标题： 基于信道预测的大量MIMO信道状态信息反馈：如何在用户设备上避免使用机器学习？ 显示英文标题

标题： Massive MIMO CSI Feedback using Channel Prediction: How to Avoid Machine Learning at UE?

摘要： 在文献中，机器学习（ML）已被应用于基站（BS）和用户设备（UE）以提高下行信道状态信息（CSI）的精度。 然而，由于UE功耗等原因，ML在UE上的实现可能不可行。 受此问题的启发，我们在BS提出了一种CSI学习机制，称为CSILaBS，以避免在UE上使用ML。 为此，通过利用BS上的信道预测器（CP），考虑了一个轻量级预测函数（PF），用于UE上的反馈评估。 CSILaBS减少了空中反馈开销，提高了CSI质量，并降低了UE的计算成本。 此外，在多用户环境中，我们提出了各种机制，通过利用PF选择反馈，旨在提高CSI准确性。 我们还讨论了各种基于ML的CP，例如 NeuralProphet（NP），一种受ML启发的统计算法。 此外，受到同时使用统计模型和ML的启发，我们提出了一种由循环神经网络和NP组成的新型混合框架，其预测精度优于单独模型。 CSILaBS的性能通过在诺基亚贝尔实验室记录的经验数据集进行了评估。 结果表明，消除UE上的ML可以保留性能优势，例如预编码质量。 显示英文摘要

摘要： In the literature, machine learning (ML) has been implemented at the base station (BS) and user equipment (UE) to improve the precision of downlink channel state information (CSI). However, ML implementation at the UE can be infeasible for various reasons, such as UE power consumption. Motivated by this issue, we propose a CSI learning mechanism at BS, called CSILaBS, to avoid ML at UE. To this end, by exploiting channel predictor (CP) at BS, a light-weight predictor function (PF) is considered for feedback evaluation at the UE. CSILaBS reduces over-the-air feedback overhead, improves CSI quality, and lowers the computation cost of UE. Besides, in a multiuser environment, we propose various mechanisms to select the feedback by exploiting PF while aiming to improve CSI accuracy. We also address various ML-based CPs, such as NeuralProphet (NP), an ML-inspired statistical algorithm. Furthermore, inspired to use a statistical model and ML together, we propose a novel hybrid framework composed of a recurrent neural network and NP, which yields better prediction accuracy than individual models. The performance of CSILaBS is evaluated through an empirical dataset recorded at Nokia Bell-Labs. The outcomes show that ML elimination at UE can retain performance gains, for example, precoding quality.