Stochastic Consensus Algorithms over General Noisy Networks

Stochastic consensus algorithms are considered for multi-agent systems over noisy unbalanced directed networks. The graph which represents a communication network of the system is assumed to contain a directed spanning tree, that is, a given digraph is weakly connected. Then two types of stochastic consensus are investigated, where one is for the agent states themselves and the other is for the time averages of the agent states. The convergence of the algorithms is investigated, which gives a stopping rule, i.e., an explicit relation between the number of iterations and the closeness of the agreement