Analysis and Augmented Spatial Processing for Uplink OFDMA MU-MIMO Receiver With Transceiver I/Q Imbalance and External Interference
Analysis and Augmented Spatial Processing for Uplink OFDMA MU-MIMO Receiver With Transceiver I/Q Imbalance and External Interference
复制标题
DOI：
10.1109/twc.2016.2521382
复制
发表时间：
2014-07
影响因子：
10.4
通讯作者：
Aki Hakkarainen;J. Werner;K. Dandekar;M. Valkama
中科院分区：
文献类型：
--
作者：
Aki Hakkarainen;J. Werner;K. Dandekar;M. Valkama
本文研究了多用户多输入多输出（MU-MIMO）系统中接收机（RX）信号的处理。我们重点研究了相关射频电子学中相位/正交（I/Q）不平衡下的上行正交频分多址（OFDMA） MU-MIMO通信。已有文献表明，收发器I/Q失衡会导致OFDM系统中镜像子载波的串扰。与通常报道的单用户研究相反，我们将研究扩展到基于ofdma的MU-MIMO通信，在频率和空间域同时进行用户多路复用，并在RX输入处纳入来自多个源的外部干扰，以模拟日益流行的异构网络中具有挑战性的条件。在信号处理的发展中，我们利用增强子载波处理，它将每个子载波与图像子载波上的对应子载波联合处理，并在所有RX天线上联合处理。此外，我们从最小化均方误差的角度推导出最优增广线性RX。该方法将多个终端的I/Q失衡缓解、外部干扰抑制和数据流分离集成到一个处理阶段，从而避免了单独的收发器校准。广泛的分析和数值结果表明，任意数据流经过RX空间处理后的信噪比（SINR）和符号误差率（SER）行为是不同系统和损伤参数的函数。结果表明，在收发器I/Q不平衡的情况下，传统的单副载波处理的性能受到严重限制，并且对外部干扰特别敏感，而本文提出的增强副载波处理提供了一种高性能的信号处理解决方案，能够检测不同用户的信号并有效地抑制外部干扰。最后，我们还将研究扩展到具有非常大的天线系统的大规模MIMO框架。结果表明，尽管RX天线数量巨大，但传统的线性处理方法仍然受到I/Q不平衡的严重影响，而增强方法则没有这种限制。
This paper addresses receiver (RX) signal processing in multiuser multiple-input multiple-output (MU-MIMO) systems. We focus on uplink orthogonal frequency-division multiple access (OFDMA)-based MU-MIMO communications under in-phase/quadrature (I/Q) imbalance in the associated radio frequency electronics. It is shown in the existing literature that transceiver I/Q imbalances cause cross-talk of mirror-subcarriers in OFDM systems. As opposed to typically reported single-user studies, we extend the studies to OFDMA-based MU-MIMO communications, with simultaneous user multiplexing in both frequency and spatial domains, and incorporate also external interference from multiple sources at RX input, for modeling challenging conditions in increasingly popular heterogeneous networks. In the signal processing developments, we exploit the augmented subcarrier processing, which processes each subcarrier jointly with its counterpart at the image subcarrier, and jointly across all RX antennas. Furthermore, we derive an optimal augmented linear RX in terms of minimizing the mean-squared error. The novel approach integrates the I/Q imbalance mitigation, external interference suppression, and data stream separation of multiple UEs into a single processing stage, thus avoiding separate transceiver calibration. Extensive analysis and numerical results show the signal-to-interference-plus-noise ratio (SINR) and symbol-error rate (SER) behavior of an arbitrary data stream after RX spatial processing as a function of different system and impairment parameters. Based on the results, the performance of the conventional per-subcarrier processing is heavily limited under transceiver I/Q imbalances, and is particularly sensitive to external interferers, whereas the proposed augmented subcarrier processing provides a high-performance signal processing solution being able to detect the signals of different users as well as suppress the external interference efficiently. Finally, we also extend the studies to massive MIMO framework, with very large antenna systems. It is shown that, despite the huge number of RX antennas, the conventional linear processing methods still suffer heavily from I/Q imbalances while the augmented approach does not have such limitations.