Control of a quadrotor with network induced time delay.

A backstepping controller augmented with a state predictor is proposed to control a quadrotor over a network subjected to both state and input time delay. The state predictor predicts the future values of the states by taking the measured delayed states as input. A backstepping control law is further designed based on these predicted states. It is shown with the aid of the Lyapunov-Razumikhin theorem that the error dynamics of the predictor is asymptotically stable. The cascade of state predictor and backstepping controller makes the tracking error dynamics of the quadrotor asymptotically stable. Simulation results are presented to validate the proposed approach.