mining classification method like Decision tree classification, Naive Bayes classification, Support Vector Machine classification, and k-NN classification are used for identification and precaution of the diseases at an early stage so that it can be curable and preventable.

Cloud computing is a very well known technology for all business people, software developers, end-users, and so on. Significant researches are going on to balance the cloud load. The migration of heavily loaded Virtual Machines (VMs) into lightly loaded Physical Machines (PMs) balances the Cloud load. In Resource Intensity Aware Load Balancing (RIAL) method, based on the weight of resources under utilization, it selected the VMs from heavily loaded PMs for migration and chosen the lightly loaded PMs as destination. An Improved RIAL was proposed to consider both lightly and heavily loaded PMs as destination. Later it was enhanced in the proposed Power Consumption Aware- Traffic Aware- IRIAL (PT-IRIAL) method with the consideration of power consumption, temperature and traffic measures to select the VMs for migration and select PMs for destination. From all these, in this current paper, the crossover and mutation process of GA is utilized to optimally select the migration VMs and choose the destination PMs. Thus this GA based load optimization algorithm optimally maps the migration VMs with the destination PMs efficiently.

An area effective, high speed and low power 8-bit Successive Approximation Register Analog to Digital converter (SAR-ADC) using 250 nm CMOS technology introduced in this paper. Among different types of ADC successive approximation is used as it is high speed architecture, suitable packed design and has good speed to power ratio. To minimize the power a Switched Inverter Quantization (SIQ) comparator and to reduce the leakage power the Multi Phase Clocking (MPC) based D-FF shift register for SAR logic are preferred for this work. The whole circuit design is bringing out with the help of Tanner EDA Tool.

CharityChain is a decentralized application, built on Ethereum that provides the user with a place in the virtual ocean, where the user can find a cause that he personally can associate with and help fund it in the fairest form using Blockchain technology. Blockchain technolgy is easy to use, lesser cost and fast access in all areas. NPO s enabe the chance for donations from global donars with the help of block chain technology.

Wireless Communication services have witnessed a dramatic growth in the past few years. With Telecommunication, there has been parallel growth in usage of different applications and services like VoIP, social network, e-commerce, entertainment etc. 5G is the next generation of telecommunication technology with elaborative and twirling services that can have a profound affect on the society. It can unleash new economic opening and societal benefits giving a potential for being a conversional force for the world. This paper elucidates the evolution of mobile wireless communication technologies. It includes details of architectures, salient features like data rate, primary services etc. To deliver next generation network services, 3GPP introduces IP Multimedia Subsystem (IMS) based on Universal Mobile Telecommunication System (UMTS). This paper also covers Software Defined Network (SDN) based architecture and explains various technologies like Network Function Virtualization (NFV), Network Slicing and certain pointers to research issues related to 5G.

This paper proposes the merging of different Techniques for the implementation of an artificially intelligent Chatbot with which the users can interact with text or voice commands. Main objective for the Chatbot to be used as an information and advising system that helps postgraduate students in answering their different queries and respond and provide the best possible options for them. It is created using Visual C# (GUI) and SIML (Synthetic Intelligence Markup Language) as a knowledge base for the Chatbot and uses Microsoft SAPI for both speech synthesizer, and speech recognition with natural language processing abilities as well as providing a few different extra help for the students as a course list, schedule and more.

Visual P300 mind-speller is a brain-computer interface that offers an easy and effective approach to track human brain responses. One major challenge in the design of this system is the unbalanced nature of its dataset, which can bias the classification process. In this work two distinct methods viz. nontarget undersampling and target oversampling were used to balance the mind-speller dataset. Since the choice of classification algorithm can impact the performance of mind-speller, the effect of dataset balancing was analyzed for a set of classifiers. The error rate, accuracy, true positive and false negative rates, true negative and false positive rates, positive predictive value, Matthews correlation coefficient, F-score, G-mean, and time consumption were the metrics used in this study. Among the various evaluated classifiers, k-nearest neighbor, support vector machine, and artificial neural network demonstrated significantly improved classification performance for the balanced (by target oversampling) mind-speller dataset while the Gaussian support vector machine yielded the highest metric scores.

