Björn Gävert, Kommunikation, Antenner och Optiska Nätverk

There is an ever-increasing demand for higher data rates in wireless communication systems for supporting, for example, the growing consumption of video streaming in today’s networks, and extremely data-intensive applications such as immersive communications in future networks. However, wireless communication equipment providing higher rates comes with more stringent hardware requirements. A large contributor to distortion is the phase noise related to up- and down-converting of signals to and from radio frequency. The main focus of this thesis is related to handling the carrier phase noise, balancing complexity and performance. For this, different alternatives for close-to-optimal low complexity pilot-based phase noise estimation methods are studied.

First, a seemingly simple approach is examined in Paper A, where a pilot tone in the form of a sinusoid is added to an unknown communication signal. The power of the transmitted signal is shared between the pilot tone and the communication signal. Several pilot-based phase noise estimation algorithms are studied, and an optimal signal-to-pilot power ratio is presented.

Second, pilot-based phase noise estimation of an OFDM signal affected by phase noise is studied in Paper B. The known pilots are allocated in the frequency domain and occupy a number of subcarriers within the OFDM symbols. A novel, low-complexity pilot-based phase estimator is proposed, which is close to optimal over a wide dynamic range of phase noise levels and signal-to-noise ratios.

Finally, Paper C examines the maximum rate of a line-of-sight MIMO system affected by phase noise, where known pilots are interleaved in the time domain together with the payload. To maximize the rate, a novel pilot-based estimator is proposed that jointly estimates both payload and phase noise.