Self-tuning average consensus in complex networks

In this paper we present an adaptive algorithm for distributed average consensus over a network of multi-agent systems. The coupling parameters defining the strength of agents interactions are locally self-tuned by each node based on the state information of its neighbors. Assuming that the underlying graph is connected, it is shown that the sequence of coupling parameters generated by normalized gradient algorithm (NGA) is convergent, and the agent states converge toward the average of the initial state values. Relation of the proposed method to synchronization phenomenon is discussed. Simulation results illustrate effectiveness of the proposed method.