Home // International Journal On Advances in Software, volume 8, numbers 3 and 4, 2015 // View article

Cloud Computing System Based on a DHT Structured Using an Hyperbolic Tree

Authors:
Telesphore Tiendrebeogo
Oumarou Sié

Keywords: Cloud Computing; DHT; Hyperbolic Tree; Consis- tent; Reliable; Cloud Services; Storage; Discovery

Abstract:
During the last decade, Cloud Computing (CC) has been quickly adopted worldwide, and several solutions have emerged. Cloud computing is used to provide storage service, computing power and flexibility to end-users, in order to access data from anywhere at any time. Thus, Cloud Computing is a subscription-based service where you can obtain networked storage space and computer resources. The cloud makes it possible for you to access your information from anywhere at any time. Distributed Hash Table (DHT) plays an important role in distributed systems and applications, particularly in environments distributed on a large scale. In the model of normal Client/Server (C/S model), as we centralize most of the resources on the server, it becomes the most important part as well as the bottleneck and the weak point of the system. On the contrary, distributed model (a typical is Peer-to-Peer (P2P) model) distributes the resources on the nodes in the system. In this paper, we propose a new system of Cloud Computing based on our DHT structured using an hyperbolic tree, which organizes the distributed services so well that peers only need to know part of the system they can get services efficiently. DHT provides two basic operations: retrieve service from DHT and store service into DHT, which is so simple and graceful, but is suitable for a great variety of resources (Applications, Infrastructures, Platforms), and provides good robustness and high efficiency, especially in large-scale systems. Resources as services are distributed by using virtual coordinates taken in the hyperbolic plane. We use the Poincar´e disk model and we perform and evaluate our cloud structure performances. First, we show that our solution is scalable, consistent, reliable. Next, we compare the performances realized by the substitution strategy, which we propose with the strategy of classic replication in a dynamic context.

Pages: 491 to 502

Copyright: Copyright (c) to authors, 2015. Used with permission.

Publication date: December 30, 2015

Published in: journal

ISSN: 1942-2628