超大规模天线阵列下的多用户快速波束训练

摘要：天线阵列孔径增加提升了波束赋形增益，但更窄的波束使得传统的波束训练方法开销成倍增加，制约了系统支持的用户数目。研究了基于哈希函数的快速、低复杂度新型波束训练方案，可以用较低的训练开销在支持更多用户的同时进行波束训练，在此基础上提出了随机及固定两种码本设计方法，并对波束对齐成功概率进行理论分析。由于推导得到的波束对齐成功概率是一个复杂函数，无法用于指导码本设计，为此，研究提出一个简单的启发式指标用于评估码本生成参数对性能的影响。仿真结果验证了理论分析的正确性，表明所提波束训练方案能够以低训练开销快速实现多用户波束对齐，同时具有高准确性。

关键词：波束训练；多用户；哈希函数；码本设计；性能分析

Abstract：Large aperture arrays can increase beamforming gain, but lead to narrower beams and multiply the traditional beam training overhead, which restricts the number of users served by the system. A fast and low-complexity beam training scheme based on Hash functions is studied, which can support multi-user simultaneous beam training with low overhead. On this basis, two codebook design methods, random and fixed, are proposed.After that, we derive the beam alignment success probability. Since the derived beam alignment success probability is a complex function and cannot be used to guide codebook design, a simple method is proposed to evaluate the impact of codebook generation parameters on performance. The simulation results verify the correctness of the theoretical analysis and show that the proposed beam training scheme can quickly achieve multi-user beam alignment with low training overhead and high accuracy.

Keywords：beam training; multiuser; Hash function; codebook design; performance analysis