Advanced Signal Processing for High-speed Optical Fiber Systems

已安装的基于多模光纤的以太网骨干网大多工作在1 Gb/S左右的比特率，不能满足当前和新兴的需求。随着高速(40 Gb/S及以上)光纤传输系统的出现，这类系统的先进信号处理和编码方案应该仔细设计，并有原则地进行优化。传统上，光学系统只执行最少量的信号处理。由于技术的进步，今天在光学系统中执行复杂的信号处理功能是可能的。这可以带来巨大的灵活性，因为有机会使用在光学领域中不实用的低成本和自适应信号处理电路。该项目致力于开发高速光纤系统的新设计范例和先进算法。特别是，它包括对用于超高速光通信的编码调制方案、高速解码算法和高效多载波传输技术的设计的深入研究。将探索大量的数字和统计信号处理工具，包括陪集码、稠密格子、框架理论和蒙特卡罗信号处理。除了进行理论分析外，还将开发计算程序以促进分析工作。在理想条件下开发的新概念和算法将被量身定做，以适用于具有各种约束的实际系统。预计这项研究不仅将加深我们对复杂光纤系统基本原理的理解，发展新的信号处理概念和方法，而且还将为未来高速光学系统的物理层解决方案提供新的强大工具。

Most installed Ethernet backbones based on multi-mode fibers operate at bit rate of about 1Gb/s, which is inadequate for current and emerging demands. With the advent of high-speed (40 Gb/s and beyond) fiber optic transmission systems, advanced signal processing and coding schemes for such systems should be carefully designed and optimized in a principled way. Traditionally, optical systems perform a minimal amount of signal processing. Owing to advances in technology, today it is possible to perform sophisticated signal processing functions in optical systems. This could bring enormous flexibility resulting from the opportunity to employ low-cost and adaptive signal processing circuits that is not practical in the optical domain. This project focuses on the development of new design paradigm and advanced algorithms for high-speed fiber optic systems. In particular, it encompasses in-depth investigations on the design of coded modulation schemes, high-speed decoding algorithms, and highly efficient multi-carrier transmission techniques, for very-high speed optical communications. A large arsenal of digital and statistical signal processing tools will be explored, including coset codes, dense lattices, frame theory, and Monte Carlo signal processing. In addition to conducting theoretical analysis, computational procedures will be developed to facilitate the analytical work. The new concepts and algorithms developed under ideal conditions will be tailored to operate in practical systems with various constraints. It is expected that the proposed research will not only enhance our understanding of the fundamental underpinnings of complex fiber optic systems and develop novel signal processing concepts and methods, but also produce new and powerful tools for physical layer solutions for future high-speed optical systems.