CPS: Medium: Collaborative Research: The Foundations of Implicit and Explicit Communication in Cyberphysical Systems

提案标题：CPS：媒介：合作研究：网络物理系统中隐式和显式通信的基础机构：加州大学伯克利分校摘要：日期：07/30/2009本研究的目的是为理解网络物理系统中隐式和显式通信提供理论基础。该方法有两个方面：(a)开发新的信息理论工具，以类似（并兼容）经典网络信息理论的方式揭示隐式通信的本质；(b)将无线生态系统本身视为一个网络物理系统，其中频谱是由异构交互网络系统操纵的物理基底，这些网络系统必须经过认证以满足安全和性能目标。这个项目的智力价值来自于正在开发的变革性技术方法。理解隐式沟通的关键是在概念上的突破，通过使用近似最优性范式结合球体包装和认知无线电信道的新思想，解决了40年来未解决的威森豪斯遭遇反例。这些技术为解决分布式控制问题开辟了全新的数学途径，这些问题长期以来被认为是根本难以解决的。它们指导了非线性控制策略的发展，这些策略可以证明比最好的线性策略好几个数量级。理解网络物理系统中显式通信的关键是在结合最坏情况和概率元素的分布式环境中主动学习、检测和估计的新方法。除了异构网络物理系统本身的许多不同应用（互联网、智能电网、智能交通等）之外，本研究还涉及无线政策：允许为下一代网络制定原则性的政府法规。培养研究生（含女）和博士后，并将研究成果纳入本科和研究生课程。国家科学基金立项摘要立项号：0932410 PI姓名：Sahai， anant印刷自

Proposal Title: CPS:Medium:Collaborative Research: The Foundations of Implicitand Explicit Communication in Cyberphysical SystemsInstitution: University of California-BerkeleyAbstract Date: 07/30/09The objective of this research is to develop the theoreticalfoundations for understanding implicit and explicitcommunication within cyber-physical systems. The approach istwo-fold: (a) developing new information-theoretic tools toreveal the essential nature of implicit communication in amanner analogous to (and compatible with) classical networkinformation theory; (b) viewing the wireless ecosystem itselfas a cyber-physical system in which spectrum is the physicalsubstrate that is manipulated by heterogeneous interactingcyber-systems that must be certified to meet safety andperformance objectives.The intellectual merit of this project comes from thetransformative technical approaches being developed. The key tounderstanding implicit communication is a conceptualbreakthrough in attacking the unsolved 40-year-old Witsenhausencounterexample by using an approximate-optimality paradigmcombined with new ideas from sphere-packing and cognitive radiochannels. These techniques open up radically new mathematicalavenues to attack distributed-control problems that have longbeen considered fundamentally intractable. They guide thedevelopment of nonlinear control strategies that are provablyorders-of-magnitude better than the best linear strategies. Thekeys to understanding explicit communication in cyber-physicalsystems are new approaches to active learning, detection, andestimation in distributed environments that combine worst-caseand probabilistic elements.Beyond the many diverse applications (the Internet, the smartgrid, intelligent transportation, etc.) of heterogeneouscyber-physical systems themselves, this research reaches out towireless policy: allowing the principled formulation ofgovernment regulations for next-generation networks. Graduatestudents (including female ones) and postdoctoral scholars willbe trained and research results incorporated into both theundergraduate and graduate curricula.NATIONAL SCIENCE FOUNDATIONProposal AbstractProposal:0932410 PI Name:Sahai, AnantPrinted from