The task of performing trustworthy computation in a network of autonomous agents in the presence of misbehaving components is the subject of this work. A solution to this problem is relevant for several coordination, formation, and synchronization tasks. Relying only on the information flow of the control protocol, uncooperative behaviors are revealed using an unknown input observer technique. Necessary and sufficient conditions to detect and correctly identify misbehaving nodes are given, together with a study of the genericity of such conditions. It is shown that generically any node of the network can correctly estimate the state of the other agents, and therefore compute any function of the state, provided that the connectivity of the communication graph is strictly greater than the number of misbehaving nodes. An intrusion detection and identification algorithm is described, and its effectiveness is confirmed through a numerical study.

JF - Proc. IEEE International Conference on Decision and Control CY - Shangai, China N1 -General Chairs Recognition Awards for Interactive Papers

ER - TY - CONF T1 - Distributed intrusion detection for secure consensus computations T2 - Proc. IEEE Int. Conf. on Decision and Control Y1 - 2007 A1 - F. Pasqualetti A1 - A. Bicchi A1 - F. Bullo KW - Embedded Control KW - Robotics AB -This paper focuses on trustworthy computation systems and proposes a novel intrusion detection scheme for averaging networks with misbehaving nodes. This prototypical control problem is relevant in network security applications. The objective is for each node to detect and isolate the misbehaving nodes using only the information flow adopted by standard averaging protocols. We focus on the single misbehaving node problem. Our technical approach is based on the theory of Unknown Input Observability. First, we give necessary and sufficient conditions for the misbehavior to be observable and for the identity of the faulty node to be detectable. Second, we design a distributed unknown input estimator, and we characterize its convergence rate in the

JF - Proc. IEEE Int. Conf. on Decision and Control ER -