Design Of Digital Baseband Predistortion For High Power Amplifier

With the rapid development of wireless communication, in order to make more efficient use of frequency spectrum and achieve greater power efficiency, the linear modulation methods and high-efficiency radio frequency power amplifiers are widely used. Since the envelope of modulated signal fluctuates, unwanted inter-modulation distortions generated when the signal is amplified by the nonlinear RF power amplifiers. The inter-modulation distortions will lead to severely inter-symbol interference and adjacent channel interference. Therefore, high linearity power amplifier is used to ensure the communication quality, but the high power amplifier (HPA) usually works at nonlinearity status which brings nonlinearity distortion. In order to solve this problem, it should improve the nonlinearity of high power amplifier by using the linearization techniques.High-Power amplifier (HPA) linearity and power efficiency are critical requirements in modern wireless communications systems. This report is oriented to the group of C type, which exhibits varying degrees of nonlinearity when operated at high-input drive levels. Such nonlinear effects cause in-band distortions (IBD) and out-of-band power emissions (OPE). The IBD results in degraded system performance, while the OPE generates spectral re-growth, which in turn leads to adjacent channel interference (ACI). One of the best ways to suppress such nonlinearities is to use linearization techniques to make a balance between PA linearization and high effects.The digital baseband predistortion being at the center of interest of this paper is one of the possible methods to compensate for HPA nonlinearities. The principle of predistortion is to distort the HPA input signal by an additional device called a predistorter whose characteristics (due to the temperature, aging,etc.). It is desirable to make a predistortion adaptive. That's why we are focusing on the adaptation of predistorter characteristic.The aim of this partnership is to study and improve adaptive filter-based digital baseband predistortion schemes using square-root-and-division-free algorithms such as dichotomous coordinate descent algorithms. The report mainly discusses pre-distortion technology, which is a most excellent, easiest to implement and has greatest development potential linearization technique, could better overcome the nonlinearity and memory effect of power amplifier. Then the report introduces several classic pre-distortion algorithms. On the basis of that, it proposes an improved iteration algorithm, dichotomous coordinate descent (DCD). The DCD algorithm is based on polynomial model. It uses simple polynomial pre-distortion model to achieve better linearization performances by using simpler operations and requiring less resources.The paper is structured in 4 chapters.The first chapter introduces the background and significance of the project, along with the research achivements and documents summarization, as well as the contents.As the good choice of the power amplifier model is a crucial task for the verification of developed predistortion algorithms, the first chapter's beginning is focused on the AM/AM, AM/PM characteristics, Error Vector Magnitude (EVM), Adjacent Channel Power Ratio (ACPR) and Signal Noise Ratio (SNR), which is used to quantify the effct of HPA nonlinearity. Several behavior models and linearization techniques are also presented in this chapter, the digital baseband predistortion with more details.The third chapter focus on the mail algorithms for predistortion. The improved DCD algorithm is simulated on MATLAB which certify the improved algorithm's efficient and quick convergence.Finally, the report proposes the design and schema for the system on FPGA.