Weighted distributed consensus algorithm based on label propagation algorithm

Abstract In the practical application of multi-agent system, the convergence rate of the distributed consensus algorithm becomes slow with the expansion of the communication topology. In order to accelerate the convergence rate of distributed consensus under complex topology, in the paper, based on the fast linear iterations distributed consensus algorithm, a weighted distributed consensus algorithm based on label propagation algorithm was proposed. In the algorithm, firstly based on label propagation algorithm, the complex topology was composed into two layer of topologies, in the first layer of topology, the information was exchange each other within the same community, in the second layer of topology, the information was exchange within the different communities . The consensus firstly was reached in the first layer of topology, then the consensus was reached in the second of layer topology. The analysis and simulation of the convergence performance and the convergence speed were done. The results show that the average consensus can be reached in the weighted distributed consensus, and the convergence rate of the proposed algorithm was higher than that of the fast linear iterations distributed consensus algorithm.