Approximate Byzantine consensus in sparse, mobile ad-hoc networks

We consider the problem of approximate consensus in mobile ad-hoc networks in the presence of Byzantine nodes. Due to nodes' mobility, the topology is dynamic. We propose a protocol based on the linear iteration method. The nodes collect information during several consecutive rounds: moving gives them the opportunity to gather progressively enough values. A novel sufficient and necessary condition guarantees the final convergence: from time to time only the correct nodes that own a value equal to (or very close to) either the minimum or the maximum value have to receive enough messages (quantity constraint) with either higher or lower values (quality constraint). Of course, nodes' motion should not prevent this requirement to be fulfilled. New concepts are introduced to prove the correctness of the protocol. Based on particular mobility scenarios, simulations are conducted to analyze the impact of some parameters on three variants of the protocol.

[1]  Jie Lin,et al.  Coordination of groups of mobile autonomous agents using nearest neighbor rules , 2003, IEEE Trans. Autom. Control..

[2]  Behrouz Touri,et al.  On Ergodicity, Infinite Flow, and Consensus in Random Models , 2010, IEEE Transactions on Automatic Control.

[3]  C.N. Hadjicostis,et al.  Distributed function calculation via linear iterations in the presence of malicious agents — Part I: Attacking the network , 2008, 2008 American Control Conference.

[4]  Ger Koole,et al.  The message delay in mobile ad hoc networks , 2005, Perform. Evaluation.

[5]  Shreyas Sundaram,et al.  Robustness of information diffusion algorithms to locally bounded adversaries , 2011, 2012 American Control Conference (ACC).

[6]  Danny Dolev,et al.  Optimal Resilience Asynchronous Approximate Agreement , 2004, OPODIS.

[7]  Antonio Bicchi,et al.  Consensus Computation in Unreliable Networks: A System Theoretic Approach , 2010, IEEE Transactions on Automatic Control.

[8]  E.M. Atkins,et al.  A survey of consensus problems in multi-agent coordination , 2005, Proceedings of the 2005, American Control Conference, 2005..

[9]  C.N. Hadjicostis,et al.  Distributed function calculation via linear iterations in the presence of malicious agents — Part II: Overcoming malicious behavior , 2008, 2008 American Control Conference.

[10]  H. Bajwa,et al.  Global convergence in partially fully connected networks (PFCN) with limited relays , 2001, IECON'01. 27th Annual Conference of the IEEE Industrial Electronics Society (Cat. No.37243).

[11]  Nancy A. Lynch,et al.  Reaching approximate agreement in the presence of faults , 1986, JACM.

[12]  Randal W. Beard,et al.  Consensus seeking in multiagent systems under dynamically changing interaction topologies , 2005, IEEE Transactions on Automatic Control.

[13]  Shreyas Sundaram,et al.  Consensus of multi-agent networks in the presence of adversaries using only local information , 2012, HiCoNS '12.

[14]  Luoyi Fu,et al.  Fundamental Relationship between Node Density and Delay in Wireless Networks with Unreliable Links , 2011 .

[15]  Sébastien Tixeuil,et al.  Brief Announcement: The BG-Simulation for Byzantine Mobile Robots , 2011, DISC.

[16]  Xinbing Wang,et al.  Fundamental relationship between NodeDensity and delay in wireless ad hoc networks with unreliable links , 2011, MobiCom '11.

[17]  Noa Agmon,et al.  Fault-tolerant gathering algorithms for autonomous mobile robots , 2004, SODA '04.

[18]  Mehran Mesbahi,et al.  On maximizing the second smallest eigenvalue of a state-dependent graph Laplacian , 2006, IEEE Transactions on Automatic Control.

[19]  Lewis Tseng,et al.  Iterative approximate byzantine consensus in arbitrary directed graphs , 2012, PODC '12.