Expansion Properties of Topology for Networking of Information in Cloud

Abstract

Toward the progress in the era of globalization and ubiquity of sensors and devices, sharing and dissemination of information dominate todays networks. Content-centric networking, cloud services, and open connectivity form the main ingredients of the future Internet architecture. With the problem of information overload, the networking paradigm of cloud computing can benefit from transitioning to a network of information in which information is the main token of communication instead of physical address. Available methods may not be efficient in exploiting the semantics of information for content dissemination. Considering a content-centric approach, we intend to tackle this problem by using the expander graphs for an enhanced network coding scheme that takes an opportunistic strategy to utilize the spectral characteristics of the network topology to achieve a better solvability and reliability and lowering the processing cost for the entire system. By simulation and analytical evaluation, we compare our proposed method with an epidemic network coding based approach. Our evaluation examines the performance of our clustering method in the presence of different random topology models as well as examining the impact on the network coding technique.