基于图论模糊聚类的室内自适应RSSI定位算法  

Indoor adaptive RSSI localization algorithm based on graph theory and fuzzy clustering in wireless sensor networks

作　　者：高鹏[1,2] 石为人[2] 周伟[2] 王小刚[2] 李洪兵[2] 

机构地区：[1]公安部四川消防研究所,成都610036 [2]重庆大学自动化学院,重庆400044

出　　处：《仪器仪表学报》2013年第9期1998-2004,共7页Chinese Journal of Scientific Instrument

基　　金：十二五国家科技支撑计划(2011BAK07B03);国家工信部2011年物联网发展专项资金"基于物联网技术的城市管网智能感知与监控系统研发与产业化"项目资助

摘　　要：针对现有RSSI定位算法无法满足室内环境下对目标的自适应定位,提出一种基于图论模糊聚类的室内自适应RSSI定位算法IAL-GT-FC。算法根据室内环境下各区域RSSI分布的差异性,建立基于图论和模糊聚类的RSSI模糊聚类模型,并基于该模型将定位区域自适应划分成若干环境差异较小的子区域;通过建立参数自适应规则,自适应调整各子区域内的环境参数,使其满足该区域内RSSI测距的需要;结合该区域内的环境参数,通过建立相邻区域间的RSSI补偿机制对未知节点进行RSSI测距补偿;最后采用权重质心法对未知节点进行位置求解。通过实验,证明该算法具有较高的定位精度和较强的自适应能力,能够满足室内环境下对目标的精确定位。Aiming at the problem that most existing RSSI-based localization algorithms can not meet the requirement of adaptive localization for targets in indoor environment. In this paper, an indoor adaptive RSSI localization algorithm based on graph theory and fuzzy clustering （IAL-GT-FC） is proposed. According to the differences of RSSIs in different regions in the indoor environment,the proposed algorithm first establishes RSSI fuzzy clustering model based on graph theory and fuzzy clustering method. And then the target area is adaptively divided into several sub-regions using this model,in each of the sub-regions, the environment difference is not obvious. In order to meet the requirement of RSSI localization, the proposed algorithm adjusts the environmental parameters of each sub-region adaptively by establishing the parameter adaptive rules. Furthermore,the RSSI compensation mechanism between two adjacent sub-regions is es- tablished to make RSSI ranging more accurate. Finally, the final position of the target is obtained using the weighted centroid localization （WCL） approach. Experiment results show that the proposed algorithm can reach high localization accuracy and good adaptive performance, which can meet the requirement of indoor adaptive localization for targets.

关 键 词：无线传感器网络 定位 图论 模糊聚类 自适应 室内 

分 类 号：TP393.0[自动化与计算机技术—计算机应用技术]