Abstract: The global optimal leader-following consensus problem of discrete-time multi-agent systems is studied. Two major challenges for the problem are the coupling effect between distributed communication and global optimal control, and the computational complexity that grows rapidly with the increase in the number of agents. To solve the above problems, weighting matrices are designed to realize the computational decoupling. An innovative information fusion method is proposed using a dynamic average consensus algorithm to realize the distributed optimal leader-follower control of the agents. Finally, the effectiveness of the proposed method is verified by numerical simulation.