The invention discloses a 
millimeter wave anti-blocking multi-unmanned aerial vehicle (UAV) deployment method based on building 
geometric analysis, and mainly solves the problems of user clustering and UAV deployment when UAVs provide communication services for users in an 
urban environment. The specific method comprises the following steps: firstly, establishing a 
millimeter wave channel model capable of predicting 
signal blocking based on building 
geometric analysis; providing a BT-K-means multi-unmanned aerial vehicle clustering 
algorithm, and performing multiple iterations according to a minimum 
path loss principle to obtain a clustering result of the user and an initial position of the unmanned aerial vehicle; in each cluster, providing an unmanned aerial vehicle position coordinate searching method based on 
path loss sudden change 
perception, searching the minimum 
path loss value in six coordinate axis directions at the same time, unmanned aerial vehicle coordinates are continuously updated through multiple iterations, unmanned aerial vehicle position optimization is achieved, and the final deployment position of each unmanned aerial vehicle is obtained. According to the 
algorithm provided by the invention, the number of 
sight distance transmission users can be increased, and meanwhile, the total path loss of the whole 
system is effectively reduced.