MS Defense: Abdullah

Thursday, November 21, 2019 - 2:00pm to 3:00pm
Atanasoff 216
Event Type: 

Data Compression Based Cost Optimization for a Multi-Cloud Data Storage System

A client wants to store data on cloud but given multiple heterogeneous cloud service providers available in the market, it is very difficult for the user to make the best choice. The cloud service providers differ in the services provided, cost of services, security, availability, latency, bandwidth etc. The number of cloud service providers in the market is already very large. Further, one cloud service provider like Amazon offers about 70 different cloud services for the users, which makes it very tedious for customers, especially small and medium size businesses to search through all the available cloud providers and make an informed decision. Cloud broker can facilitate users by selecting the best cloud service provider for their data meeting their service level agreement (SLA) requirements. The cloud broker keeps up-to date information about the cloud providers in the market, like the services provided, price of services, QoS, latency, datacenter regions, availability and security etc. The questions of concern are; How to select the datacenter which meets user's SLA requirements with minimum cost? And Can data compression be helpful in saving storage space and bandwidth use? We develop the datacenter selection problem as an optimization problem with an objective function and constraints. To solve the optimization problem, we develop heuristic solutions which include three algorithms namely Simple Greedy Algorithm, Partly Compressed Algorithm and Fully Compressed Algorithm. First algorithm stores the data in plain text with no compression, while the cloud broker explores effects of data compression to save storage space and bandwidth in the next two algorithms. Simulation experiments are conducted to evaluate the optimal datacenter selection results from three algorithms.

Committee: Wensheng Zhang (major professor), Johnny Wong, Guang Song