A joint peer-to-peer and network coding approach for large scale information management

Abstract

The widespread availability of connectivity to the Internet allows to share large amount of information generated by the most heterogeneous, possibly mobile, sources. One scenario where this situation arises is given by smart cities, which are envisioned to generate and consume relevant information about their statuses to enhance the security and lifestyle of their citizens. In this context, a very challenging question is how the information can be maintained and distributed among the city itself. In this paper, we propose a system architecture based on the creation of a distributed geographic overlay network, which allows to achieve the desired goals. Moreover, information is redundantly encoded by means of randomized network coding, in order to dynamically and distributedly preserve the resource availability. By means of simulations, we investigate the behavior of the proposed solution, in terms of efficiency and speed in data publication/search, as well as resource availability and storage occupancy requirements.