Model-Based Signal Processing for Digital Communications

PIS最近开发了一种使用基于模型的逐符号最大后验概率(MAP)算法的数字通信信号解调范例。同时，我们还开发了一个实验物理层无线调制解调器，该调制解调器实现了通信理论中的许多先进算法。这个项目利用这些过去的发展来产生高质量的研究：-为重要的无线数据通信应用开发新的算法；-产生分析这些算法的性能和稳定性的技术；-在现实的无线信道上部署和试验新的算法，以验证理论并制定更准确的分析信号模型。逐符号地图解调在各种数字通信应用中非常重要。Turbo码由于其接近的容量性能和相对简单的解调结构，最近引起了人们的极大兴趣。MAP译码是Turbo译码中非常重要的一环，本项目将原则性地探索降低标准Turbo译码复杂度的方法，并探索基于逼近最优MAP译码的Turbo译码的非迭代技术。此外，在无线通信中常用的交织编码系统中，MAP解调非常重要。本项目探索用于时变频率平坦和频率选择性衰落信道的软输出算法。特别是，我们正在探索用于IS-136等时分多址应用和数字电视的高性能解调器。在高容量应用中，干扰抑制的重要性大大增加。该项目将扩展基于模型的MAP范例，以开发在存在同频道干扰的情况下使用传统的单天线调制解调器和“时空”调制解调器架构的高性能信号解调器。我们项目的另一个独特之处是，我们将软输出解调/解码与接收器前端设计(如均衡、同步)集成在一个框架中。这种方法对于以利用强大的码(例如，Turbo码)所需的低信噪比进行操作是必不可少的。这个项目的另一个独特方面是，我们打算以比传统尝试更深入的方式研究产生的算法。这个项目将应用复杂的分析技术来分析许多提议的算法的性能和稳定性。这种分析将提供对这些算法的性能的更好的洞察，而不是仅仅通过模拟研究来实现。我们还打算实现和现场测试许多提出的算法。我们目前有一个工作正常的无线调制解调器，通过简单的软件更改就可以添加新的算法。算法开发、算法分析和算法实现的这种组合将导致对通信理论的重大新见解，特别是在无线领域。这种结合的方法还将为学生参与者提供更全面的教育体验。

A demodulation paradigm using model-based symbol-by-symbol maximum a posteriori (MAP) algorithms for digital communication signals has recently been developed by the PIs. In parallel, an experimental physical layer wireless modem has been developed that implements many advanced algorithms from communication theory. This project uses these past developments to produce high quality research to: - Develop new algorithms for important wireless data communication applications; - Produce techniques to analyze the performance and stability of these algorithms,; - Deploy and experiment with new algorithms on realistic wireless channels to verify the theory and to formulate more accurate analytical signal models. Symbol-by-symbol MAP demodulation is important in a wide variety of digital communication applications. Turbo codes have recently sparked significant interest due to their near capacity performance and their relatively simple demodulation architecture. MAP decoding is very important in Turbo decoding and the project will explore methods to reduce complexity of standard Turbo decoding in a principled fashion and explore noniterative techniques for Turbo decoding based on approximations to optimal MAP decoding. Additionally MAP demodulation is very important in interleaved coded systems commonly used in wireless communications. This project explores soft output algorithms for use in time-varying frequency flat and frequency selective fading channels. In particular, we are exploring high performance demodulators for time-division multiple-access applications like IS-136 and for digital television. In high capacity applications, interference suppression has greatly increased in significance. This project will extend the model-based MAP paradigm to develop high performance demodulators for signals in the presence of co-channel interference using both a conventional single antenna modem and ''space-time'' modem architectures. Another unique feature of our project is that we consi der soft output demodulation/decoding in an integrated framework with receiver front-end design (e.g., equalization, synchronization). Such an approach is essential for operating at the low SNR's necessary to take advantage of powerful codes (e.g., Turbo codes). A further unique aspect of this project is that we intend to study the resulting algorithms in a much greater depth than is traditionally attempted. This project will apply sophisticated analytical techniques to analyze the performance and stability of many of the proposed algorithms. This analysis will provide a greater insight into the performance of these algorithms than can be achieved via simulation study alone. We also intend to implement and field test many of the proposed algorithms. We currently have a working wireless modem and new algorithms can be added by a simple change of software. This combination of algorithm development, algorithm analysis, and algorithm implementation will lead to significant new insight into communication theory especially in the wireless area. This combined approach will also provide a more well-rounded educational experience for the student participants.