摘要
针对无线传感器网络中的扫描覆盖问题,建立了同时满足兴趣点覆盖需求和数据投递要求的扫描覆盖数学模型,并通过与组合覆盖中经典的车辆路径问题类比分析,证明了该问题是NP-hard问题.在此基础上,提出了一种基于兴趣点分类的扫描覆盖机制FCSC(FDBSCAN_clustering_based sweep coverage).该机制利用FDBSCAN聚类算法,根据兴趣点位置信息将兴趣点分类,针对每类兴趣点,利用启发式算法生成移动传感器节点对兴趣点的访问路径,完成数据采集.仿真结果表明,在相同的网络场景下,相较于传统的扫描覆盖机制,提出的机制有效地降低了算法复杂度,节约了50%以上的算法运行时间.
A sweep coverage model for wireless sensor network which satisfies both POI (point of interest) coverage and data delivery was proposed. The sweep coverage problem was proved to be a NP-hard after translating it to a vehicle routing problem which is a classic combinatorial optimization problem. Based on this model, a novel sweep coverage scheme FCSC (FDBSCAN_clustering_based sweep coverage) was proposed by introducing clustering algorithm to classify POIs in accordance with their locations. FCSC was operated with two steps. In the first step, all the POIs in monitoring area were classified using FDBSCAN algorithm. Then, in the second step, an insertion heuristic algorithm was applied for the POIs in the same cluster to generate the trajectory of mobile sensor nodes. The simulation results show that compared with traditional sweep coverage approaches, FCSC effectively reduces the computational complexity and decreases overall execution time by more than 50% while achieving the similar performance in the same network scenarioes.
出处
《西南交通大学学报》
EI
CSCD
北大核心
2014年第1期165-172,共8页
Journal of Southwest Jiaotong University
基金
教育部博士点基金优先发展领域资助项目(20110181130007)
四川省科技支撑项目(2011RZ0004
2010GZ0169)
关键词
无线传感器网络
覆盖机制
扫描覆盖
聚类分析
启发式算法
wireless sensor networks
coverage approach
sweep coverage
cluster analysis
heuristicalgorithm