Low-Complex Digital Cancellation of Transmitter Harmonics in LTE-A/5G Transceivers

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.