基于亚频带采样的融合ROF链路和射频功放的多维记忆型线性化技术研究

包含ROF链路和功放的ROF前传系统（Front Haul Systems）存在严重的多频带互作用动态非线性，因此，ROF前传系统动态非线性抑制成为亟待解决的问题。传统的非线性化抑制方法是采用模拟预失真或数字预失真分别对ROF前传系统中的ROF链路和功率放大器分别进行线性化,且对实时大动态无线信号引起的非线性抑制较差。模拟预失真带宽宽，但忽略记忆效应，抑制能力有限；数字预失真受A/D转换器速率的限制，难以直接应用于宽带的ROF前传系统线性化。本项目提出将ROF前传系统的ROF链路和功放融合成一个整体进行系统级线性化研究，在理解ROF前传系统非线性的多频带互作用机理上，应用亚频带采样，多维构建能精确描述动态信号非线性和多频带互作用动态非线性的记忆型非线性模型，提出一种能够消除多频带互作用动态非线性的多维记忆型数字和模拟混合宽带预失真新技术，降低A/D采样率，为建立宽带超线性的ROF前传系统奠定基础。

There are severe interactive dynamic nonlinearities between multi-frequency bands in front-haul systems that consist of radio over fiber (ROF) links followed by RF power amplifiers. Therefore, it is a pivotal technical challenge to suppress the above nonlinearities in the ROF front-haul systems. Conventionally, suppression of nonlinearities in separate ROF links and RF power amplifiers is completed using either analog predistortion or digital predistortion techniques， and also it is difficult for the conventional linearization techniques to be used for real-time large dynamic wireless signals induced nonlinearities. The analog predistortion techniques can be wideband, but are limited in linearization efficiency due to the memory effects ignored. The digital predistortion techniques are difficult to be used in broadband ROF front haul systems, limited by A/D conversion rate. In this project, it is proposed to linearize the ROF front haul systems by combining the linearization of the ROF links with the RF power amplifiers, instead of linearization of ROF links and RF power amplifiers separately. By understanding the mechanisms of interactive dynamic nonlinearities between multi-frequency bands in the ROF front-haul systems, using sub-frequency band sampling, a nonlinear model will be established, which can be used for accurate modeling of dynamic signal nonlinearities, interactive nonlinearities between multi-frequency bands, and multi-dimensional memory effects of the multi-frequency bands. Eventually, based on sub-frequency band sampling, an analog and a digital combined predistortion technique including multi-dimensional memory effects will be proposed and verified, to suppress the above nonlinearities and thereby the requirements for A/D sampling rate reduced. Consequently, this project will pave a way for achieving super-high linearity’s ROF front-haul transmission systems.