Chinmaya Dehury

Chinmaya Dehury
University of Tartu · Institute of Computer Science

Doctor of Philosophy

About

26
Publications
14,987
Reads
How we measure 'reads'
A 'read' is counted each time someone views a publication summary (such as the title, abstract, and list of authors), clicks on a figure, or views or downloads the full-text. Learn more
158
Citations
Introduction
Chinmaya Dehury currently works at the Department and Graduate Institute of Computer Science and Information Engineering, FUN Lab, Chang Gung University. Chinmaya does research in Parallel Computing, Operating Systems and Algorithms. Their current project is 'Design and Implementation of Data Processing Algorithms for Analyzing Real Time Big Data in Cloud'.
Featured research
Article
Full-text available
Identifying and anticipating potential failures in the cloud is an effective method for increasing cloud reliability and proactive failure management. Many studies have been conducted to predict potential failure, but none have combined SMART (self-monitoring, analysis, and reporting technology) hard drive metrics with other system metrics, such as central processing unit (CPU) utilisation. Therefore, we propose a combined system metrics approach for failure prediction based on artificial intelligence to improve reliability. We tested over 100 cloud servers’ data and four artificial intelligence algorithms: random forest, gradient boosting, long short-term memory, and gated recurrent unit, and also performed correlation analysis. Our correlation analysis sheds light on the relationships that exist between system metrics and failure, and the experimental results demonstrate the advantages of combining system metrics, outperforming the state-of-the-art.
Poster
Full-text available
The proposed edited book is going to cover the relationship of recent technologies (such as Blockchain, IoT, and 5G) with the cloud computing as well as fog computing, and mobile edge computing. The relationship will not be limited to only architecture proposal, trends, and technical advancements. The book will also explore power of predictive analytics of Blockchain, IoT, and 5G data in Cloud computing with its sister technologies. Since, the amount of computing, storage and network resources increases day-by day, artificial intelligence (AI) tools are becoming more popular due to their capability which can be used in solving wide variety of issues, such as minimize the energy consumption of physical servers, optimize the service cost, improve the quality of experience, increase the service availability, efficiently handle the huge data flow, manages the large number of IoT devices, etc. Considering the popularity of above mentioned technologies and their dependence on cloud computing, we felt that there is a need to provide the perspective and practice of Blockchain, IoT, and 5G in the context of cloud computing. Objective The purpose of this edited book is to provide an in-depth understanding of issues, challenges, and solutions to process Blockchain, IoT, and 5G data in cloud computing and its sister technologies such as fog computing and edge computing.. Moreover, the edited book is targeting to provide the application-specific issues of cloud computing in the smart healthcare, smart city, and 5G wireless communications. The readers will get the fresh exposure to applying the concepts to applications and will get to know about the recent tools, software, and simulations available to experiment cloud computing ideas in relation with diversified data. In addition to that, the book will bring value by exposing the reader with Artificial Intelligence tools and algorithms available to design systems to tackle the cloud computing issues.
Preprint
Full-text available
With the increasing number of Internet of Things (IoT) devices, massive amounts of raw data is being generated. The latency, cost, and other challenges in cloud-based IoT data processing have driven the adoption of Edge and Fog computing models, where some data processing tasks are moved closer to data sources. Properly dealing with the flow of such data requires building data pipelines, to control the complete life cycle of data streams from data acquisition at the data source, edge and fog processing, to Cloud side storage and analytics. Data analytics tasks need to be executed dynamically at different distances from the data sources and often on very heterogeneous hardware devices. This can be streamlined by the use of a Serverless (or FaaS) cloud computing model, where tasks are defined as virtual functions, which can be migrated from edge to cloud (and vice versa) and executed in an event-driven manner on data streams. In this work, we investigate the benefits of building Serverless data pipelines (SDP) for IoT data analytics and evaluate three different approaches for designing SDPs: 1) Off-the-shelf data flow tool (DFT) based, 2) Object storage service (OSS) based and 3) MQTT based. Further, we applied these strategies on three fog applications (Aeneas, PocketSphinx, and custom Video processing application) and evaluated the performance by comparing their processing time (computation time, network communication and disk access time), and resource utilization. Results show that DFT is unsuitable for compute-intensive applications such as video or image processing, whereas OSS is best suitable for this task. However, DFT is nicely fit for bandwidth-intensive applications due to the minimum use of network resources. On the other hand, MQTT-based SDP is observed with increase in CPU and Memory usage as the number of...<truncted to fit character limit in Arxiv
Preprint
Full-text available
Identifying and anticipating potential failures in the cloud is an effective method for increasing cloud reliability and proactive failure management. Many studies have been conducted to predict potential failure, but none have combined SMART (Self-Monitoring, Analysis, and Reporting Technology) hard drive metrics with other system metrics such as CPU utilisation. Therefore, we propose a combined metrics approach for failure prediction based on Artificial Intelligence to improve reliability. We tested over 100 cloud servers’ data and four AI algorithms: Random Forest, Gradient Boosting, Long-Short-Term Memory, and Gated Recurrent Unit. Our experimental result shows the benefits of combining metrics, outperforming state-of-the-art.
Preprint
Full-text available
The recent advances in virtualization technology have enabled the sharing of computing and networking resources of cloud data centers among multiple users. Virtual Network Embedding (VNE) is highly important and is an integral part of the cloud resource management. The lack of historical knowledge on cloud functioning and inability to foresee the future resource demand are two fundamental shortcomings of the traditional VNE approaches. The consequence of those shortcomings is the inefficient embedding of virtual resources on Substrate Nodes (SNs). On the contrary, application of Artificial Intelligence (AI) in VNE is still in the premature stage and needs further investigation. Considering the underlying complexity of VNE that includes numerous parameters, intelligent solutions are required to utilize the cloud resources efficiently via careful selection of appropriate SNs for the VNE. In this paper, Reinforcement Learning based prediction model is designed for the efficient Multi-stage Virtual Network Embedding (MUVINE) among the cloud data centers. The proposed MUVINE scheme is extensively simulated and evaluated against the recent state-of-the-art schemes. The simulation outcomes show that the proposed MUVINE scheme consistently outperforms over the existing schemes and provides the promising results.
Additional affiliations
April 2019 - present
University of Tartu
Position
  • PostDoc Position
Description
  • Working in RADON project (Horizon 2020 EU program) in Mobile and Cloud Lab.
September 2013 - January 2019
Chang Gung University
Position
  • PhD Student
Description
  • Artificial Intelligence & Big Data Computing Lab http://abc.csie.cgu.edu.tw/

