Dmytro Bondar

Digital baseband predistortion of wideband power amplifiers with improved memory effects

A new digital baseband predistortion of wideband power amplifiers with improved memory effects is presented and investigated. Distortion compensation is increased by implementing proposed iterative digital baseband injection technique. Frequency-domain equalization in baseband is developed to minimize memory effects. The non-linear and memory parts of the proposed linearizer are implemented in baseband without additional feedback loop. The spectral regrowth improvement of 20 dB for a 5-MHz 16-QAM signal and a 10 dB reduction of 8 MHz offset adjacent channel power ratio (ACPR) are achieved. Experimental results are presented to validate the argument.

Linearization of high-efficiency power amplifiers using digital baseband predistortion with iterative injection

This paper presents the linearization of a high-efficiency compressed power amplifier (PA) by iterative digital predistortion with suppression of memory effects. Iterative injection of the odd-order in-band distortion components is used to enhance the nonlinearity compensation and achieve the optimal ratio of linearization degree to computational complexity by adjusting the number of injections. Baseband equalization of the input signal minimizes memory effects. The proposed linearizer is designed within the baseband block, and does not require additional radio-frequency components. The experimental investigation is carried out with EDGE signals for an up-link PA. Up to 18 dB improvement in spectral re-growth and 10 dB reduction in a 500-kHz offset adjacent channel power ratio (ACPR) are achieved for 200-kHz EDGE signal at 880 MHz.

Distortion improvement of power amplifiers with digital predistortion

This paper presents an optimized digital predistortion (DPD) technique based on iterative injecting of distortion components into the baseband signal. The ratio of computation complexity to linearization degree can be regulated by choosing the appropriate number of injections. The proposed DPD technique allows achieving a significant increase in linearization degree and overcoming the distortion compensation limit. Experimental results for a 21-dB gain GaAs power amplifier fed with a 5-MHz digitally modulated signal show significant improvement in spectral regrowth.

A new approach for non-linear analysis of power amplifiers

This paper presents a new approach for non-linear analysis of power amplifiers based on the n-order polynomial modelling. A general expression for the fundamental frequency output is developed in this paper. It allows considering any order of polynomial model and obtaining analytical expression for the fundamental-frequency output without complex computations, using developed formulas. This model significantly simplifies non-linear analysis of power amplifiers and the process of designing linearizers. This paper also presents a simple approach for obtaining polynomial model from a frequency response of power amplifier. The proposed method is verified by experiments. Obtained polynomial model shows simulation results similar to the measurements for real amplifier. Developed n-order output modelling and proposed model extraction method can significantly simplify, optimize and enhance power amplifierspsila non-linear analysis and behavioural modelling. Proposed modelling approach can be also used to design linearizers.

Compensation of nonlinear distortion in RF power amplifiers for wireless communications

This paper presents a new, simple and efficient technique for linearizing power amplifiers by combining iterative digital predistortion and baseband equalization. Iterative injection of in-band distortion components at baseband is applied to improve the performance of digital predistortion, whereas the equalization technique is used to minimize memory effects. The digital signal processing (DSP) algorithm is experimentally tested for a 5-MHz QPSK-modulated signal on a 751 MHz carrier amplified by a highly-efficient and highly nonlinear power amplifier. Up to 16 dB improvement in spectral re-growth is achieved by the proposed digital predistortion with 53% efficiency and with no changes to RF or modulation hardware.

Minimizing memory effects in OFDM transmitters using adaptive baseband equalization

This paper presents a simple and effective approach for eliminating memory effects in OFDM transmitters. It uses advantages of OFDM systems to provide pre-compensation of the frequency-dependent distortions, which are results of the power amplifiers (PA) memory effects. The process of memory effects quantification is carried out in this paper by obtaining a frequency-dependent PA gain, phase shift and intermodulation products. The memory effects are eliminated at baseband using equalization of the IDFT signal. Implementation of the equalization procedure at baseband makes the process of minimizing memory effects simple and effective, because no additional RF components or feedback loops are used. Memory effects are compensated in DSP part using simple multiplication of the frequency-domain digital signal by coefficients, which are calculated adaptively for each OFDM sub-carrier frequency and input power. The approach is tested with Motorola MOSFET MRF9742 power amplifier model in Advanced Design System (ADS). Simulations show significant improvement in minimizing memory effects. Received constellation of the 16-QAM OFDM signal after implementing baseband pre-compensation technique looks alike ideal one, whereas without pre-compensation it shows high dispersion due to the presence of PA memory.

Linearization of power amplifiers by baseband digital predistortion for OFDM transmitters

For the signals with high peak-to-average ratio (PAR), such as WiMax OFDM, power amplifiers need to operate in a compression mode. It results in a nonlinear distortion of the output signal. To compensate for this distortion, linearizers are used. However, they decrease efficiency of amplifiers. This paper proposes to use a digital predistortion of baseband signals, which is characterized by a small impact on efficiency and a good linearizing performance, for linearization of OFDM transmitters. The paper describes design of a baseband predistorter and verifies its performances by WiMax OFDM simulations in ADS. Proposed predistorter brings significant improvements in AM/AM characteristic, eliminates spectrum re-growth, decreases adjacent channel power ratio (ACPR) and error vector magnitude (EVM).

Spectrum regrowth reduction in power amplifiers using digital baseband signal injection

This paper addresses the issue of improving distortion compensation in a memoryless digital predistorter. It presents a developed iterative baseband injecting method aimed at overcoming the distortion compensation limit. The concept of the proposed technique is described in the following section. Further section presents experimental verification of the considered PA modelling technique and results of Matlab-ADS co-simulations for demonstrating performances of the proposed predistortion approach.