A Feasibility Study of Digital Predistortion (DPD) on the Linearization of Millimeter- Wave Power Amplifiers
This thesis presents two effective Digital Pre-Distortion (DPD) methods that successfully linearized four mm-Wave (Millimeter-Wave) power amplifiers (PA). A state-of-art literature survey of DPD on both the Sub-6 GHz and mm-Wave PAs are discussed. This thesis describes the mm-Wave hardware setup in Dr. Lie’s lab which is used to perform DPD on the mm-Wave PAs presented. For example, complex waveform creation and DPD operation using the NI (National Instruments)-RFmx are discussed. Look-Up Table (LUT) DPD and Generalized Memory Polynomial (GMP) DPD are the DPD methods used in this thesis and the algorithms are discussed. The PAs were tested in both small-signal and large-signal CW (continuous-wave) conditions, and also measured using 64/ 256-QAM (Quadrature Amplitude Modulation) 4G (4th Generation) LTE (Long Term Evolution) modulated signals across a large bandwidth and signal power levels. This thesis also discusses the modulated signal bandwidth’s effects on PA linearity, specifically contrasting the Narrowband (NB) vs. Broadband (BB) signals on the ACLR (Adjacent Channel Leakage Power Ratio) and EVM (error vector magnitude) for RF PAs. The experimental results suggest that DPD can be rather useful ineffectively linearizing mm-Wave PAs to improve their linearity vs output power (POUT) and power efficiency design trade-off. My future work will involve designing and applying DPD on mm-Wave 5G (5th Generation) and 6G (6th Generation)-like broadband PAs in silicon and III-V technologies and realizing highly power-efficient RF (Radio Frequency) circuits and systems for communication and sensing.
Embargo status: Restricted until 09/2028. To request the author grant access, click on the PDF link to the left.