嵌入分布式系统中的多节点自适应重构

结合多UAV、传感器网络等嵌入分布式系统的自适应计算需求，分析嵌入分布式系统中节点间相似度的表示机制与相似度的计算方法。提出嵌入分布式系统中相似节点的自动发现机制以及相似节点上的任务自动部署方法。结合“感知评估调整感知”的自适应计算思想，研究任务/节点失效时的可适度降级分布重构技术。该方法对于提高嵌入分布式系统的自适应计算与容错能力有一定的应用价值。

Adaptive Reconfiguration Among Multi-nodes in Distributed Embedded System

According to the requirement for adaptability in distributed embedded system such as UAVs, wireless sensor network and so on, some common methods are proposed to define similar nodes, compute their similarity degree, discover similar nodes and deploy the other’s tasks on them automatically. Based on the adaptive computing thoughts of “sensing  evaluating  adjusting  sensing”, some reconfiguration ways for fault nodes and tasks are studied. The methods are more efficient to improve the adaptability of distributed embedded systems, in which some nodes are similar.