Vellore Institute of Technology (VIT) is a private university located in&nbsp;Tamil Nadu, India. Founded in 1984, as Vellore Engineering College, the institution offers 20 undergraduate, 34 postgraduate, four integrated and four research programs. It has campuses in Vellore, Amravati, Bhopal and Chennai.

VIT is one of the top ranked private universities in India according to NIRF, THE and QS Rankings.&nbsp;Govt. of India has recognized&nbsp;VIT, Vellore as an&nbsp;Institution of Eminence. This has allowed VIT to take independent quality initiatives and move up in world ranking.

&nbsp;

&nbsp;

VIT University

Vehicular Ad Hoc Networks (VANETs) is the most growing research area in wireless communication and has been gaining significant attention over recent years due to its role in designing intelligent transportation systems. Wireless multi-hop forwarding in VANETs is challenging since the data has to be relayed as soon as possible through the intermediate vehicles from the source to destination. This paper proposes a modified fuzzy-based greedy routing protocol (MFGR) which is an enhanced version of fuzzy logic-based greedy routing protocol (FLGR). Our proposed protocol applies fuzzy logic for the selection of the next greedy forwarder to forward the data reliably towards the destination. Five parameters, namely distance, direction, speed, position, and trust have been used to evaluate the node's stability using fuzzy logic. The simulation results demonstrate that the proposed MFGR scheme can achieve the best performance in terms of the highest packet delivery ratio (PDR) and minimizes the average number of hops among all protocols. © 2020 - IOS Press and the authors. All rights reserved.

World University root node

Department of Software Systems

School of Computer Science and Engineering

Vellore Campus

Journal of Intelligent and Fuzzy Systems

Modified fuzzy-based greedy routing protocol for VANETs

India is the cultivating country and our country is the biggest maker in agricultural products. So, we have to classify and exchange our agricultural products. Manual arranging is tedious and it requires works. The automatic grading system requires less time for grading of the agricultural products. Image processing technique is helpful in examination and evaluating the products. In this paper we proposed a vegetable disease detection system for recognizing diseased vegetables. Here we utilize the Image processing system for reviewing the vegetables. Vegetables are recognized dependent on their features. The features are color, shape, size, texture. We extract these features utilizing algorithms to distinguish the vegetables. We develop a recognition system for 2D input images. The main aim of this work is detecting infected vegetable based on features with K-means clustering algorithm. Algorithm includes three main steps namely enhancement, segmentation and classification. Vegetable samples are collected as images from high resolution camera and data acquisition is carried out for database preparation. The image segmentation process is based on pixel of the image and is applied to get the segmented and infected vegetables using K-Means Clustering algorithm. The image classification is based on Support Vector Machine (SVM) which perform supervised leaning for classification. © 2020 IEEE.

Department of Software and Systems Engineering

School of Information Technology and Engineering

2020 6th International Conference on Advanced Computing and Communication Systems, ICACCS 2020

Vegetable Disease Detection Using K-Means Clustering and Svm

Agriculture has been the most important sector of our country. As the world is growing and moving in a fast pace, and depending on the automation of most of the things there is also a need to maintain plants by doing some automation. In general, every plant has some specific needs that have to be addressed for its survival. Therefore, a system must be developed where plants can communicate with the user through IoT. By monitoring these parameters, we can ensure that the plants are healthy. There is also the need to analyse, collect and make the best use of the parameters for classifications about their state. In this project, the focus is mainly on two parts. The first part aims to model a system where we keep track of the requirements of the plants using sensors and IoT. Data collection is done through sensors and sent to the Blynk app. Automatic water controller activates only when the moisture content falls below a certain threshold. The second part aims to analyse the Blynk collected data to classify the plant based on its conditions, i.e. healthy or unhealthy. Finally, we compare the performance results of support vector machine, random forest and logistic regression for classification. © Springer Nature Singapore Pte Ltd 2020.

VIT Unitversity

Advances in Intelligent Systems and Computing Artificial Intelligence and Evolutionary Computations in Engineering Systems

An Effective Approach for Plant Monitoring, Classification and Prediction Using IoT and Machine Learning

Data mining is a viable innovation to break down and extract patterns from crude information, which can change the original data into up-to-date information. Predictive analytics includes an assortment of factual systems that analyze present and historical facts to make forecasts about future or generally obscure occasions. Machine learning incorporates statistical methods for regression and classification. The objective of machine learning is to create a predictive model that is unclear from the correct model. The assessed relative execution qualities were evaluated by Ein-Dor and feldermesser utilizing a linear regression method considering the properties machine cycle time, minimum main memory, maximum main memory, cache memory, minimum channels, and maximum channels. This relationship is communicated as a mathematical statement that predicts the reaction variable published relative performance as a linear function of the parameters. The proposed technique utilizes machine learning work to re-phrase prediction as an optimization problem. Confidence prediction and polynomial regression include imaginative application utilization and promising research. The experimental evaluation platform contains detailed performance analysis of the preferred methods. It is expected that this machine learning approach gives a quick and straightforward approach to fabricate applications. © 2020 IEEE.

