Signal Shaping and Sampling-based Measurement Techniques for Improved Radio Frequency Systems

Abstract: Wireless communication systems are omnipresent in our day-to-day life, with high expectations regarding capacity, reliability and power efficiency. In order to satisfy the capacity and reliability expectations, today's wireless systems are adopting sophisticated modulation schemes, such as orthogonal frequency division multiplexing (OFDM), which shape today's wireless signals with large bandwidths and high crest factors. On top of that, it is anticipated that different wireless systems/standards will co-exist and share the same radio frequency (RF) front-end in order to reduce the network implementation cost.  Such signals characteristics and systems coexistence put high requirements on the amplification stage which in best scenarios is considered weakly nonlinear. As a result, the power amplifier needs to be backed-off for linear operation. However, such power back-off reduces the operation's efficiency. Reducing the crest factor of the wireless signal and the possibility to linearize by means of digital pre-distortion the operation behavior of the power amplifier when operated near its maximum allowed continuous wave (CW) operating power range would lead to the optimal linearity and efficiency of operation.  In order to achieve a good linearization performance, accurate baseband behavioral models are needed which requires measuring time domain signals whose spectra spread largely due to the nonlinear operation of the power amplifier. Such spectrum spreading, denoted by spectral regrowth, puts high requirements on today's sampling-based measurement systems as a trade-of between the sampling rate and amplitude resolution exists in today's generation of analog-to-digital converters, in addition to a limitation in the available analog bandwidth. Overcoming such measurement challenges could lead to the design of expensive measurement systems which is not favorable.  In this thesis, the performance of RF transmitters is improved by combining the use of a smart crest factor reduction technique with an enhanced digital pre-distortion technique which allows operating the power amplifier near its CW 1-dB compression point, offering a significant increase in the efficiency of operation while satisfying the standard constraints on information error and spectral emission. Furthermore, the performance of RF measurement receivers is improved by reducing the requirements on the digital bandwidth by means of an evolved harmonic sampling technique, and by reducing the requirements on the analog bandwidth and design cost by means of a digital bandwidth interleaving technique and a signal separation technique based on an advanced sparse reconstruction methodology.