Information Flow Theory in Dense Wireless Networks

摘要：未来的无线网络是由大量密集分布在大片地理区域的节点组成的。这种网络被用于广域环境保护、交通运输、自动化农业或民用基础设施系统中的无线微传感等应用。分布式源生成的信息通过相互转发流量的节点的协作努力传输到目的地；因此，信息在由许多跳组成的路由上传播，每一跳都有助于短距离传输。仔细分析具有如此大量无线节点的网络涉及到令人望而却步的复杂程度。这种复杂性是由于网络中每对节点之间的不协调交互(例如，由于同时传输的干扰)；因此，人们对大规模密集无线网络的行为知之甚少。空间连续信息流模型是克服传统离散空间方法过于复杂的一种很有前途的方法。该项目预计将产生强大的设计和分析工具，用于不同的目的，如最小化给定信息源空间分布所需的无线节点数量，寻找空间中所需的最小节点密度，确定节点的位置，最小化节点的传输功率，负载平衡等。该项目是朝着发展密集无线网络中的信息流理论迈出的重要一步。这样的理论是与其他科学和工程分支的经典流动理论(如流体动力学、热交换和静电学)的无线联网对应。

Abstract:Future wireless networks are composed of a massive number of nodes densely distributed in large geographical areas. Such networks are used in applications such as wide area environmental protection, transportation, automated agriculture, or wireless micro-sensing in civil infrastructure systems. Information generated by distributed sources is transported to the destinations by collaborative efforts of nodes that forward the traffic of each other; therefore, the information is diffused over routes composed of many hops, each contributing to a short range transmission. Careful analysis of the networks with such large number of wireless nodes involves a prohibitive level of complexity. This complexity is due to the uncoordinated interactions between each pair of nodes in the network (e.g., due to interference of simultaneous transmissions); as a result, little is known about the behavior of massively dense wireless networks.The introduced methodology models a dense wireless network by a continuum of nodes. The spatially continuous model of information flow is a very promising methodology to overcome the prohibitive complexity of conventional discrete space methods. The project is expected to produce powerful design and analysis tools for different purposes such as minimizing the number of required wireless nodes for a given spatial distribution of information sources, finding minimum required density of nodes in space, determining the placement of nodes, minimizing the transmit power of nodes, load balancing, etc. The project is an important step towards development of a theory of information flow in dense wireless networks. Such a theory is the wireless networking counterpart of the classical flow theory of other branches of science and engineering such as fluid dynamics, heat exchange, and electrostatics.