The physical layer security is a new security paradigm based on the principles of information theory. Several methods for achieving secrecy in physical layer is proposed. This paper proposes PLS based on the deep learning architecture. In which the deep learning model will transform the channel coefficients, the beamforming based on this transformed channel coefficients can be decoded using deep learning architecture in the receiver. The secrecy rate and secrecy outage probability of proposed system is compared with the zero forcing based beamforming and superior performance is verified by the simulation using popular deep learning library TensorFlow.

This paper proposes a rehabilitation idea for partially paralyzed people due to neurological disorders or the ones who cannot move their arm due to an accident. Those people need aid for their routine activities which can be provided by a controllable robotic arm. This paper deals with the kinematics and dynamics of the robotic arm used for tracking and positioning the end effector within the workspace of the arm. An Arduino based low-cost robotic arm controlled by a joystick is developed. An alternate control strategy has also been developed to control the servo actuators in a wireless technique by an Android application via Bluetooth. The arm is designed to perform pick and place things which are within the designed payload. The simulation mechanism of the 4R robotic arm was designed using “RoboAnalyzer” software. The arm has been manufactured by a strong and light weighted PLA filament of 1.75 mm thickness. Parts of the arm and joints made of the servo motor casing were designed in Solidworks platform which was manufactured by FlashForge 3D printer. By this work, it is proposed to have low cost, a light weighted robotic arm which may be developed further for automatic control.

A landslide can be termed as the movement of a mass of rock, debris or earth down the slope, when the shear stress exceeds the shear strength of the material due to factors like accumulated rainfall, moisture or a wide range of ground movements. The local people who are living near the hilly areas are the ones that will be affected by this disaster as they have no time to react at all. It causes death, injury and property damage and adversely affects a variety of resources. Efforts to measure and to monitor potential landslides are required to ensure human safety and to protect civil infrastructure. Proper monitoring and awareness can help in saving lives and reduce property damage. The recent floods in Kerala, India caused numerous Landslides which lead to many deaths and major destruction in the mountain regions of the state. Kerala, India is very much vulnerable to landslide hazards. Existing methods include manual inspection using surveying equipment such as electronics distance measurement (EDM), levels, theodolites and total stations. The aim for our project is to develop a system using IoT to inform human individuals about potential landslides automatically, if any pre-failure slope deformations are identified. It uses a network of IOT that helps in updating the information about landslide on the Android App. Moisture Sensor, Vibration Sensor and Accelerometer are employed in the system that detects landslides as some critical values will be set for these sensors. If the value exceeds these critical values, then people will be notified the forthcoming landslide and huge loss can be prevented. So, this telemetry project helps to make the residents aware about the forthcoming disaster and can help to tackle the situation better. The fundamental concern of this project is to build up an early landslide warning system while using minimum resources without yielding on the efficiency of the unit.

In recent days, parents are discomfit about their school going children as kidnapping cases of students has been increasing from day-to-day. It is very important for every school to have an incorruptible and reliable transport service to make sure that the students are safeguarded. Nowadays students wait for a longer time for their school bus arrival standing near roads which is unsafe for them. Even though there are some of the technologies to ensure the safety of students, they were not able to provide abundant information to parents. Without their notice, they might exit at wrong stops. To reduce parent’s uneasiness about their school going children, this project presents the evolution of a school bus monitoring system which provides information about location, speed and bus arrival prediction time. It also identifies student’s location within the school campus. This is done through the technology – (IoT) Internet of Things. Tracking of school bus is done by GPS (Global Positioning System) technology. The system alerts parents by notifying them. This is achieved through GSM (Global System for Mobile applications). RFID (Radio Frequency Identification) identifies unique id given to each individual. The bus track system provides solution for tracking school bus on smart phones. Also an algorithm is implemented to calculate the arrival time of the bus in addition to tracking. The information can be accessed by parents through mobile application and they can continuously monitor their child.

