Title: Community Networks, a theme for networking research
Time: 13:00
Location: Meeting Room A, Building Zeta
Type: Research talk
Speaker: Leonardo Maccari
Abstract:
In the last few years, thanks to a series of European research projects, the scientific interest in community networks (CNs) raised considerably.
A CN is a wireless mesh network built by a community of people that grows in an unplanned way when the community enlarges.
Due to the improved performance of wireless standards, today we have mesh networks made of hundreds of nodes, or even thousands of nodes, and many stakeholders consider them a key instrument to reduce digital divide (50% of the world population was still disconnected in 2017).
CNs are also a stimulating playground for networking research to be applied to other fields. As an example, in this talk I will describe one of the challenges we tackled, which is the scalability of routing protocols.
The solution we proposed exploits the concept of betweenness centrality to fine-tune routing protocols in an automated and back-compatible way.
While we designed, tested and implemented it in CNs, we also extended the concept to other kinds of networks. Specifically, we are now considering how to apply the same principle to Internet routing, in order improve the convergence of the BGP protocol. In the process, we formalized the first fully distributed exact algorithm for centrality computation on a generic graph, which is needed to compute centrality when the full network graph is unknown. We can now study centrality-based optimizations on BGP but also in other application domains, like sensor networks.
I will conclude with a future step of this research, which is a data-based approach to generate realistic network topologies using several communication technologies. The final goal is to characterize the high level features of the network graph (cost, population coverage, robustness etc.) in order to take informed decisions on the choice of the best technology for a specific context.
The talk is based on the following publications:
1) Leonardo Maccari, Renato Lo Cigno. “Pop-routing: Centrality-based tuning of control messages for faster route convergence”. Proceedings of the International Conference on Computer Communications (INFOCOM), April 2016.
2) Leonardo Maccari, Renato Lo Cigno, “Improving Routing Convergence with Centrality: Theory and Implementation of Pop-Routing”. IEEE Transactions on Networking, vol. 26, pp. 2216–2229, Oct. 2018
3) Leonardo Maccari, Lorenzo Ghiro, Alessio Guerrieri, Alberto Montresor, Renato Lo Cigno, “On the Distributed Computation of Load Centrality and Its Application to DV Routing”. Proceedings of the International Conference on Computer Communications (INFOCOM), April 2018.