Publications

Publications (26)
Article
The devices at the edge of a network are not only responsible for sensing the surrounding environment but are also made intelligent enough to learn and react to the environment. Clustered Edge Intelligence (CEI) emphasizes intelligence-centric clustering instead of device-centric clustering. It allows the devices to share their knowledge and events...
Article
Full-text available
Identifying and anticipating potential failures in the cloud is an effective method for increasing cloud reliability and proactive failure management. Many studies have been conducted to predict potential failure, but none have combined SMART (self-monitoring, analysis, and reporting technology) hard drive metrics with other system metrics, such as...
Poster
Full-text available
The proposed edited book is going to cover the relationship of recent technologies (such as Blockchain, IoT, and 5G) with the cloud computing as well as fog computing, and mobile edge computing. The relationship will not be limited to only architecture proposal, trends, and technical advancements. The book will also explore power of predictive anal...
Preprint
Full-text available
With the increasing number of Internet of Things (IoT) devices, massive amounts of raw data is being generated. The latency, cost, and other challenges in cloud-based IoT data processing have driven the adoption of Edge and Fog computing models, where some data processing tasks are moved closer to data sources. Properly dealing with the flow of suc...
Article
The serverless platform allows a customer to effectively use cloud resources and pay for the exact amount of used resources. A number of dedicated open source and commercial cloud data management tools are available to handle the massive amount of data. Such modern cloud data management tools are not enough matured to integrate the generic cloud ap...
Article
With the increasing number of Internet of Things (IoT) devices, massive amounts of raw data is being generated. The latency, cost, and other challenges in cloud-based IoT data processing have driven the adoption of Edge and Fog computing models, where some data processing tasks are moved closer to data sources. Properly dealing with the flow of suc...
Preprint
Full-text available
Identifying and anticipating potential failures in the cloud is an effective method for increasing cloud reliability and proactive failure management. Many studies have been conducted to predict potential failure, but none have combined SMART (Self-Monitoring, Analysis, and Reporting Technology) hard drive metrics with other system metrics such as...
Article
The applications that are deployed in the cloud to provide services to the users encompass a large number of interconnected dependent cloud components. Multiple identical components are scheduled to run concurrently in order to handle unexpected failures and provide uninterrupted service to the end user, which introduces resource overhead problem f...
Preprint
Full-text available
The recent advances in virtualization technology have enabled the sharing of computing and networking resources of cloud data centers among multiple users. Virtual Network Embedding (VNE) is highly important and is an integral part of the cloud resource management. The lack of historical knowledge on cloud functioning and inability to foresee the f...
Preprint
Full-text available
In Cloud Computing, the tenants opting for the Infrastructure as a Service (IaaS) send the resource requirements to the Cloud Service Provider (CSP) in the form of Virtual Network (VN) consisting of a set of inter-connected Virtual Machines (VM). Embedding the VN onto the existing physical network is known as Virtual Network Embedding (VNE) problem...
Preprint
Full-text available
The serverless platform allows a customer to effectively use cloud resources and pay for the exact amount of used resources. A number of dedicated open source and commercial cloud data management tools are available to handle the massive amount of data. Such modern cloud data management tools are not enough matured to integrate the generic cloud ap...
Preprint
Full-text available
The applications that are deployed in the cloud to provide services to the users encompass a large number of interconnected dependent cloud components. Multiple identical components are scheduled to run concurrently in order to handle unexpected failures and provide uninterrupted service to the end user, which introduces resource overhead problem f...
Preprint
Full-text available
Fog computing is introduced by shifting cloud resources towards the users' proximity to mitigate the limitations possessed by cloud computing. Fog environment made its limited resource available to a large number of users to deploy their serverless applications, composed of several serverless functions. One of the primary intentions behind introduc...
Chapter