Department of Information Security

Proceedings of the Confluence 2020 - 10th International Conference on Cloud Computing, Data Science and Engineering

Efficient predictive model for utilization of computing resources using machine learning techniques

Swarm Intelligence (SI) is increasing day by day in the various research fields. There are many swarm-based optimizations introduced since the early '60s, Evolutionary Algorithms (EA) is the most updated one. All Evolutionary Algorithms have proved their capability to resolve most of the optimization problems. These algorithms are using for training the neural networks in this paper. The main difficulty for any optimization problem is selecting the correct values of parameters to get possible results. The main idea to get the best convergence rate and best performance is to vary the parameters of the algorithms. This paper provides a comparison of the most used and essential swarm-based optimization algorithms. Here, comparing the optimization algorithms, Particle Swarm Optimization (PSO), and Multi-Verse Optimization (MVO) before and after tuning the parameters with three different datasets. © 2020 IEEE.

Department of Analytics

Disease prediction using hybrid optimization methods based on tuning parameters

The Wastewater Management procedure, via all intents with purposes, be an extremely constructive process moreover addition into the cleanliness maintenance and upkeep of a community. It gives the inhabitants the assurance that all their wastes and garbage are processed and treated that, by the time, these excretes make get in touch through the environment, they do not cause any damage. Wastewater management is a procedure used towards exchange wastewater into an effluent that can be recurred to the water cycle with least impact lying on the environment, or else directly reused. The later is called water renovation and implies escaping of clearance with use of treated wastewater effluent for different purposes.” Proposed system configuration is to enhance the significance and the need to expand the effectiveness of the cleaning procedure of the remaining waters from waste production. The ideas about sustainable development permit solving the issues which arrive in a present movement, so that, not to occur negatively parts of that later on. It was resolved that a generally little measure of water could be gathered. After examining and assessing the systems utilized in various case studies, a hybrid system is proposed. This depends above the significance of plant-microbe collaborations in solar aquatic systems and the cost-effectiveness of regular treatment. It was resolved that potential exists for the reuse of drainage and sludge in trial applications in farming, aquaculture, residential and industrial settings. Proposed idea deals also with giving a database by utilizing Hadoop device we can check out no constraints about information also simple calculate number of tool to the group and gain conclusion among inadequate time, large throughput moreover supply is minor along with we are utilizing partitions, joins and bucketing procedure in Hadoop. © BEIESP.

International Journal of Innovative Technology and Exploring Engineering

Smart environmental waste water monitoring system and analysis using big data

Ranking of QoS provides significant facts for choosing best services offered by the cloud service providers from a group of service candidates who are functionally equivalent. Currently available frameworks must have the capacity to recommend service providers that are chosen probably by the clients. In maximum number of frameworks, rating score is used for representing the degree of preference. Customers who availed the services previously will give the ratings on a group of attributes. This input from the customers is given to traditional collaborative filtering algorithms for giving rating for the services which are not rated. We propose a multicriteria-based QoS ranking prediction framework, which guides the customers to choose the best service provider based on requirements of the customers and preferences for various services offered by the cloud by considering the benefits of past customers service usage and experiences. The framework proposed by us does not require any further real-world service problems from the intended cloud users for predicting QoS ranking. © Springer Nature Singapore Pte Ltd. 2019.

Advances in Intelligent Systems and Computing Soft Computing and Signal Processing

Multicriteria-Based Ranking Framework for Measuring Performance of Cloud Service Providers

Cloud computing is a network of computing resources that are delivered to various customers as a service over internet. Resources are available dynamically to the users; however, management of resources in a dynamic fashion stands a vast area for researchers. Resource provisioning is the process of determining, how available computing resources are allocated among various user's request in such a way that, it has to reduce the overall response time and cost. Various customers want to run different kinds of applications on the cloud server. Each customer may have different quality of service (QoS) requirements based on their workload. This makes resource provisioning a complex task. To simplify this task resource provisioning model is needed, which will enhance cloud system usability and dependability as well as quality of service. In this paper, we proposed a hybrid cuckoo search algorithm for scheduling resources to data centres in multi-cloud environment by minimising the overall response time for various user's requests. Copyright © 2017 Inderscience Enterprises Ltd.

