Connectivity-preserving distributed algorithms for removing links in directed networks

Abstract This article considers the link removal problem in a strongly connected directed network with goal of minimizing dominant eigenvalue network’s adjacency matrix while maintaining its strong connectivity. Due to complexity problem, this focuses on computing suboptimal solution. Furthermore, it is assumed that knowledge overall topology not available. calls for distributed algorithms which rely solely local information available each individual node and exchange between neighbors. Two different strategies based perturbation analysis are presented, namely simultaneous iterative strategies. Key ingredients implementing both include novel estimating eigenvectors an verifying connectivity under removal. It shown via numerical simulations type networks general strategy yields better However, comes at price higher communication cost comparison strategy.