Improving Routing Convergence With Centrality: Theory and Implementation of Pop-Routing

Abstract

One of the key features of a routing protocol is its ability to recover from link or node failures, recomputing routes efficiently without creating temporary loops. Link-State protocols perform better than Distance-Vector ones and they are often presented as ideal from this perspective. Indeed, in real conditions there is always a trade-off between the overhead due to the periodic generation of control messages and route convergence time. This work formalizes the problem of the choice of timers for control message generation as an optimization problem that minimizes the route convergence time constrained to a constant signaling overhead. The solution requires the knowledge of nodes' centrality in the topology and can be obtained with a computational complexity low enough to allow on-line computation of the timers. Results on both synthetic and real topologies show a significant shrinkage of the transient duration with the consequent performance gain in terms of reduced number of unreachable destinations and routing loops that lead to traffic loss. Next, we present the extension of OLSRv2 with our proposal, named Pop-Routing, and discuss its performance and the stability of centrality metrics in three large-scale real wireless mesh networks. This exhaustive analysis on traces of the topology evolution of real networks for one entire week show that Pop-Routing outperforms the non-enhanced protocol in every situation, even when it runs with sub-optimal timers due to centrality computation on stale information. This situation must be taken into account, as, albeit computationally lightweight, the optimization cannot run in real-time on wireless routers.

Optimized Framework for QoS Efficient Routing Scheme in Wireless Mesh Network

“…This model allows a dedicate communication channel in order to carry out transmission and its outcome was found to support futuristic communication protocols too. The work of Maccari et al [22] has addressed the problem associated with node failures as well as link failures and issues associated with their recovery in mesh network. The author has presented a solution by using the centrality concept of the information of the node over the topology with a target to improve the link state routing.…”

Towards scalable Community Networks topologies

“…The network-aware strategy does not necessarily add the fastest new end if 25: end for 26: return(nodes below threshold, min(bottleneck)) link. It rather evaluates all the candidate neighbor nodes as to how a link to them impacts the performance of the whole network (through the checkStopCondition at line 6 in Algorithm 3 that returns R min ) and then selecting n maximizing this value (lines [11][12][13][14][15]. The fastest link is used as a fallback node selection criterion only if there is a tie.…”

“…Network overhead is small for distance vector protocols and it grows roughly linearly with the nubmer of nodes (less than 200 bit/s for a 25 node network, and still less than 0.5-1 kbit/s for a 100 node network for a typical topology) while Link State protocols (like OLSRv2) generate a higher overhead. Yet its impact, albeit non negligible, can be controlled with proper tuning [15]. Therefore, we do not consider such metrics in this work.…”

A Big Data and machine learning approach for network monitoring and security

“…For 1 week, every 5 minutes we dumped the network topology annotated with the ETX value per link. The whole dataset has been used for several publications and is available online. For this paper we use only one network, the “ninux” network of Rome, made of about 130 nodes.…”