Distributed Control of the Laplacian Spectral Moments of a Network

摘要: It is well-known that the eigenvalue spectrum of Laplacian matrix a network contains valuable information about structure and behavior many dynamical processes run on it. In this paper, we propose fully decentralized algorithm iteratively modifies agents in order to control moments spectrum. Although individual have knowledge their local only (i.e., myopic information), they are collectively able aggregate decide what links most beneficial be added or removed at each time step. Our approach relies gossip algorithms distributively compute spectral matrix, as well ensure connectivity presence link deletions. We illustrate our nontrivial computer simulations show good final approximation target achieved for cases interest.