Fairness-oriented Overlay VPN topology construction

An important issue in dynamically constructed Virtual Private Networks (VPN) is how the overlay topology is created and maintained. Classical VPN topologies, such as hub-and-spoke or full-mesh, fail to remain convenient and viable when the number of nodes grows to as little as a few tens. Convenient topology formation mechanisms should be distributed, should permit incremental and dynamic operations, and should limit the number of nodes a new entry connects with. In this work, we show that approaches devised to create “short” networks, while yielding a significant total network throughput, may be severely affected by unfairness issues, i.e., different pair of nodes may experience a widely different throughput performance. Hence, we introduce a fairness-oriented topology formation algorithm for VPN. The proposed algorithm is incremental, meaning that the addition of a new node to the overlay topology does not imply rewiring of already established overlay links. Simulation results show that our proposed approach achieves high fairness levels, as quantified in terms of well known Jain's fairness index, meanwhile retaining satisfactory throughput performance.

[1]  Azer Bestavros,et al.  EGOIST: overlay routing using selfish neighbor selection , 2008, CoNEXT '08.

[2]  Raj Jain,et al.  A Quantitative Measure Of Fairness And Discrimination For Resource Allocation In Shared Computer Systems , 1998, ArXiv.

[3]  Markus Brede,et al.  Networks That Optimize a Trade-Off between Efficiency and Dynamical Resilience , 2009, Complex.

[4]  Chase Cotton,et al.  Packet-level traffic measurements from the Sprint IP backbone , 2003, IEEE Netw..

[5]  Azer Bestavros,et al.  Implications of Selfish Neighbor Selection in Overlay Networks , 2007, IEEE INFOCOM 2007 - 26th IEEE International Conference on Computer Communications.

[6]  Caterina M. Scoglio,et al.  On graph-based characteristics of optimal overlay topologies , 2009, Comput. Networks.

[7]  S. Gago,et al.  Notes on the betweenness centrality of a graph , 2012 .

[8]  Prasant Mohapatra,et al.  The impact of topology on overlay routing service , 2004, IEEE INFOCOM 2004.

[9]  Oliver Paridaens,et al.  Scalability implications of virtual private networks , 2002, IEEE Commun. Mag..

[10]  Antonio Capone,et al.  Routing and resource optimization in service overlay networks , 2009, Comput. Networks.

[11]  Ashfaq A. Khokhar,et al.  Virtual private networks: an overview with performance evaluation , 2004, IEEE Communications Magazine.

[12]  U. Brandes A faster algorithm for betweenness centrality , 2001 .

[13]  Bert de Vries,et al.  Networks That Optimize a Trade-Off between Efficiency and Dynamical Resilience , 2009, COMPLEX 2009.

[14]  Francesca Lo Piccolo,et al.  OPSS: an overlay peer-to-peer streaming simulator for large-scale networks , 2007, PERV.

[15]  Albert-László Barabási,et al.  Statistical mechanics of complex networks , 2001, ArXiv.