基于环扇区的无线传感器网络K近邻查询处理算法

K Nearest Neighbor Query Processing Algorithm Based on Ring Sectors for Wireless Sensor Networks

现有传感器网络环境下的K近邻查询处理算法没有考虑节点失效对查询处理的影响,导致查询结果极易丢失.提出一种基于环扇区的数据收集算法RISC,以保证查询处理过程的鲁棒性.给出理论上最节省能量的环扇区大小设置,以减少算法的能量消耗.系统地分析了算法在不同节点密度、节点失效概率和查询区域条件下的查询成功率和能量消耗.实验结果表明,在绝大多数情况下,提出的基于环扇区的K近邻查询处理算法优于现有的基于路线的算法.

The existing KNN query processing algorithm for wireless sensor networks does not take node failures into consideration.Therefore,the query result is very prone to be lost.In this paper,we propose a data collection algorithm called RISC based on ring sector to ensure the robustness of query processing.We study the ring sector size parameter to reduce energy consumption in the process of data collection.Finally,we analyze the influence of node density,the probability of node failure,query region size on the query success rate and energy consumption.Experimental results show that in most cases the KNN query processing algorithm proposed in this paper outperforms the Itinerary-based algorithm.