Research On Beam Search And Channel Estimation Methods For Millimeter Wave Communication Systems

As millimeter-wave signals suffer from high attenuation and weak penetration,millimeter-wave system shall apply large-scale array antennas to achieve high-gain directional beamforming,which enables compensation for the propagation loss.Current millimeter-wave communication system mainly employs two types of architectures,i.e.,analog beamforming and hybrid beamforming.Therefore,from the perspective of reducing time delay and pilot overhead,this thesis aims at improving the effectiveness and reliability for millimeter-wave communication systems under different architectures.Corresponding signal processing methods are studied in the following two aspects:1)For the millimeter-wave communication system with analog beamforming,a multi-power-level beam sensing algorithm is first proposed.Received signals are quantized in the energy as well as angular domain to establish the decision and access criterion.Simulations are then given to analyze the influence of several factors,e.g.,sector width and numbers of power level,to the detection performance of the proposed algorithm.Further,a beam search scheme aiming to maximize the transmission throughput is investigated.We formulate the optimization problem by analyzing the time correlation between the access and transmission process within each frame.Simulation results prove the existence of peak value for the optimization problem,thus giving the optimal search time in each sector direction.Also,compared with exhaustive search,the proposed method reveals lower access time delay.2)For the millimeter-wave communication system with hybrid beamforming,we center on the dynamic vehicle-to-everything scenario and propose a multi-user oriented time-varying channel estimation algorithm.The estimation issue is decomposed into a median filtering-based adaptive angle estimation and a complex exponential basis expansion model-based gain estimation.Simulation results demonstrate that the proposed algorithm possesses higher accuracy and can reduce the normalized mean square error of more than 5dB.Further,to reduce the frequent channel re-estimation process,this thesis improves two types of channel tracking algorithms as follows:i)A channel tracking algorithm in view of the time correlation between angles of arrival is presented.Simulation results illustrate that the proposed algorithm yields higher tracking accuracy.When the signal-to-noise ratio is 0dB,the mean square error can reach the order of 10-2.ii)A channel tracking algorithm considering vehicles' moving states is developed.Simulation results show that the proposed algorithm is more capable of capturing the random movement characteristics,which has higher robustness and maintains the tracking accuracy on the order of 10-4.Finally,we summarize the contents of this thesis and give an outlook for future research.