| Mm Wave-massive MIMO systems can obtain ultra-high transmission rate and positioning accuracy by beamforming,which is an important supporting technology for mobile network communication.Accurate and low-complexity direction estimation of angle-of-departure(AoD)and angle-of-arrival(AoA)is critical for channel estimation and positioning in mm Wave-massive MIMO systems.In this thesis,the normal GoB(grid of beam)codebook is designed for beam scanning for common antenna linear array and antenna plane array,and a direction estimation algorithm based on the normal GoB s is proposed.The direction can be estimated in a closed form based on the receive power of several strongest normal GoBs.Since the phase information is not needed,the proposed direction estimation algorithm has lower computational complexity than most existing algorithms.By using LDP(Large Deviation Principle)and calculating the CRLB(Cramer-Rao Lower Bound),this thesis further analyzes the performance of the direction estimation algorithm based on the normal GoB,including the probability of beam misalignment and estimation bias,and the simulations verify the theoretical performance of the direction estimation algorithm.The simulation results also show that the estimation error of the proposed direction estimation algorithm can approach its CRLB.In order to improve the efficiency of direction estimation,comb-type beam(CTB)codebook is designed by aliasing multiple equispaced normal GoBs in a hybrid beamforming system,and a direction estimation algorithm based on the CTB is proposed.The beam pattern of CTB is particularly designed such that every lobe of a CTB can be distinguished as a normal GoB,and thus the direction can also be estimated in a closed form based on the receive power of several strongest CTB.Further theoretical analyses and simulations show that the beam alignment performance of CTB depends on a higher SNR than normal GoB.When the probability of beam misalignment is low enough,the estimation deviation of the direction estimation algorithm based on CTB and normal GoB can reach its lower bound,and the proposed direction estimation algorithm based on CTB has a lower time complexity in beam scanning. |