Cloud storage is an efficient solution for storage needs of organizations and individuals. But the data outsourcing and data transfer using the same network makes the cloud data more sensitive for attacks. Hence providing security to the cloud data while in storage or transit requires a prime concern. The denial of access to a third party also must be ensured for the confidentiality of stored data. In this paper, Rubik’s cube based encryption with data spitting technique is introduced to securely storing and transferring cloud stored medical records. And the role of the third party auditor also limited to verifying the user only. The multiple number of encryption and denial of access to the third party ensure that the data is protected in the cloud.

Optimization of processes and functions on Software side or on Hardware side are constantly remaining under the research consideration. Optimization can decrease the average time taken by a functional element to complete a particular task. Space-Time Complexity of various algorithms has been determined. These algorithms are widely used in the real-time systems. One of the algorithms is Face detection Algorithm. This project focuses on finding the easiest way of implementing the algorithm so that it can work in real time. In this comparative study, the Viola Jones Algorithm for Face Detection is implemented in 4 forms – CPU, Multi-threaded CPU, OpenCV and GPU using CUDA may be on cloud. The Algorithm is tested over Face Detection Dataset by FDDB and the results are framed on a graph to get the comparison among the methods used. This research project also discusses the limitations, future scope and implementation of the algorithm in real-time video streaming in the most efficient way.

Cloud computing is a growing environment in the IT industry. Many of the users are interested to outsource their data in cloud. However, load balancing in cloud is still at risk. Resource allocation plays a major role in load balancing. In this scheduling problem, independent tasks in cloud computing can allocate resources by the use of fuzzy c means algorithm (FCM). To allocate tasks to their corresponding resources, particle swarm optimization algorithm (PSO) is used. This paper proposes a hybridization of the FCM and PSO algorithm which is called H-FCPSO algorithm. FCM uses Euclidean distances and PSO optimizes the cluster centers. FCM requires the number of clusters used in advance and thus PSO comes into action to find the number of best clusters. Hence, H-FCPSO identifies the number of clusters and enhances the load balancing. Since our proposed system selects resources based on parallel execution kit reduces the load imbalance in cloud. When compared to Genetic algorithm (GA), Ant Colony Optimization algorithm (ACO), PSO algorithm showed better results in terms of memory. Similarly, FCM was compared with k-means clustering algorithm, Hierarchial algorithm and it showed outputs with better accuracy. The proposed system evaluated data sets and proved to overcome the issues in load balancing and load scheduling which is proved by its precision in the outputs.

Unlike a traditional network architecture, Software Define Network (SDN) provides central control over the Network. When network visibility and network device programmability is concerned, standard documented open flow enabled physical or open vSwitch is a priority. Because of the central control over the network devices can be programed for different functionality of network application such as switches, routers, firewall and load balancer. One can program a network switch even though it is physical or virtual switch which supports standard southbound API interfacing to controller like open flow to act as hub, switch, firewall, etc. using those network application which can be interfaced with restful API to the controller platform choosen. In this paper, we study the behavior of different network functionality such as hub, switch, firewall and load balancer of open vSwitches over the SDN in linear topology. To create required virtual environment, we used mininet a network emulator tool for analysis and demonstration for network behavior of open vSwitches as hub, switch, firewall and load balancer role by running pox controller remotely outside the network of topology running machine and analyzed whether does it affect on throughput if we run our controller locally or remotely outside network boundary e.g. running on AWS cloud.

Computer performance improvement is one of the biggest challenges. The technology has moved towards increasing the performance by using Quantum computing which improves 20 times to decrypt the code compared to the classical computer. Quantum computing is computing which uses quantum mechanical phenomena. The main purpose of the quantum computing is to find algorithms which are considerably faster than the classical algorithms for solving the same problem. In this paper we are proposing the tools of quantum computing and different quantum algorithms. These ideas are first applied to classical computer and then to the quantum computer. We are also focusing on the architecture of quantum computing which is existing in literature.

