Encoding and modeling techniques for center frequency agile digital RF transmitters

Abstract

Current and the future radio frequency (RF) transmitters have to cope with growing data rates and support simultaneously numerous wireless standards. To this end, one of the increasingly desired features is the ability to adjust the center frequency of the transmission signal over a wide bandwidth setting. This capability leads the way to flexible RF front-ends supporting multi-mode and multi-channel operation. Advanced, quadrature type digital transmitter concepts combine re-configurable RF signal generation with high efficiency power amplification. However, most of the state-of-the-art transmitter solutions are optimized for limited center frequency settings. Therefore, in this thesis, novel pulsed encoding and behavioral modeling techniques are presented to improve the center frequency agility. Firstly, a novel way of digital pulse-width modulation (PWM) of quadrature sequences is given. With suitable noise shaped encoding methods, the distortion due to aliasing can be shifted away from the signal band. Thus, improved transmission signal dynamic range can be obtained for a large scale of intermediate frequency settings. Secondly, the inherent quadrature imbalance related to the digital up-conversion technique is analyzed and suitable compensation techniques are presented. The presented quadrature noise shaped encoding algorithms enable suppression of the conjugate quantization noise and conjugate image components in the signal band of up to 50 dB. Thus, digital encoding of RF pulsed sequences with improved center frequency tuning can be enabled. Thirdly, two complex baseband behavioral models, which capture the analog circuitry based non-linear effects of the transmitter are introduced. According to the simulation and measurement based validation results, the models are effective for a multitude of center frequency settings.