Linearization of High Efficiency Transmitters for Wireless Communications

Doctoral thesis, 2011

In recent years, reducing energy consumption in wireless communication systems has gained prominence. In order to achieve this energy saving, high power efficiency is necessary for the wireless transmitters used. Unfortunately, wireless transmitters usually exhibit severe distortion when acquiring high efficiency, mainly due to the design of power amplifiers. Also, RF modulators, which are building blocks of wireless transmitters, suffer from I/Q imbalance and nonlinear distortion. Consequently, these affect the fidelity of the transmitted signals. In order to obtain distortion-free transmitted signals while maintaining high efficiency, novel linearization approaches based on digital predistortion are proposed in this thesis. In the first part of this thesis, the distortion introduced by an RF modulator is first investigated. A novel nonlinear modeling approach is proposed to compensate for the nonlinear distortion and I/Q imbalance in the RF modulator. Later, this novel approach is applied for the entire transmitter chain to compensate for distortions resulting from the RF modulator together with RF power amplifiers. Compared with the conventional and recently proposed approaches, the approach presented in the thesis has large improvements, which makes it a promising candidate for use in wireless transmitters. The second part of the thesis investigates high-efficiency power amplifier architectures, such as envelope tracking and varactor-based dynamic load modulation, and their linearization issues. These architectures allow an extra degree of freedom to achieve simultaneously low distortion and high efficiency. Different dedicated linearization techniques are proposed for such high-efficiency power amplifier architectures. Practical problems like time-misalignment and bandwidth expansion of envelope signals are considered together with the linearization techniques. Among all the proposed linearization techniques, the dual-input linearization technique has the best linearization performance, giving more than 5 dB improvement in terms of adjacent channel leakage ratio compared with other techniques. It is also very robust, and achieves good results under time-misalignment and reduced-bandwidth envelope signal. The methods proposed in this thesis resolve the efficiency and distortion dilemma of modern wireless transmitters, which, as a result, contributes to better service quality of wireless communications and energy savings.