The life of human beings is getting more dynamic in this era. So, every person needs to have a safer driving feature which supports their dynamic life. Self-driving or driverless car is getting popular day by day due to its low transportation cost which can result in better accessibility of those persons belonging to economically backward classes. A driverless car can drive in an autonomous way towards the destination safely with the help of GPS based road map which can be downloaded from a GPS server in collaboration with GPS satellites. But the problem with this present system is that if the number of the driverless car increases in a city that will lead to the congestion which results in overcrowding of GPS server. So the delay in the system will increase proportionally. One solution to this problem may be the cooperative communication through which vehicular communications can improve the wireless transmission capacity and reduces the server overloading due to its distributed features. In this paper, we have proposed a next-generation vehicular technology using an ad-hoc cooperative communication system consisting of vehicles and BTSs with small database consists of surrounding destinations.

ID card location innovation assumes a significant job in the present society to recognize the data database in different areas. Histogram of Oriented Gradients is the most broadly perceived part descriptors in picture planning and PC vision and is used for article ID. High-security ID check is an extremely intense activity. Present day object distinguishing proof and acknowledgment has progressively transformed into a critical issue in various on-going applications like security frameworks and government oversaw reserve funds systems. The ID card can be definitely perceived by utilizing an institutionalized scanner, for example, a standardized tag scanner. The specific customer honestly takes a gander at the given information on the ID card and after that thinks about it to the information effectively given to check whether it matches. This sort of ID is used precisely to strikingly separate customers faster and is dynamically powerful to give unmistakable access control courses of action. The trial results led show that this technique can be utilized to get the careful data of an individual and the required information.

Cognitive Radio (CR) is an adaptive, intelligent radio and network technology which automatically detects available channels in a wireless spectrum and changes transmission parameters enabling communications to run at the same time as and also develops radio operating behavior. It is viewed as a novel approach for improving the utilization of a precious natural resource: the radio electromagnetic spectrum. Cognitive radio network consist of primary network and secondary network. Primary users are licensed users and secondary users are unlicensed users. Whenever the channel band is found idle secondary users can occupy it without jeopardizing the communication of primary user. Priority is to serve primary user first on the channel. But there are lot of interferences caused by both the users. Solutions proposed are not able to deal with multiple CRN’s without centralized coordinator. This problem arises in coexistence heterogeneous networks. Several Markov chain models are proposed to deal with the solution which consist of decentralized MAC protocol to solve the unfairness issues. Fairness is one of the issues to solve and for this purpose probing is done. Simulation results shows how fairness is achieved and each user has got fair chance to serve on the channel.

As the beginning of twenty-first century, PC framework describes the improving Updation in the form of network efficiency, several hand holders & kind of operations that achieve on the system. As accompanied by latest generation comfortable machines for ex: Internet mobile, tabs, smart instruments i.e. updated machines & software also several calculating devices, no. connected hand holders progressing most & most. Therefore, safety on connection has been the key process which support complete hand holders. Intrusion detection has been the procedure in protecting intrusion. An intrusion detection technique may predict the complete upcoming & on- going intrusion at a structure. Intrusion detection technique may investigate complete priority under safety procedure with the help of managing the infrastructure movement. As Intrusion detection system (IDS) emerges as the obvious class under safety layout, it may manage capacity with support to determine the safety points in a framework. Several system supports under intrusion detection. Primary objective of this research work is representing i.e. With support to hybrid document opening approaches, it may minimize the duration difficulty in process as compared with mono approach. Particular structures were certified with support to kdd’99 document pair. An observational outcome remains significantly describing in nature i.e. hybrid approaches with support to k-means & Projective Adaptive Resonance Theory may uniquely minimize structure practicing duration of the framework & balances the perfectness of detections.

