In wireless sensor networks, topology control plays an important role for data forwarding efficiency in the data gathering applications. In this paper, we present a novel topology control and data forwarding mechanism called REMUDA, which is designed for a practical indoor parking lot management system. REMUDA forms a tree-based hierarchical network topology which brings as many nodes as possible to be leaf nodes and constructs a virtual cluster structure. Meanwhile, it takes the reliability, stability and path length into account in the tree construction process. Through an experiment in a network of 30 real sensor nodes, we evaluate the performance of REMUDA and compare it with LEPS which is also a practical routing protocol in TinyOS. Experiment results show that REMUDA can achieve better performance than LEPS.
数据融合是降低无线传感器网络的冗余能耗、延长网络生存期的有效手段之一.传输延时的分配是数据融合中的重要问题,它对网络融合的效果和数据分组的传输延迟具有很大影响.提出了一种基于融合贡献的传输延时分配算法ACDA(aggregate contribution based delay-time allocation),量化了路由树中不同位置的节点对融合效果的影响,并依此按比例分配融合等待时间.算法充分考虑到了各节点在路由树中的位置差异性和节点间的相互影响.仿真结果表明,同现有的几种延时分配算法相比,ACDA能够更加有效地改进数据融合效果,提高网络实时性.
