Abstract: In order to improve the network performance of airborne multi-agent systems, a model-based adaptive motion control method for UAV communication relay is proposed in this paper. The relay motion control problem is solved by jointly considering the unknown RF channel parameters, unknown multi-agent mobility, and unavailable angle of arrival information of received signals. An estimation algorithm based on Gaussian process learning and online data measurement is proposed to estimate wireless channel parameters between the UAV and each agent. Two different relay application scenarios are considered: end-to-end communication and multi-node communication. For the former, a line search algorithm is proposed, and its stability and convergence are given and proved; for the latter, a general gradient-based algorithm is proposed to provide a target relay position at each decision time step, reducing the two-dimensional search to one-dimensional search. Simulation results show that the proposed relay motion control algorithm can drive the UAV to reach or track the motion of the optimal relay position and improve the network performance.