Collaborative Research: ITR: Fundamental Performance Trade-offs for Ad Hoc Networks

摘要：各种类型的自组织无线网络最近一直是人们感兴趣的主题。就我们的目的而言，这些网络可以被描述为一组节点，它们使用很少或不固定的基础设施通过无线通道相互通信。在这样的网络中，为了到达目的地，消息可能需要通过几条中间链路转发。这种类型的网络的例子包括移动自组织网络(MANET)、传感器网络和各种混合自组织/蜂窝体系结构。成功的网络具有可快速部署且几乎不需要固定基础设施的优势。然而，有效地利用这种类型的网络是一项具有挑战性的任务，部分原因是节点之间的干扰、通信信道的时变特性、非拴系节点的能量限制以及例如通过基站的集中控制的缺乏。本项目是正在进行的研究的一部分，目的是为了更好地理解自组织网络的基本性能权衡。这项工作采用跨层的观点，既考虑了物理层的限制，也考虑了更高层的性能指标，如分组吞吐量。这项工作将运用排队论、最优化和信息论的技术。与我们以前在单跳无线环境下这类问题上的工作不同，也就是更适合于蜂窝网络等情况的模型，在这些情况下，所有的通信都是从一个点进行的，在这个项目中，我们专注于具有多个发射器和多个接收器的多跳自组织网络。在这种环境下，几个新问题使分析变得更加复杂。其中包括网络内流量的路由和转发，以及每个节点发送和接收的调度。正在解决的具体问题包括表征网络稳定性和确定使数据包吞吐量最大化的自适应资源分配策略。

Abstract:Various types of ad hoc wireless networks have been the subject ofmuch recent interest. For our purposes, these networks can becharacterized as a set of nodes thatcommunicate with each other over wireless channels using little or nofixedinfrastructure. In such a network, to reach itsdestination, a message may need to be forwarded over severalintermediate links. Examples of this type network include mobilead hoc networks (MANET's), sensornetworks, and various hybrid ad hoc/cellular architectures. Suchnetworkshave the advantages of being rapidly deployable and requiring little ornofixed infrastructure. However, efficiently utilizing this type ofnetwork isa challenging task due, in part, to the interference between nodes, thetime-varying nature of the communication channels, the energylimitations of untethered nodes, and thelack of centralized control via, for example, base stations.This project is part of ongoing research with the goal offurthering understanding of basic performance trade-offs for adhoc networks. This work takes a cross layer view andconsiders both physical layer limitations as well as higher layerperformance metrics such as packet throughput. Techniques fromqueueing theory, optimization, and information theory will beutilized in this work. In contrast with our previous work on problems of this type insingle-hop wireless settings, or, that is, models more appropriatefor cases such as cellular networks, where all communication is toor from a single point, in this project, we are focusing on multi-hopad hoc networks, with multiple transmitters and multiplereceivers. In this environment, several new issues complicate theanalysis. These include the routing and forwarding of trafficwithin the network and the scheduling of transmission andreception by each node. Specific problems that are being addressedincludecharacterizing network stability and identifying adaptive resourceallocation policies which maximize packet throughput.