Initially, a number of frequent itemset mining (FIM) algorithms have been designed on the Hadoop MapReduce, a distributed big data processing framework. But, due to heavy disk I/O, MapReduce is found to be inefficient for such highly iterative algorithms. Therefore, Spark, a more efficient distributed data processing framework, has been developed with in-memory computation and resilient distributed dataset (RDD) features to support the iterative algorithms. On the Spark RDD framework, Apriori and FP-Growth based FIM algorithms have been designed, but Eclat-based algorithm has not been explored yet. In this paper, RDD-Eclat, a parallel Eclat algorithm on the Spark RDD framework is proposed with its five variants. The proposed algorithms are evaluated on the various benchmark datasets, which shows that RDD-Eclat outperforms the Spark-based Apriori by many times. Also, the experimental results show the scalability of the proposed algorithms on increasing the number of cores and size of the dataset.

The number of accidents in the world is increasing day by day and most of these accidents are effectuated due to Driving While Intoxicated (DWI). Therefore, since the death due to the Drunken-Driving or Driving Under the Influence (DUI) of alcohol has assumed proportion larger (60%) than one can visualize. In order to combat such life-risking problems, we have designed a mock-up. This safety serious system is actualized using ARDUINO and the main unit of this project is alcohol detecting sensor. Here alcohol sensor (MQ3) is used in a steering wheel and also in the passenger seat to sense the alcohol molecules present inside the vehicle. When sensor equipped in a steering wheel senses the alcohol content (i.e)., when the level of alcohol of driver crosses a permissible limit, fails the attempt of the driver to start the engine irrespective of the passenger. To thwart the accidents due to immediate engine failure, a timer is proposed in the exemplary. This framework is also embodied with widely used GSM MODEM and it will automatically send the distress message to the owner of the vehicle or to the family member.

Fresh water availability, conservation, and management are intense areas that need consideration. The paper focuses on automated mechanism to find effective solution to certain problems like, leakage detection when there occurs a break in the pipelines, water wastage due to overflow of water from buckets and storage tanks, no facility to check the quantity and quality of water in storage tank, use of water pipelines without the knowledge of the household owner and user needs to check manually the condition of lid of the storage tank. The system comprises of Raspberry Pi, pH value detector probe, ultrasonic sensor, solenoid operated valve, PIR sensor, light sensor, water flow meter, and a push button switch. The work also aims to notify the user of their water usage for each activity by monitoring the quantity of water used at each pipe and analyzing the consumption rate.

Cuckoo algorithm which is inspired from the breeding strategy of cuckoo bird is a type of meta-heuristics, which is widely used in solving NP-hard problems. Furthermore, CS has only one input parameter i.e., discovery probability. So, it is generally applicable to vast set of problems as compared to other multi parameter algorithms. Moreover, Cuckoo search uses Levy flight which is supposed to be global random walk as compared to other distribution schemes. In this paper, we have deeply studied Cuckoo Search Algorithm and Levy flight with important variants of both. We reviewed and compared Binary cuckoo search, Improved (Dynamic)cuckoo search, as variants of CS while Mantegna, Gaussian, and Cauchy distribution as variants of Levy walk. Furthermore, on the behalf of this we have proposed Dynamic Strategy based cuckoo search (DSBCS) which is supposed to give better results due to dynamic discovery probability, best global walk, good neighbor strategy after discovery and solution preserving strategy for replacement.

The Internet of Things (IoT) ML and AI is an enormous gathering of gadgets containing sensors or actuators associated together over wired or remote systems. With a gauge of more than 25 billion gadgets associated together by 2020, IoT using ML and AI has been quickly becoming over the previous decade. Amid the development, security has been distinguished as one of the weakest territories in ML IoT. While actualizing security inside an IoT using AI arrange, there are a few difficulties including heterogeneity inside the framework just as the amount of gadgets that should be tended to. To approach the difficulties in verifying IoT gadgets in ML, we propose utilizing AI inside an IoT door to help secure the framework. We explore utilizing Artificial Neural Networks in a portal to distinguish oddities in the information sent from the edge gadgets. We are persuaded that this methodology can improve the security of IoT frameworks.

