Research Of Key Technologies Of Beamforming For Millimeter Wave Wireless Communication

The significant increase in wireless data traffic in recent years has driven the need to explore suitable areas in the radio spectrum to meet projected demand.For this reason,millimeter-wave communication has received considerable attention in the research community.Because the millimeter wave has a shorter wavelength,it will cause higher path loss and penetration loss,so the beamforming technology of large-scale antenna arrays plays a pivotal role in establishing and maintaining robust communication links.Due to the wide channel bandwidth,unique channel characteristics and hardware limitations of millimeter wave,using beamforming technology directly on millimeter wave communication will bring a variety of challenges.In the current mm Wave communication,there is already a standard beam training protocol,but it has higher complexity and lower accuracy,so there is still a great prospect for development in terms of complexity and accuracy.Faster and more accurate beam training scheme.When deploying larger arrays,the overhead due to beam training is usually higher.When the number of users increases,a base station transmits a beam to cover more users,so the problem of users that need to be distinguished on a beam arises.To this end,this thesis conducts research on the key technologies of beamforming for millimeter-wave wireless communication,and designs it through beam training.The main work and innovation of this article are summarized as follows:1.To reduce the training overhead,a novel position-aided fast millimeter-wave beam training using compressive sensing(PAF-CS)is proposed.Aiming to resolve the position uncertainty,the PAF-CS method first identifies a critical area by using the statistical information of the position error,and finds the best beams by using CS method.On the other hand,the simulation results also verify that the proposed method not only has extremely low beam training complexity,but also the performance is close to the upper bound.2.Aiming at multi-user beam training,in order to reduce the beam training overhead and solve the problem of beam misalignment due to position error,a hybrid beamforming method based on robust position-aided millimeter wave beam training is proposed.The method divides the hybrid beamforming design into two stages.In the first stage,a robust beam training scheme is used to find the best beam pair without explicit channel estimation,and directly design the analog precoder of the base station and the analog combineder of the user.The digital precoding of the base station is designed by selecting the strongest equivalent channel gain to eliminate inter-cluster interference.Finally,the simulation results verify that the proposed algorithm reduces training overhead and the performance is close to the optimal method.3.For the hybrid millimeter wave-NOMA communication system,in order to solve the problem of distinguishing users of the same beam,a beam training scheme suitable for the millimeter wave-NOMA system is proposed.In order to overcome the physical limitations of narrow analog beams,a beamwidth adjustment algorithm based on Hierarchical Codebook based on mm Wave-NOMA(HCM)is proposed to adaptively widen the analog beamwidth to cover all users of the same NOMA cluster.The proposed method assumes that BS do not know the SCI of each user.In our design,at first BS uses the highest layer of HCM to sweep beams,then all users feed back the beam indexes.Next we proposed a K-means learning framework for user clustering,which exploits the beam indexes of users.Subsequently,for each cluster,we select its analog beamforming vector from the HCM depending on its beam indexes of each cluster and the digital precoding is designed to cancel the inter-cluster interference by selecting the strongest equivalent channel gain.At last,we derive an optimal power allocation method for the sum rate maximization by taking into account the quality of service(QOS)requirement.The new hybrid hierarchical beamforming codebook can adjust the beamwidth to maximize the system sum-rate.Compared with similar algorithms,the simulation results show that the proposed algorithm has better performance.