Topology Inference on Partially Observable Mobile Robotic Networks under Formation Control

Abstract

The interaction topology is critical for efficient cooperation of mobile robotic networks (MRNs).We focus on the local topology inference problem of MRNs under formation control, where an inference robot with limited observation range can manoeuvre among the formation robots. This problem faces new challenges brought by the highly coupled influence of unobservable formation robots, inaccessible formation inputs, and unknown interaction range. The novel idea here is to advocate a range-shrink strategy to perfectly avoid the influence of unobservable robots while filtering the input. To that end, we develop consecutive algorithms to determine a feasible constant robot subset from the changing robot set within the observation range, and estimate the formation input and the interaction range.Then, an ordinary least squares based local topology estimator is designed with the previously inferred information. Resorting to the concentration measure, we prove the convergence rate and accuracy of the proposed estimator, taking the estimation errors of previous steps into account. Extensions on nonidentical observation slots and more complicated scenarios are also analyzed. Comprehensive simulation tests and method comparisons corroborate the theoretical findings.

I Can Read Your Mind: Control Mechanism Secrecy of Networked Dynamical Systems under Inference Attacks

“…To obtain a reliable local topology estimate, an alternative method is to shrink the inference range [12]. Let V H be a subset of V F such that…”

“…From the observed information, the attacker can infer the real historical states which may contain sensitive location information of the robot base [9], or predict the future trends of the trajectory evolution [10]. Also, if the robots exhibit a process of reaching certain regular pattern, then the attacker can infer the internal interaction structure among the robots [11], [12], and find the critical robot that has dominant impacts on the system [13]. Even worse, the control laws about how the robots are driven are also likely to be learned if sufficient data of the control process are collected [14], [15].…”

Sneak Attack against Mobile Robotic Networks under Formation Control

“…The differences between this paper and its conference version [1] include i) a systematic intelligent attack design is introduced rather than mere topology inference, ii) the learnability of both the internal interaction structure within the MRN and the external interaction mechanism with the environment is explored, and iii) the sneak attack is introduced with comprehensive feasibility and performance analysis. The main contributions of this paper are summarized as follows.…”