In recent years, intrusion becomes one of the greatest problems which affect the community network. In Windows, firewall takes action on the bases of rules defined in it and there is no such software which has multi-faceted features i.e. against attacks, viruses and location tracking. This system is easy to understand and use by novice users as they cannot do the same with big Intrusion Prevention System like Snort, Suricata, etc. It is a user-friendly graphical interface system that provides all basic networks and system securities just on one click. It will provide secured environment in term of detection against intrusions in personal computers by hackers. The system is tested over Windows operating system, based on the signature method for the novice users in a community network. The system will scan the port and track the location of the attacker. If any suspicious IP is found, then those IP will be automatically blocked by the system. The system will have its own database which maintains the activity log.

In today’s scenario where everything is internet based, data security is on high demand. Data is considered as a crucial asset and is required to be protected. Data Security includes data confidentiality, integrity, authenticity and much more. Despite of several approaches to data security such as encryption, compression, steganography, data is still prone to potential security threats. The major reason identified is the shortcomings of these individual techniques when certain factors are taken into account. Cryptography ensures security but it comes at cost in terms of time and space. Compression techniques encrypts data and saves disk space but these compressed files can easily be uncompressed by attackers. Steganography a technique which hides data instead of encrypting it also has its merits and demerits. In this paper, approaches for data security (Encryption, Compression, Steganography) and their potential advantages and disadvantages are studied in detail. A comparison of these approaches has been generated successfully and a hybrid approach is proposed named as Extended ECS (Encryption, Compression, and Steganography) by taking selected sub approaches only with aims to overcome the flaws of existing approaches keeping high data security and low complexity as aims in mind.

Graphene-based antennas are artistic concept for wireless communications. In this paper, a Dual Band antenna using Graphene and FR4 is proposed. Graphene is used a patch and ground plane whereas FR4 is used as a substrate material. The graphene antenna is evaluated in terms of return loss, radiation pattern and current distribution. The radiation pattern of Graphene antenna shows omnidirectional pattern with no null. The simulated impedance bandwidth of the graphene-based antenna is 5.06% (2.43–2.30 GHz) and 3.80% (5.52–5.12 GHz), which makes the antenna suitable for Dual-Band Wireless Local Area Network (WLAN)/WiMAX communication.

In cloud storage services, users can store their data remotely on the cloud and exchange data with others. The cloud files consist of sensitive information. At the time of file sharing, may realizing the sensitive information of files hide, but this shared file cannot be used by others. The problem is to how to perform data sharing with sensitive information. To solve this problem, propose a Revocation algorithm to avoid the use of sensitive files from unauthorized users. In the system first, The system sanitizes the data blocks and data blocks corresponding to the organization’s sensitive information, sanitizes the sanitized data blocks signatures, and finally, sanitizers the sanitized files and their corresponding signatures in the cloud. These signatures are used to verify the integrity of files sanitized during the integrity audit phase. TPA then checks the integrity of the disinfected file. In this way, we are provide security for sharing data.

System security is one of the real worries of the difficult time. With the fast advancement and monstrous utilization of web over the previous decade, the vulnerabilities of system security have turned into an important issue. Interruption identification framework is utilized to distinguish unapproved get to and uncommon assaults over the verified systems. High volume, assortment and fast of information produced in the system have made the information examination procedure to identify assaults by conventional strategies extremely troublesome. To comprehend the present status of usage of Machine and Deep learning methods for tackling the interruption recognition issues, this study paper listing out the related examinations in the continuous period focusing. This overview paper gives the various models of the detection system and briefly on Machine and Deep learning algorithms.

In this digital era, the usage of internet and information grows rapidly. Every fraction of second, huge volume of data is transferred from one network to another. This information and information system are subjected to attack. It is necessary to protect this valuable information and network from intruders generally named as crackers or hackers who are threat to system security. System security is a common, current and critical problem which is a challengeable task to researchers. Intrusion Detection System (IDS) offers good solution to this problem. With aim of boost up the performance of IDS, it is integrated with data mining. Various data mining techniques in IDS, based on certain metrics like accuracy, false alarm rate, detection rate and issues of IDS have been analyzed in this paper. A total of 43 papers were reviewed in the period 2008 to 2018. It is observed that more number of articles support SVM or ANN Techniques. Also it is observed that hybrid methods produce better performance than single. This survey shows that in hybrid methods, frequently K-means or SVM technique are combined with others and gives good result.

