CAREER: An Integrated Digital Signal Processing Framework for Optimized Wireless Communications

无线通信系统，特别是码分多址（CDMA）系统，是快速发展的全球信息基础设施的重要组成部分。由于物理网络层对数字信号处理的要求越来越高，数字信号处理（DSP）在无线通信中发挥着核心作用。CDMA系统设计中主要的dsp挑战来自物理层处理的三个关键因素：(1)信道传播动力学；(2)多址干扰；(3) DSP算法的固有复杂性。提出的研究有两个主要目标来满足物理层CDMA挑战：*开发DSP工具和理论，以发起时变多径色散，时变多径衰落和多址干扰的集成攻击。开发实用高性能系统的设计策略，优化DSP技术固有的复杂性和性能权衡。为了实现这些目标，我们将开发一种基于DSP的标准多路径多普勒坐标统一框架。规范坐标是由一个固定的基础来定义的，这个基础是根据信号波形的某些离散多径延迟和多普勒频移来的信道传播动力学的基本特征来定义的。规范坐标自然地连接了关键的物理层面，从而促进了统一的DSP开发。首先，所有与传播效果相关的处理都可以直接在坐标中执行。其次，相同的坐标提供了期望信号和多址干扰的基于规范子空间的表示。最后，坐标的简单计算和子空间结构提供了对复杂性的直接处理。该研究的核心是一个特别有前途的创新：现有的干扰抑制技术将传播效应视为一种滋扰，而规范坐标利用由信道效应引起的子空间结构来增强干扰抑制。该项目的教育目标包括DSP和通信的综合课程开发，促进概念思维的主题教学方法，以及利用技术扩展传统教学和学习的界限。

Abstract for CCR-9875805Wireless communication systems, particularly code-divisionmultiple access (CDMA) systems, are vital building blocks of the rapidlyevolving global information infrastructure. Digital signal processing (DSP)is playing a central role in wireless communications due to the increasinglysophisticated processing required in the physical network layer. The majorDSP challenges in CDMA system design stem from three key factors thatdominate the processing in the physical layer: (1) channel propagationdynamics; (2) multiaccess interference; and, (3) the inherent complexity ofthe DSP algorithms. The proposed research has two primary goals for meetingthe physical layer CDMA challenges:* Development of DSP tools and theory to launch an integrated attackon time-varying multipath dispersion, time-varying multipath fading, andmultiaccess interference.* Development of design strategies for practical high-performancesystems that optimize the complexity versus performance tradeoffs inherentin the DSP techniques.A new unified framework based on DSP in canonical multipath-Dopplercoordinates will be developed to meet these goals. The canonical coordinatesare defined by a fixed basis that is derived from a fundamentalcharacterization of channel propagation dynamics in terms of certain discretemultipath delays and Doppler shifts of the signaling waveform. The canonicalcoordinates naturally connect the key physical layer facets and thusfacilitate a unified DSP development. First, all processing relating topropagation effects can be directly performed in the coordinates. Second,the same coordinates provide a canonical subspace-based representationof the desired signal and multiaccess interference. Finally, the simple computation and subspace structure of the coordinates affords a direct handle on complexity. At the heart of the research is a particularly promisinginnovation: whereas existing interference suppression techniques treatpropagation effects as a nuisance, canonical coordinates exploit the subspacestructure induced by channel effects for enhanced interference suppression.The educational objectives of the program include integrated curriculumdevelopment in DSP and communications, a thematic approach to teaching tofacilitate conceptual thinking, and leveraging of technology to extend theboundaries of traditional instruction and learning.