EAGER: A New Algorithmic and Graph Model for Networking in Challenged Environments

EAGER：EAGER提出了一种通用的图模型，它可以捕获网络环境中的移动性（NICE）。这种模型被称为加权演化图，它捕捉时空动态，同时保持足够简单，以保持传统图模型的大部分优雅结构。我们fi RST提出了基于不同度量的若干路径优化问题，包括最早完成、最小跳、最快和最大可靠性。然后，我们在这个新模型中扩展了几个图的概念，该建议解决了一个基本问题：我们能否开发一个本地化的解决方案，其中图是“修剪”（通过删除节点/链接跨越时间和空间），只使用本地信息在每个节点？修剪图的优点是降低了路由和广播的搜索复杂度。我们计划将所提出的模型应用于三个不同的应用：（a）具有频繁打开/关闭的传感器的动态传感器网络，（B）具有循环运动轨迹的移动的网络（如车载网络），以及（c）大学生携带iMotes/智能手机在定期会议和课堂上保持联系记录。所提出的捕捉NICE的图论模型是一个简单的模型。该提案提出了一种有前途的和独特的方式，应用此图模型来解决一组新的路径优化问题和其他图形概念。该模型可应用于动态传感器网络、DTN和人际网络等三个重要应用。我们设想，这项研究的见解和结果将为建模和分析NICE提供指导方针。本研究也将在广义图抽象下开拓和贡献于动态和挑战网络的基础理论。

EAGER: A New Algorithmic and Graph Model for Networking in Challenged EnvironmentsThis EAGER proposal presents a generalized graph model that can capture mobility in Networking in Challenged Environments (NICE). This model is called a weighted evolving graph which captures time-space dynamics while remaining simple enough to maintain most of the elegant structure of the traditional graph model. We first present several path optimization problems based on different metrics, including earliest completion, minimum hop, fastest, and maximum reliability. We then extend several graph concepts in this new model.The proposal addresses one fundamental issue: can we develop a localized solution in which the graph is "trimmed" (by removing nodes/links across time and space) using only local information at each node? The merit of a trimmed graph is the reduction of searching complexity for routing and broadcasting.We plan to apply the proposed model to three different applications: (a) dynamic sensor networks with frequently switched on/off sensors, (b) mobile networks with cyclic movement trajectories (such as vehicular networks), and (c) people networks in which college students carrying iMotes/smart phones maintain contact records during periodic meetings and classes.The proposed graph theoretic model to capture NICE is a simple one. The proposal presents a promising and unique way of applying this graph model to address a new set of path optimization problems and other graph concepts. The proposed model can be applied to three important applications in dynamic sensor networks, DTNs, and people networks. We envision that insights and results from this research will provide guidelines for modeling and analyzing NICE. This research will also exploit and contribute to fundamental theories on dynamic and challenged networks under the generalized graph abstraction.