Consensus Seminorms and their Applications

Abstract

Consensus is a well-studied problem in distributed sensing, computation and control, yet deriving useful and easily computable bounds on the rate of convergence to consensus remains a challenge. We study the applications of seminorms for this goal. We revisit a previously suggested family of seminorms and correct an error made in their original presentation where it was claimed that the a certain seminorm is equal to the well-known coefficient of ergodicity. We then propose a wider family of seminorms which guarantee convergence at an exponential rate of infinite products of matrices which generalizes known results on stochastic matrices to the class of matrices whose row sums are all equal one. Finally, we show that such seminorms cannot be used to bound the rate of convergence of classes larger than the well-known class of scrambling matrices, and pose several open questions for future research.