Abstract: In the application scenario of hot spots coverage with UAV (unmanned aerial vehicle) swarm communications, a channel estimation algorithm for virtual large-scale multiple input multiple output channel in UAV swarm communication is proposed. The proposed channel algorithm includes a direction of arrival (DOA) estimation algorithm and sub-array spacing estimation algorithm in the steering-vector of the channel state information. Considering that the air to ground channel state depends on the angle domain information of the ground users, the auxiliary user is used to estimate the direction angle of the UAV. Based on this, a reduced rank based DOA estimation algorithm is proposed to obtain high-precision DOA information. Furthermore, considering that the dynamic position change of UAV results in the relative position change of antenna arrays of different UAVs, a sub-array spacing estimation algorithm based on optimization search is proposed to avoid the high computational complexity caused by large-scale search. Simulation results show that the proposed DOA and sub-array spacing estimation algorithm can improve the accuracy of channel estimation.

Key words: UAV Communications, channel estimation, massive MIMO