链路质量感知无线传感器网络K近邻查询处理算法

Link-quality aware k-nearest neighbor query processing algorithm in WSN

针对现有传感器网络K近邻(KNN)查询处理算法没有考虑通信链路的质量,导致大量数据包需要重新传输从而产生高额能耗的问题,该文通过感知链路质量,提出了一种基于链路感知数据收集协议的K近邻查询处理算法(LAC-KNN)。算法通过链路感知的地理路由协议发送查询消息到离查询点最近的节点,并使用链路感知数据收集(LAC)协议沿着一条或多条路线收集查询结果。LAC可对从查询点到多个子区域的最近节点包含K近邻节点的查询区域进行动态分区,并在每个子区域中选择合适的簇头。仿真结果表明,LAC-KNN算法能耗高效,数据重传次数比现有方法大大减少。

In view of that the existing k-nearest neighbor(KNN)query processing algorithms in the wireless sensor network(WSN)do not take the quality of communication links into account,resulting in high packet re-transmissions hence and wasting a large amount of energy,a link aware data collection protocol based k-nearest neighbor(LAC-KNN)query processing algorithm is proposed here.The algorithm sends a query message to the nearest node from the query point through a link-aware geographic routing protocol,and then uses the link aware data collection(LAC)protocol to collect the query results along one or multiple itineraries.LAC dynamically partitions the query area containing the k-nearest nodes from the query point into multiple sub-regions and elects a proper cluster head in each sub-region.Simulation results show that this algorithm outperforms the previous work in terms of the number of packet transmissions.