International Journal of Intelligent Enterprise

Towards data centre resource scheduling via hybrid cuckoo search algorithm in multi-cloud environment

Cloud computing provides huge storage capacity to store all important and sensitive data as well as process the large data sets. Sensitive data means personal data about a person, which means there is something inside a person that is considered characteristically unique. For example individual people shares private data like bank account details, personal details, health records and financial data. All these private data which are stored in cloud environment needs high security and privacy. In this paper Optimized balanced scheduling is applied to perform anonymization on data sets. Here scheduling is based on the time and the size of the data sets. Anonymization approach provides privacy on individual people personal data. In large data sets it's very difficult to provide anonymizing approach; so two phase top down specialization approach is introduced to provide privacy as well as handling of large data sets in cloud. In first phase, the process of splitting large data sets into small dataset and applying anonymization on individual data sets takes place. In second phase intermediate results are merged into one and further anonymization process is applied to get the desired output. Here anonymization process is implemented using map reduce framework on cloud environment. © 2010 RJPBCS.

Research Journal of Pharmaceutical, Biological and Chemical Sciences

Optimized balanced scheduling of two phase top down specialization for diabetes patients using map reduce

"Cloud computing is a network of computing resources, that are delivered to various customers as a service over internet. Resources are available dynamically to the users; On the other hand provisioning of resources in a dynamic manner makes resource provision task more complex for scientists. As a result allocation of these resources plays a vital role in management of resources. In collaborative cloud environment, various cloud resources from different cloud servers are connected together to offer undisrupted service to consumers. Enhanced resource scheduling techniques are necessary to preserve separation among resource users. We require a resource provisioning model that will improve cloud system usability and reliability along with quality of service. In our paper we will review the work carried out by various scientists in developing resource provisioning model and their challenges in cloud computing environments and also compare the characteristics features of these resource scheduling algorithms. © International Science Press.

International Journal of Control Theory and Applications

A survey on resource provisioning models for collaborative cloud computing environment

The Cloud Computing is a next generation platform, which provides virtualization with resource pool. There are three types of cloud service models, Infrastructure-as-a-Service (IaaS), Platform-as-a-Service (PaaS) and Software-as-a-Service (SaaS). Most of the scientific research focus on IaaS model, which manage virtualization and storage. IaaS allows customer to scale based on user demand and user only pays for the resource usage. Data security plays a crucial role in cloud environment and user trust is most challenging problem of cloud services. This research paper proposed new methodology that secures data and provide privacy to the customer in cloud. Our technique providing security by using data partition approach and that partitioned data will be proceed further parallel for encryption mechanism. Here privacy is given by centric key management scheme. © 2014 IEEE.

International Conference on Information Communication and Embedded Systems (ICICES2014)

Data security and privacy using data partition and centric key management in cloud

Cloud Computing has significantly made its landmark in the field of information technology. A concept which initially stood nebulous is now used as a synonym for internet. It has paved a way to increase capacity as well as add capabilities on the fly without investing in new infrastructure, training new personnel, or licensing new software."Cloud computing in broad sense can be understood as a connected pool of computing resources (including software and hardware) which can eventually be delivered as a service over a network (in broad sense the Internet)". Resources are available to the user in utility-style infrastructure. However management of resources in a dynamic fashion stands a vast area for researchers. These resources are precisely available at certain fixed times and also for fixed intervals of time. Thus Scheduling of these resources constitutes a major part of resource management. An optimized scheduling of resources is required to maintain separation between users of the resources. In this paper, we discuss various resource scheduling strategies that have been and are being implemented in cloud computing environments. We also propose a comparison among the characteristics features of these resource scheduling algorithms.

International Journal of Engineering and Technology

Comparison of probabilistic optimization algorithms for resource scheduling in cloud computing environment

This paper addresses the very low resolution (VLR) problem in face recognition in which the resolution of the face image to be recognized is lower than 16 × 16. With the increasing demand of surveillance camera-based applications, the VLR problem happens in many face application systems. Existing face recognition algorithms are not able to give satisfactory performance on the VLR face image. While face super-resolution (SR) methods can be employed to enhance the resolution of the images, the existing learning-based face SR methods do not perform well on such a VLR face image. To overcome this problem, this paper proposes a novel approach to learn the relationship between the high-resolution image space and the VLR image space for face SR. Based on this new approach, two constraints, namely, new data and discriminative constraints, are designed for good visuality and face recognition applications under the VLR problem, respectively. Experimental results show that the proposed SR algorithm based on relationship learning outperforms the existing algorithms in public face databases.