On the Distributed Computation of Load Centrality and its Application to DV Routing

Abstract

Centrality metrics are a key instrument for graph analysis and play a central role in many problems related to networking such as service placement, robustness analysis and network optimization. Betweenness centrality is one of the most popular and well-studied metric. While distributed algorithms to compute this metric exist, they are either approximated or limited to certain topologies (directed acyclic graphs or trees). Exact distributed algorithms for betweenness centrality are computationally complex, because its calculation requires the knowledge of all possible shortest paths within the graph. In this paper we consider load centrality, a metric that usually converges to betweenness, and we present the first distributed and exact algorithm to compute it. We prove its convergence, we estimate its complexity and we show it is directly applicable-with minimal modifications-to any distance-vector routing protocol based on Bellman-Ford. We finally implement it on top of the Babel routing protocol and we show that, exploiting centrality, we can significantly reduce Babel's convergence time upon node failure without increasing signalling overhead. Our contribution is relevant in the realm of wireless distributed networks, but the algorithm can be adopted in any distributed system where it is not possible, or computationally impractical, to reconstruct the whole network graph at each node and compute betweenness centrality with the classical approach based on Dijkstra's algorithm.

Exact Distributed Load Centrality Computation: Algorithms, Convergence, and Applications to Distance Vector Routing

“…Lorenzo Ghiro PhD grant is partially supported by the IEEE Smart Cities Initiative at Trento. This paper revises and extends a paper presented at IEEE INFOCOM in 2018 [1], by studying the behavior of the protocol under partial information. This work was initiated when all the authors were with the University of Trento global shortest paths passing through a vertex [2].…”

“…The silent propagation of subTLVs is implemented when sending route advertisements (line 8), sent periodically when the node sends DVs to all its neighbors 1 . Beyond the funda- Compared to a legacy node, an upgraded one is also able to parse subTLVs.…”

Where have all the MPRs gone? On the optimal selection of Multi-Point Relays

“…While HELLO messages are used to locally detect these changes, TCs are needed to notify the rest of the network in order for all the routing tables to be recalculated. In [38,39,40] we introduced Pop-Routing as a technique (fully compatible with MPRs) that addresses this issue and we give a detailed analysis of the relationship between TCs and network convergence showing how this timer influences route convergence time.…”

Network Meta-Analysis of Chicken Microarray Data following Avian Influenza Challenge—A Comparison of Highly and Lowly Pathogenic Strains

“…The following supporting information can be downloaded at: , Table S1: Previous network based studies in chicken [ 44 , 45 , 46 , 47 , 48 , 49 , 50 , 51 , 52 , 53 , 54 , 55 , 56 , 57 , 58 , 59 , 60 ], Table S2: Network metrics used in this study [ 61 , 62 , 63 , 64 , 65 , 66 , 67 ], Table S3: Differentially expressed genes from LPAI data analysis, Table S4: Differentially expressed genes from HPAI data analysis, Table S5: Gene Ontology of up-regulated HPAI DEGs, Table S6: Gene Ontology of down-regulated HPAI DEGs, Table S7: Gene Ontology of up-regulated LPAI DEGs, Table S8: Gene Ontology of down-regulated LPAI DEGs, Table S9: HPAI upstream activators.…”