Signal Comparison Average Consensus Algorithm Under Binary-Valued Communications

The average consensus problem under binary-valued communications is investigated in the paper. Measurement noises and fixed quantizers are considered. A signal comparison algorithm is proposed for the problem. Neither the noise distribution information nor assumptions on the states' approximate locations are required for the algorithm design. The signal comparison algorithm is proved to achieve average consensus both in the almost sure and mean square sense. The algorithm's mean square convergence rate is also calculated. The efficiency of the algorithm is demonstrated by a numerical example.