Design Of Terahertz Array Beamforming Based On Low-Bit Phase Shifters

Terahertz(THz)communication with frequencies ranging from 0.1 to 10 THz is considered a promising solution for high-speed wireless communication in the future.While THz signals have usable bandwidths of tens of gigahertz,they suffer from severe free-space diffusion and molecular absorption losses,which limit wireless transmission distances.Beamforming technology can generate high-gain directional beams in multiple-input and multiple-output(MIMO)systems to compensate for transmission losses.In THz communication scenarios,the small size of THz arrays and the integration of a large number of antenna elements require reducing hardware costs further by using hybrid beamforming structures or fully analog beamforming structures commonly used in millimeter wave(mm Wave)communication scenarios.Currently,a feasible solution widely recognized in the industry is to implement beamforming using low-bit shifters that are small in size,low in power consumption,and high in integration.However,most existing beamforming technologies focus on fully digital beamforming structures,and the resulting fully digital beamforming vectors do not constitute feasible solutions for analog beamforming using low-bit shifters.In light of this,thesis proposes designs suitable for low-bit analog beamforming for three specific THz communication scenarios to meet the requirements of different communication systems.Specifically,the thesis studies wide-coverage beam design,multicast beam design,and sidelobe suppression beam design for signal broadcasting,signal multicasting for multiple users,and interference suppression for multiple users,respectively.The existing fully digital beamforming design methods,such as the complex target tracking method,window function method,and Chebyshev optimal uniform approximation method,are extended to the scenarios considered in the thesis.Since the feasible domain of low-bit analog beamforming is a subset of the feasible domain of fully digital beamforming,these fully digital beamforming solutions serve as performance upper bounds.In addition,the thesis obtains feasible solutions by directly projecting the solution of the fully digital beamformer onto the feasible domain of the low-bit analog beamforming using a simple projection method.Since this method results in serious performance losses,it is used as the performance benchmark for simulation comparison.To obtain solutions superior to the benchmark,the thesis studies the optimization problems of low-bit analog beamforming in these three scenarios.The coordinate ascent optimization method is used in wide-coverage beam design,the cross-entropy optimization method is used in multicast beam design,and the greedy algorithm is used in sidelobe suppression beam design.Simulation results show that the optimization methods proposed in the thesis for low-bit analog beamforming outperform the performance benchmark in terms of performance.Finally,the thesis introduces the challenges and future research directions of THz MIMO systems in beamforming technology.