One-Bit Massive MIMO Precoding for Frequency-Selective Fading Channels

One-Bit Massive MIMO Precoding for Frequency-Selective Fading Channels
复制标题

DOI:
10.1109/ssp53291.2023.10207951
复制
发表时间:
2023-03
期刊:
2023 IEEE Statistical Signal Processing Workshop (SSP)
影响因子:
--
通讯作者:
Ly V. Nguyen;Lu Liu;N. Linh-Trung;A. L. Swindlehurst
Ly V. Nguyen;Lu Liu;N. Linh-Trung;A. L. Swindlehurst
中科院分区:
其他
文献类型:
--
作者:
Ly V. Nguyen;Lu Liu;N. Linh-Trung;A. L. Swindlehurst

文献摘要

相似文献

1位数模转换器(DAC)是一种实用且有前途的解决方案,可用于大规模多输入多输出(MIMO)系统中降低成本和功耗。然而,一比特预编码问题是NP难的,并且与平坦衰落场景相比,在频率选择性衰落信道中更具挑战性。在频率选择性衰落信道中,分块处理(BWP)可以有效地解决符号间干扰(ISI)问题,但其计算复杂度和处理延迟对于实际实现来说太高。缓解处理复杂性和延迟问题的替代解决方案是按符号处理(SWP),其顺序地设计发射信号。然而,现有的SWP工作为以后的信号设计留下了不必要的干扰。在本文中,我们提出了一个SWP的方法,可以有效地解决ISI,即使在符号率。其思想是将发射信号设计为不仅对其时隙有益,而且还为后续符号提供相长干扰。我们开发了两种主动ISI处理方法,其性能显著优于传统方法,其中之一甚至在低SNR下优于BWP方法。
One-bit digital-to-analog converters (DACs) are a practical and promising solution for reducing cost and power consumption in massive multiple-input multiple-output (MIMO) systems. However, the one-bit precoding problem is NP-hard and even more challenging in frequency-selective fading channels compared to the flat-fading scenario. While block-wise processing (BWP) can effectively address the inter-symbol-interference (ISI) in frequency-selective fading channels, its computational complexity and processing delay can be too high for practical implementation. An alternative solution to alleviate the processing complexity and delay issues is symbol-wise processing (SWP) which sequentially designs the transmit signals. However, existing SWP work leaves unwanted interference for later signal designs. In this paper, we propose an SWP approach which can efficiently address the ISI even at the symbol rate. The idea is to design the transmit signal to not only be beneficial for its time slot, but also to provide constructive interference for subsequent symbols. We develop two active ISI processing methods that significantly outperform a conventional approach, one of which that even outperforms the BWP approach at low SNR.