Nowadays, mobile communication radio frequency transceivers for LTE-A/5G usually operate in frequency-division duplex mode and support carrier aggregation. The resulting complexity requires compromises in the analog front-end. One consequence is a limited stop-band attenuation of the duplex filters that separate the transmit and the receive paths. The resulting transmit leakage signal combined with other non-idealities, like power amplifier (PA) harmonics, can create self-interference problems that desensitize the receivers. In this work, we propose an optimized digital self-interference cancellation scheme to mitigate the unacceptable performance degradation caused by PA induced harmonics. We derive an appropriate interference model and quantify the interference power levels for a specific PA. Based on these results, our low-complex cancellation approach solely targets the dominating components. Our algorithm relies on the auto-covariance matrix for the second and third power of LTE signals, which we derive for the first time in an analytic form. This statistical information can be combined with the transform-domain concept to noticeably improve the performance of a least-mean squares based cancellation algorithm. The capabilities of the approach are demonstrated by means of simulations and are validated on measurement data.

中文翻译：

---

LTE-A/5G 收发器中发射机谐波的低复杂度数字消除


目前，LTE-A/5G的移动通信射频收发器通常工作在频分双工模式下，并支持载波聚合。由此产生的复杂性需要在模拟前端做出妥协。结果之一是分离发射和接收路径的双工滤波器的阻带衰减有限。由此产生的发射泄漏信号与功率放大器 (PA) 谐波等其他非理想信号相结合，可能会产生自干扰问题，从而降低接收器的灵敏度。在这项工作中，我们提出了一种优化的数字自干扰消除方案，以减轻 PA 感应谐波引起的不可接受的性能下降。我们推导出适当的干扰模型并量化特定 PA 的干扰功率水平。根据这些结果，我们的低复杂度消除方法仅针对主要组件。我们的算法依赖于 LTE 信号的二次方和三次方的自协方差矩阵，这是我们首次以解析形式推导出来的。该统计信息可以与变换域概念相结合，以显着提高基于最小均方的消除算法的性能。该方法的功能通过模拟进行了演示，并通过测量数据进行了验证。