Cyber Systems: Adaptive Distributed Systems Based on Cooperative and Combination Strategies

基于合作与组合策略的自适应分布式系统。这项研究的目标是开发能够赋予分布式网络实时适应和学习能力的处理算法。分布式算法应该能够利用分布式节点的空间分集以及数据的时间维度，以便通过相邻节点之间的合作来提供改进的性能水平。该方法基于利用增量和组合策略，并允许信息在网络中实时流动。其目的是开发自适应分布式策略，这些策略需要有限的本地通信和计算，并且对拓扑变化具有健壮性。这种自适应网络的推动力是这样一种认识，即连接个人电脑、笔记本电脑、手机、传感器和执行器的分布式系统将构成未来数据和通信网络的主干。应用将包括传感器网络、精准农业、环境监测、目标定位、救灾管理和医疗检查。分布式处理的目的是将在分布式节点本地捕获的数据组合在一起，以便在全局范围内提供良好的性能。处理需要自适应和实时进行，因为不仅环境条件随时间和空间变化，而且网络拓扑也可能变化。这项拟议的研究展示了如何结合来自不同学科的工具来开发自适应的分布式网络。这项工作的进展将影响信息技术基础设施，并将有助于在一个与未来通信和网络系统基本相关的领域对学生进行培训。

Adaptive Distributed Systems Based on Cooperative and Combination StrategiesScope and Intellectual Merit. The objective of this research is to develop processing algorithms that endow distributed networks with real-time adaptation and learning abilities. The distributed algorithms should be able to exploit the spatial diversity of the distributed nodes, as well as the temporal dimension of the data, in order to deliver improved levels of performance through cooperation between neighboring nodes. The approach is based on exploiting incremental and combination strategies and allows for real-time flow of information across the network. The purpose is to develop adaptive distributed strategies that require limited local communications and computation and that are robust to topology changes.Broader Benefits. The driving force for such adaptive networks is the realization that distributed systems linking PCs, laptops, cell phones, sensors and actuators will form the backbone of future data and communication networks. Applications will include sensor networks, precision agriculture, environment monitoring, target localization, disaster relief management, and medical examinations. The purpose of distributed processing is to combine the data captured locally at the distributed nodes in order to deliver good performance globally. The processing needs to be done adaptively and in real-time because not only the environmental conditions vary with time and space, but the network topology may also vary. The proposed research shows how to combine tools from across disciplines in order to develop adaptive distributed networks. Progress in this work will impact the information technology infrastructure and will contribute to the training of students in an area of fundamental relevance to future communications and networking systems.