FourierNet: Routing in Dense Networks

这项研究项目解决了与密集网络中的路由相关的一些基本问题，这些问题有望成为以嵌入式、无处不在和普适计算为例的计算新方面的关键推动因素。具体地说，提出了一种全新的密集网络路由框架，称为基于轨迹的转发。这种新的方法是基于源路由和笛卡尔转发的推广，其中轨迹由源设置为连续轨迹，但转发决策是基于与轨迹的关系而不是最终目的地。基于轨迹的转发具有许多特征，使其成为密集自组织网络中作为低层原语的理想候选者。首先，它将路径名与路径本身分离。这是密集网络中最关键的方面，在密集网络中，节点可能会移动、进入打瞌睡模式或出现故障，从而使基于源的路径无用或使路由维护成本高昂。其次，弹道的规范与目的地无关。这使得基于轨迹的转发既可用作指示目的地时的路由支持，也可用作发现支持原语，当目的地未知时，或用作洪泛替代。除了简单的单播，轨迹路由和转发对于许多其他重要的网络功能，如洪泛、发现、组播和广播、路径恢复甚至定位，都具有显著的优势。本研究的目的是测试使用轨迹的路由方法，探索指定轨迹的方法，并实施许多网络功能，如发现、泛洪和网络管理。可以有效部署、使用和评估基于轨迹的转发的一个特定领域是监控交通系统或物理基础设施。基于轨迹的转发有望从根本上改变我们对网络中路由问题的看法，并导致一个具有丰富问题空间的新的研究领域。

This research project addresses some fundamental issues related to routing in dense networks that are expected to be the key enabler of new facets of computing exemplified by embedded-, ubiquitous-, and pervasive computing. Specifically, a radically new framework for routing in dense networks called Trajectory Based Forwarding is proposed. This novel method is a generalization of source based routing and Cartesian forwarding in that the trajectory is set by the source as a continuous trajectory, but the forwarding decision is based on the relationship to the trajectory rather than the final destination. Trajectory Based Forwarding has a number of features that makes it an ideal candidate as a low level primitive in dense ad-hoc networks. First, it decouples the path name from the path itself. This is the most critical aspect in a dense network, where nodes might move, go into doze mode, or fail, thereby rendering a source based path useless or make route maintenance expensive. Second, the specification of the trajectory is independent of the destination. This makes the Trajectory Based Forwarding usable both as a routing support, when the destination is indicated, as a discovery support primitive, when the destination is not known, or as a flooding replacement. In addition to the simple unicast, trajectory routing and forwarding have significant advantages for many other important network functions such as flooding, discovery, multicast and broadcast, path resilience and even positioning. The objective of this research is to test routing methods using trajectories, to explore ways of specifying trajectories, and to implement many of the networking functions such as discovery, flooding, and network management. A specific domain where Trajectory Based Forwarding can be effectively deployed, used, and evaluated is in monitoring transportation systems or physical infrastructure. It is expected that Trajectory Based Forwarding can radically change the way we think about routing in networks and lead to a new area of research with a rich problem space.