Skip to main content
QUICK REVIEW

[论文解读] A Scalable and Statistically Robust Beam Alignment Technique for mm-Wave Systems

Xiaoshen Song, Saeid Haghighatshoar|arXiv (Cornell University)|Aug 30, 2017
Millimeter-Wave Propagation and Modeling参考文献 36被引用 4
一句话总结

该论文提出了一种可扩展的、非交互式的毫米波MIMO系统波束赋形(BA)技术,利用二阶信道统计特性和非负最小二乘法(NNLS),实现同时多用户训练。该方法在时变信道下表现出鲁棒性能,且训练开销与用户数量无关,优于现有交互式和基于压缩感知(CS)的先进方法,在多用户场景和动态信道条件下表现更优。

ABSTRACT

Millimeter-Wave (mm-Wave) frequency bands provide an opportunity for much wider channel bandwidth compared with the traditional sub-6 GHz band. Communication at mm-Waves is, however, quite challenging due to the severe propagation path loss. To cope with this problem, directional beamforming both at the Base Station (BS) side and at the user side is necessary in order to establish a strong path conveying enough signal power. Finding such beamforming directions is referred to as the Beam Alignment (BA) and is known to be a challenging problem. This paper presents a new scheme for efficient BA, based on the estimated second order channel statistics. As a result, our proposed algorithm is highly robust to variations of the channel time-dynamics compared with other proposed approaches based on the estimation of the channel coefficients, rather than of their second-order statistics. In the proposed scheme, the BS probes the channel in the Downlink (DL) letting each user to estimate its own path direction. All the users within the BS coverage are trained simultaneously, without requiring "beam refinement" with multiple interactive rounds of Downlink/Uplink (DL/UL) transmissions, as done in other schemes. Thus, the training overhead of the proposed BA scheme is independent of the number of users in the system. We pose the channel estimation at the user side as a Compressed Sensing (CS) of a non-negative signal and use the recently developed Non-Negative Least Squares(NNLS) technique to solve it efficiently. The performance of our proposed algorithm is assessed via computer simulation in a relevant mm-Wave scenario. The results illustrate that our approach is superior to the state-of-the-art BA schemes proposed in the literature in terms of training overhead in multi-user scenarios and robustness to variations in the channel dynamics.

研究动机与目标

  • 解决现有交互式波束赋形(BA)方案在多用户毫米波系统中训练开销高、可扩展性差的问题。
  • 克服基于压缩感知(CS)的BA方法对时不变信道的假设局限,该假设在毫米波频段因高多普勒扩展而通常不成立。
  • 在基站(BS)覆盖范围内实现所有用户的同时波束赋形,无需迭代的下行链路/上行链路反馈轮次。
  • 开发一种基于二阶信道统计而非瞬时信道估计的统计鲁棒BA算法,提升对时变信道动态的鲁棒性。
  • 在采用混合数字-模拟波束成形架构的多用户毫米波MIMO系统中,实现低训练开销和高可靠性。

提出的方法

  • 基站(BS)使用结构化的波束成形序列探测下行链路(DL)信道,使覆盖范围内的所有用户能够同时估计其自身的信道方向。
  • 用户设备(UE)在接收信号上执行压缩感知(CS),将信道建模为到达角(AoA)/离开角(AoD)域中的非负稀疏信号。
  • UE处的信道估计被表述为一个非负最小二乘法(NNLS)问题,可在非负性约束下高效恢复主路径方向。
  • 该算法基于信道的二阶统计特性,因此对信道系数的时间变化具有鲁棒性,而基于瞬时估计的CS方法则不具备此特性。
  • 基站采用一种波束成形策略,结合UE处的全波束赋形增益和基站端的扇区化波束成形,以在初始训练阶段最大化信噪比(SNR)。
  • 训练开销与用户数量无关,因为所有用户均在单次下行链路传输阶段完成训练,无需迭代反馈。

实验结果

研究问题

  • RQ1如何在不依赖迭代下行链路/上行链路反馈轮次的前提下,实现多用户毫米波系统中波束赋形的可扩展性?
  • RQ2与基于瞬时信道估计的压缩感知(CS)BA方法相比,依赖二阶信道统计在多大程度上提升了对时变信道动态的鲁棒性?
  • RQ3非交互式、单阶段训练方案在检测延迟和准确性方面,能否实现与交互式分层或迭代BA方法相当或更优的性能?
  • RQ4在固定系统参数下,UE处最优的射频链路数量是多少,可使波束赋形训练时间最小化?
  • RQ5在信道相关性变化时(如从i.i.d.到高度相关),所提出的基于NNLS的BA方法在检测概率方面与现有CS方法(如OMP)相比如何?

主要发现

  • 所提方案在单次下行链路传输阶段即可完成全部多用户训练,训练开销与活跃用户数量无关,而交互式方法的开销随用户数线性增长。
  • 该方法在时变信道下表现出卓越的鲁棒性,在信道相关系数α(0到1之间)的广泛范围内均保持高检测概率,而文献[15]中的基于CS的方法在中等至高时变条件下失效。
  • 仿真结果表明,所提方案中成功对准用户的比例K(T)/K随时间槽T迅速增加,即使在理想反馈假设下,也优于文献[11]中的交互式二分法。
  • 波束赋形的期望检测时间(以时隙计)在UE处存在最优射频链路数量,仿真显示硬件成本与训练延迟之间存在权衡。
  • 所提NNLS方法的检测概率在信道相关系数α ≥ 0.6时保持在90%以上,而文献[15]中的OMP方法在相同条件下低于50%。
  • 该算法在低基带信噪比(如-33 dB)下仍保持高性能,证明其在射频链路有限的低信噪比毫米波环境中具有有效性。

更好的研究,从现在开始

从阅读论文到最终审阅,大幅缩短您的研究时间。

无需绑定信用卡

本解读由 AI 生成,并经人工编辑审核。