SOFTWARE-DEFINED-SOLUTIONS FOR NONLINEARITIES IN SUB 6 GHZ TO mmWAVE TRANSMITTERS

Abstract

The software-defined radios (SDRs) have changed radio engineering’s landscape due to the flexibility provided by its programmable features. The programming functionality can control digital signal processing, general-purpose processors, and field-programmable gate arrays. Moreover, due to the rapid growth in miniaturization technology, very compact SDRs have been developed. The availability of flexible hardware architectures and adaptability in software architectures provides additional capability to SDR to interact with other devices, networks, as well as data and air interfaces. The SDRs have the provision of tuning several parameters such as sampling rate, bandwidth, transmitter frequency, modulation type, and encryption/security, automatic gain control. Among the various SDR architectures, the homodyne architecture is a prominent architecture due to its smaller size and low-cost implementation. However, it suffers from several inherent limitations such as distortion due to DC offset, the local oscillator (LO) leakage, and in-phase/quadrature-phase (I/Q) imbalance. Moreover, to provide a higher data rate, communication technologies like long term evolution (LTE) and LTE-advanced are being used by incorporating higher quadrature amplitude modulation schemes. However, such a signal imposes a strict linearity requirement on the radio-frequency (RF) power amplifier (PA) due to the high peak-to-average power. Therefore, it requires operating the PA in the saturation region for the highest efficiency, but it produces unwanted nonlinear distortions; hence it is essential to compensate for such distortions by the well-established linearization technique. The predistortion (PD) is an important linearization technique that is very useful for the compensation of the nonlinearity of the PA. But the PD has the challenge of complexity, bandwidth requirement, along with transceiver imperfections. This thesis investigates the limitations of the available components, algorithms, and SDRs for the system-level integration for the experimental testbed development, which will be used throughout this work. This investigation of the characterization setup allows focusing on system-level imperfection so that a testbed can be developed. Further, to enhance the numerical stability of the state-of-art models in the presence of transceiver noise, a wavelet multi-scale principal component analysis based technique is investigated as a solution to the numerical stability problem. Moreover, to consider the overall imperfection on a very low-cost setup to enhance digital predistortion (DPD) performance, a novel post compensation method is developed.