Downloads: 126 | Views: 194
M.Tech / M.E / PhD Thesis | Computer Science & Engineering | India | Volume 3 Issue 6, June 2014 | Rating: 6.6 / 10
Real Time Game Theoretic Load Balancing With Cluster Based Load Classification for Private and Public Virtual Cloud
Srikant Bagewadi | Rekha Patil [9]
Abstract: Cloud is computational and architecture ion of multiple server communication where a broker provides a layered ion to the user from the underneath services; Hence load balancing is an integral and essential part of every cloud. Conventional load balancing techniques mainly focuses on providing an optimum service to the user by tactically devising a suitable load balancing strategy such that users request can be processed with optimum speed. However the existing system does not takes into account the computational constraints of server. In these project we devised a technique based on game theory approach that solve the load balancing problem and also it take into account the computational constraints of server; so we first introduce a unique method of classifying the load based on clustering technique and we propose a unique system where the service acquires and exposes the information like available bandwidth; latency; memory and other informations specific to the server which is used to performs context switching and load balancing and also with the help of real time service deployed in a real private and public virtual cloud; we show that the proposed system performs better in-terms of optimum utilization of resources; load equality and execution time.
Keywords: Cloud Computing, Load Classification, k-means Clustering, Load Balancing, and Game Theory Approach
Edition: Volume 3 Issue 6, June 2014,
Pages: 1427 - 1431