A Mobile ad-hoc Network is set of nodes which are self organized in a wireless system. The attack on the network will be either attack or passive attack. The malicious node in the network can act as a base node for vampire attack which drains the life of other nodes in the network. In this study, we have proposed a reputation based framework for detecting malicious and selfish nodes in manets using AODV and PLGP protocol and illustrating several cases on how vampire attack can be prevented in the network using backtracking technique.

In modern world, high performance is achieved with the advancement of science and technology. In the fast way of life, the developments of automation projects are required. Automation systems are mainly developed by using Internet of Things. Likewise, the project represents the implementation of the Smart Mirror for an individual use instead of a normal mirror. The project is based on the Intel Atom Processor- Z-series which runs with the help of an open source operating system. Using speech processing techniques, the pilot can interact with the Smart Mirror through verbal commands. It actively listens to the pilot’s command and once the subsequent voice command is recognized, it performs the function associated with it. Smart Mirror provides common functionalities such as daily weather, time corresponding to the location, using social applications and more. The Smart Mirror can make the people to use other devices with low contribution and get them the entire world into hands through their voice commands itself.

Extreme Learning Machine (ELM) is a single layer feedforward neural network (SLFN) that has shown remarkable results in regression and classification (multi-class) problems. The theories on ELM indicates that the hidden neurons can be randomly generated. In this paper, we introduce a hybrid Recurrent Neural Network(RNN) – ELM hybrid structure for crime hotspot classification. The RNN extracts the features from the data and learns using Long Short-Term Memory (LSTM) and finally ELM is applied at the end of the layers for our classification problem. The dataset used for this study is Philadelphia’s crime data. The dataset is also tested with RNN using backpropagation without ELM. With ease of implementation, fast learning speed and better accuracy, RNN-ELM clearly outperformed RNN with backpropagation.

Crowdsourcing is developing as a conveyed critical thinking and business creation in recent years. The expression “crowdsourcing” was authored by Jeff Howe in 2006. From that point forward, a great deal of work in Crowdsourcing has concentrated on various parts of publicly supporting, for example, computational procedures and performance analysis. Declarative crowdsourcing frameworks help diminish the complexities and conceal them from users and manages the weight of the crowd. Crowdsourcing has been a critical perspective with regards to locate a specific information in a database. Crowdsourcing gives an amazing platform to execute inquiries that require progressively human talents, insight and investigation rather than simply counterfeit canny computers, which use picture acknowledgment, information filtration and tagging. Crowd optimization realizes how to adjust among cost and latency and accordingly query optimization targets are increasingly effective. CROWDOPT for upgrading three sorts of questions: selectionquires, join quiries and complex quires. In this paper, we give the outline of the survey of Crowdsourcing worldview which are arranged by the Crowdsourcing operators and datasets. In view of this study we sketch the vital components that essential to be estimated to improve Crowdsourced data management.

In IDMA conspire interleavers are utilized to recognize the distinctive clients that subsequently influence the framework throughput. In this paper the execution of Random interleaver (RI), Tree based Interleaver (TBI) and clamorous interleavers are dissected dependent on memory necessity, multifaceted nature and mistake rate. The reproduction results demonstrate that disorder based interleaver configuration can accomplish the better execution.

Automation could be a trending topic in twenty first century since it’s a significant job in our day by day lives. Automation decreases the human work, time, exertion and a couple of human mistakes moreover. This paper aims to debate home automation systems for Dysarthric persons by speech recognition and to work home appliances by their own voice. Defect of speech could be a neurologic incapacity that damages the management of motor speech articulators. It can even be outlined as a speech that’s characteristically unintelligible and slow. Someone with defect of speech might also suffer from issues like dominant the pitch, loudness, rhythm, and voice qualities of his or her speech. Several home automation systems are developed for old and disabled persons. Similar manner, this subject discussion will certainly pave the manner for several researchers in future to develop a sensible connected home for dysarthric persons.

