Effects of hardware and design parameters of adaptive digital predistortion on system performance

Due to the increasing need for data rate and the limit of bandwidth in communication systems, nowadays, usage of high order modulation types has become essential. Communication systems, ever increasing long distance communication as well as satisfy less error rates in order to reach more output power levels at transmitter outputs. The necessity of reaching higher power levels has led to development of power amplifiers operating at compression and saturation region. Power amplifiers present nonlinear input-output behavior as the power level applied to the input of power amplifiers increases. The nonlinear effects of power amplifiers degrade the quality of communication systems. Signals with high order modulation are highly affected by nonlinear power amplifier behaviour because of their higher peak to average power ratios. To this end, a trade-off should be obtained between efficiency and linearity for power amplifiers. At this point, linearization techniques help reduce the distortions arising from the nonlinear effects of signals applied to the system input. This can be accomplished by extending the linear range of power amplifier. Adaptive predistorters are linearization systems with RF feedback path. They are flexible systems as they do not require a power amplifier model and they are implemented on digital domain. Nonlinear systems increase bandwidth of input signal. For this reason, the performance of linearization system reduces if the bandwidth of feedback path is not large enough. Since the signal, which is to be digitized on the feedback path, has an increased bandwidth, analog to digital converters with high sampling rate are required. The system performance gets worse if sampling rate of analog to digital converter is high enough. This thesis analyses the effects of analog to digital converter sampling rate and bandwidth of feedback path on the performance of adaptive digital predistortion systems.