Computer vision is a pioneering sub-field of artificial intelligence that is used in computers for throwing a light on the visual world and better understanding of it. In crucial times like COVID-19, computer vision is used to combat all the challenges that are been faced. In healthcare field, computer vision has been used to enhance the productivi...
Chapter
Consider we are living in a remote place which is far away from any near-by hospitals or we don’t have enough time to take a leave from the work to visit a hospital or we can’t afford the rapidly increasing medical services costs. This is one of the few scenarios where we can relate to the fact that computer systems and algorithms if they may help...
Chapter
To provide cost effective cloud resources with high QoS, serverless platform is introduced that allows to pay for the exact amount of resource usage. On the other hand, a number of data management tools are developed to handle the data from a large number of IoT sensing devices. However, the modern data-intensive cloud applications require the powe...
Article
In Cloud Computing, the tenants opting for the Infrastructure as a Service (IaaS) send the resource requirements to the Cloud Service Provider (CSP) in the form of Virtual Network (VN) consisting of a set of inter-connected Virtual Machines (VM). Embedding the VN onto the existing physical network is known as Virtual Network Embedding (VNE) problem...
Article
Full-text available
The amount of data generated by millions of connected IoT sensors and devices is growing exponentially. The need to extract relevant information from this data in modern and future generation computing system, necessitates efficient data handling and processing platforms that can migrate such big data from one location to other locations seamlessly...
Article
Virtual Network Embedding (VNE) is the process of embedding the set of interconnected virtual machines onto the set of interconnected physical servers in the cloud computing environment. The level of complexity of VNE problem increases when large number of virtual machines with a set of resource demand need to be embedded onto a network of thousand...
Article
Full-text available
Virtualization technology boosts up traditional computing concept to cloud computing by introducing Virtual Machines (VMs) over the Physical Machines (PMs), which enables the cloud service providers to share the limited computing and network resources among multiple users. Virtual resource mapping can be defined as the process of embedding multiple...
Article
With advent of new technologies, we are surrounded by several tiny but powerful mobile devices through which we can communicate with the outside world to store and retrieve data from the Cloud. These devices are considered as smart objects as they can sense the medium, collect data, interact with nearby smart objects, and transmit data to the cloud...
Conference Paper
We propose an implementation of Radix Sort algorithm with radix 2. In this paper we are trying to optimize the time as well as space taken by Radix sort algorithm to sort a large data set. Though the proposed algorithm and Radix sort both required a space complexity O(n), but implementation of the new proposed algorithm reduces the space by 10 time...

Questions

Questions (7)
Question
Southbound interface is the bridge between Control plane and Data plane. C-DPI is also the interface between Control and Data plane. Are both the terms functionally synonym to each other?
Question
Today I came across a different data center containing thousands of servers, like OVH from Europe containing more than 1 million physical servers spreaded over 8 DCs. So what actually do these servers consists of, in terms of hardware and software.
Question
In many paper related to Cloud Computing, I come across the terms virtual link and virtual node. So what do these two terms refers to?
Question
From the link "http://archives.opennebula.org/documentation:rel4.4:schg", I came to know that Match-making scheduler is used in OpenNebula for scheduling VMs onto physical processing nodes. But I am unable to find how a large number of small requests or data-intensive jobs or compute-intensive scientific requests are being served. In other words how are incoming jobs mapped to VMs?
Question
What is the meaning of Cloudlet in Cloudsim? Is it similar to an incoming job from user?
Question
Is there any similarity between a job scheduler and a parallel data processing framework? As I know a job scheduler receives a set of jobs and based on some algorithm sends them to start their execution. In contrast a data processing framework receives a job and executes in such a way that resources can be utilized efficiently and makespan of the job can be minimized. It seems there is a relation between the two.
Question
Is there any alternative of MapReduce programming model? There are so many frameworks at present based on MapReduce like Google's MapReduce, Microsoft's Dryad, Yahoo's Map-Reduce-Merge. Is there any other framework available in Cloud industry?

Network

Cited By

Projects

Project (1)
Project
Prognosis models, prediction. Traffic congestion analysis and prediction.