Data mining is a method, used to extract usable and valuable information from the bulk of the data. Frequent data mining is an interesting task that is used to find frequent patterns from the database. In this paper, we used to perform frequent item set mining on temporal data. Temporal data contains data that primarily ranges over time. The idea of the time hierarchy is introduced to generate rules from temporal data. In this paper, we try to solve the problems using three popular data mining algorithms such as FP growth, Eclat and Apriori algorithm. The main focus of this study is to generate efficient algorithms that consume very less runtime and present the more frequent item set from the dataset. We evaluate our algorithms through experiments.

In mid seventies a new method of exchanging messages between electronic devices originated which revolutionized the global community into a new world of computer networks called internet. The users identified the potential usage of this method presently known as email and started using it as the means of communication and marketing. But the competence of this method was lessened by the wide spread proliferation of spam. Researchers have come up with many proposals and tools to fight against spam. But the dynamic nature of spam makes the tools ineffective and raises the requirement for developing a filter that is to be successful over time in identifying spam. Hence spam filtering is a particularly exigent machine learning task as the data distribution and concept being learned changes over time. This paper explores this phenomenon called concept drift seen in email datasets and proposes a new framework in identifying the strategies for developing spam detection systems.

Product design is an important step in creation of a new product or service. It involves the modelling of the product and also various other steps to make it unique. Conventional CAD modelling involves the usage of standard peripheral like keyboard or mouse. In this paper, Virtual Reality (VR) is used for developing a 3D product design tool. This supports more control over the dynamic models and gives a more improved user experience. Using this tool user can access 3D modelling area along with the toolbox through the VR headset. User can then create the model and interact with it using his/her hand. These interactions are monitored which in turn gives response to the processing unit to generate the model. Once the modelling is done, the model can be saved to the system in a format supported by a 3D printer or any other format supported by different modelling tool.

Light-Fidelity is a wireless technology which uses visible light as a medium instead of radio waves. To improve the performance in this technology, various studies have been carried out. Development of new technologies in the today’s world everything is extended to smart and smarter. Our environment needs to become smarter to match these smart things. In the smart museum environment, Li Fi is an efficient technology with superior bit error rate. Free space optical communication is also called as visible light communication referred ad Li-Fi. In this paper we transmit the data using LEDs in visible light region of electromagnetic spectrum. Using simple light sources we attain the transmission of data. In this paper, the Li-Fi technology is used to transmit the data in museums in order to provide the visitors a better understanding and knowledge of the ancient artefacts and technologies present around them.

Citizens living in the city are the real asset for the city. The main aim of the smart city concept is to have a satisfied citizen with better quality of life. Constant innovation in the field of ICT allows smart citizens uses this opportunity to sense the city surrounding and express their opinions and concerns on online platforms like Social Media. Hence in the era of a smart city, instead of using the traditional ways of collecting the needs and complaints of citizens like surveys and polls, government can also uses social media posts to understand the civic problems and provide the upgraded service in a timely manner. So this study presents the survey of the existing work where social media has been used to identify needs and issues faced by the people.

Light Fidelity (Li-Fi) is a visible light communication technology that uses light as a medium to provide high-speed data communication. Since the spectrum utilized by Li-Fi does not overlap with the spectrum utilized by Radio Frequency (RF) spectrum, they can be hybridized so as to improve quality of service and quality experience (Qoe) of the users. However in a hybrid Li-Fi/Wireless-Fidelity (Wi-Fi) network, movement of users may prompt frequent handover which may degrade the system throughput. Therefore there is the need to mitigate frequent handover in a hybrid Li-Fi/Wi-Fi network. This paper surveys various types of handovers that could be used to mitigate unnecessary handovers in a hybrid Li-Fi/Wi-Fi network. This paper also focuses on the working principles of Li-Fi, its applications, modulation techniques and areas of applications.

The chapter published in the book “Second International Conference on Computer Networks and Communication Technologies”, pages 342--348, https://doi.org/10.1007/978-3-030-37051-0_40 has been retracted because it reports research undertaken by Shafiya Afzal Sheikh and M Tariq Banday.