The state estimation problem of unmanned aerial vehicle (UAV) system with random parameter uncertainty and data packet missing under the phenomenon of network communication link failure and system parameters uncertainty is studied. Based on moving horizon estimation theory and random least squares theory, a recursive distributed moving horizon estimation algorithm is proposed. For parameter uncertainty and data packet missing problems, the model adopts system matrix disturbance noise and Markov random series based on given probability. The simulation results show that the proposed algorithm is better than a new Kalman filtering algorithm. Finally, the effects of system compression, data packet missing probability and time window length on the estimated performance of the proposed algorithm are analyzed. In the case where the system uncertainty and the data packet missing probability are unknown, an appropriate increase in the window length can improve the estimation performance of the algorithm.