Receiver Side Signal Processing for Nonlinear Distortion Compensation in 5G AND Beyond
Babaroglu, Huseyin (2023)
Babaroglu, Huseyin
2023
Master's Programme in Electrical Engineering
Informaatioteknologian ja viestinnän tiedekunta - Faculty of Information Technology and Communication Sciences
This publication is copyrighted. You may download, display and print it for Your own personal use. Commercial use is prohibited.
Hyväksymispäivämäärä
2023-04-24
Julkaisun pysyvä osoite on
https://urn.fi/URN:NBN:fi:tuni-202304183880
https://urn.fi/URN:NBN:fi:tuni-202304183880
Tiivistelmä
Trading between transmit waveform quality and power efficiency is one of the most challenging issues in radio transmitter implementation. To this end, digital predistortion is the de-facto solution for mitigating power amplifier (PA) nonlinear distortion in cellular base-stations due to its high flexibility and good linearization performance. Theoretically, it is convenient to describe predistorter (PD) transfer function as the mathematical inverse of the PA transfer function, and PD modeling is often performed through parametric methods. Thus, an additional feedback loop is required in the system for PD model parameter estimation. PA is an analog device and DPD is a part of digital front-end, implying that PA output signal is needed to be downconverted to baseband and sampled in the parameter estimation path. Consequently, it is required to employ additional components in the feedback loop such as attenuator, downconverter, and analog-to-digital converter (ADC). In order to be able to capture higher order nonlinearities, it is necessary to perform upsampling operation, which implies that in addition to digital-to-analog converters (DACs) in the forward loop, the components in the feedback loop should support higher bandwidths than the original transmission bandwidth. Additionally, to have a good linearization performance, a high resolution ADC is required. Having an ADC/DAC that supports wide bandwidth and has high resolution is directly increasing the material cost and power consumption. When future millimeter-wave (mmWave) systems are considered, adopting DPD becomes even more complex and costly due to wider waveform bandwidths and employing active antenna arrays.
Alternative to DPD, receiver based approaches, referred to as digital post-distortion (DPoD), can be utilized to mitigate the nonlinear effects of transmitter PA. Naturally, receiver side techniques do not provide any improvement in terms of out-of-band (OOB) emission issues, rather they aim to improve received signal error vector magnitude (EVM). As the radiated power at mmWave is typically EVM limited and OOB emission requirements are relaxed compared to sub-6 GHz band, DPoD can offer means for improved network energy-efficiency. Several iterative DPoD methods are proposed in the literature such as power amplifier nonlinearity cancellation (PANC), and reconstruction of distorted signals (RODS). In this thesis, we present a non-iterative computationally efficient receiver side nonlinearity mitigation technique, referred to as digital post-inverse (DPoI), along with the parameter estimation approach targeting existing 5G NR standard-compliant reference signal. The receiver EVM performance of presented approach is analyzed by using computer simulations. It is seen that DPoI can reach similar or improved performance compared to the iterative PANC method, which is chosen as a reference DPoD method. Moreover, it is shown that both DPoD methods overperform ideally linearized transmitter PA under strong nonlinear conditions, which allows higher power efficiency when receiver side techniques are employed.
Alternative to DPD, receiver based approaches, referred to as digital post-distortion (DPoD), can be utilized to mitigate the nonlinear effects of transmitter PA. Naturally, receiver side techniques do not provide any improvement in terms of out-of-band (OOB) emission issues, rather they aim to improve received signal error vector magnitude (EVM). As the radiated power at mmWave is typically EVM limited and OOB emission requirements are relaxed compared to sub-6 GHz band, DPoD can offer means for improved network energy-efficiency. Several iterative DPoD methods are proposed in the literature such as power amplifier nonlinearity cancellation (PANC), and reconstruction of distorted signals (RODS). In this thesis, we present a non-iterative computationally efficient receiver side nonlinearity mitigation technique, referred to as digital post-inverse (DPoI), along with the parameter estimation approach targeting existing 5G NR standard-compliant reference signal. The receiver EVM performance of presented approach is analyzed by using computer simulations. It is seen that DPoI can reach similar or improved performance compared to the iterative PANC method, which is chosen as a reference DPoD method. Moreover, it is shown that both DPoD methods overperform ideally linearized transmitter PA under strong nonlinear conditions, which allows higher power efficiency when receiver side techniques are employed.