The invention discloses a predictive cooperative beam forming and
resource allocation method based on multi-
station information fusion, which is suitable for a cellular-free large-scale
MIMO unmanned aerial vehicle communication
system. According to the method, a low-overhead
beam tracking frame structure is designed on a ground
base station side, and
position tracking of a mobile unmanned aerial vehicle user is realized by using an uplink
pilot signal and based on an extended Kalman filtering distribution mode; and then, designing beam forming based on position prediction to realize downlink
data transmission. Each
base station returns
user state estimation and
covariance information in a low-overhead mode through a forward link, and the
central processing unit fuses information of multiple base stations to improve
position tracking precision. Furthermore, according to the method, the
system effectiveness and the spectrum efficiency are taken as optimization targets, the uplink
pilot frequency length, the downlink
base station user association and the power distribution are jointly optimized, and a low-complexity iterative
algorithm based on
compressed sensing and fractional
programming is used for solving. According to the invention, the
pilot frequency training and backhaul overhead can be obviously reduced.