Abstract:
In modern ages, data is increasing day by day using internet of things for various purposes.
Data intensive analysis is the major challenge because of the ubiquitous deployment of
various kinds of sensors. Traditional cloud computing infrastructure is not enough for
processing these large amount, variety or velocity of data that means big data. Traditional
cloud computing structure is geographically centralized. Balancing load for big data is a
crucial issue. A preprocessing stage is necessary to handle these big data for real time
service oriented application. In this thesis, a big data management strategy is proposed
using a preprocessing stage that is fog computing. Providing real time services from cloud
is too many time consuming for big data. Here, Fog Computing plays a vital role. The
maximum functions of processing data are implemented outside of cloud in the case of Fog
Computing and one thing considered in Fog Computing is that here memory of fog devices
is very little. So, a well-organized communication system is needed for data processing.
Here, in this work an affordable, robust and secure power supply or third party memory
management has been suggested which is Smart Grid and Smart Local Grid. The smart
grid and smart local grid or third party memory management support the customer's real
time services using fog infrastructure. In this work a hierarchical architecture has been
proposed for creating well-organized communication system for data processing in the case
of smart grid and smart local grid or third party memory management using fog
infrastructure or nodes. For task scheduling of nodes in case of fog infrastructure queue
based scheduling technique is used. In this work, an effective result for big data
management and providing real time services has been found. Here, different parameters
such as network latency, throughput have been used for measuring performance in real
time services. The overall network latency is minimized and throughput is increased in
case of fog computing.
Description:
This thesis is submitted to the Department of Computer Science and Engineering, Khulna University of Engineering & Technology in partial fulfillment of the requirements for the degree of Master of Science in Engineering in Computer Science and Engineering, November 2018.
Cataloged from PDF Version of Thesis.
Includes bibliographical references (pages 31-34).
