摘要： 提出一种利用传感器网络精细电子地图定位建立信息安全边界的方法。在离线阶段，采用离散概率密度函数统计接收信号强度，绘制精细电子地图并定义信息安全边界，在线运行阶段，用位置概率矩阵描述目标位置，用矩阵运算融合多观测点、不同时刻的观测，并采用贝叶斯滤波器动态跟踪目标位置。实验结果表明，该方案在1 m和2 m的过渡区域宽度条件下均有较高的安全判定准确性。

关键词: 电子地图, 安全边界, 传感器网络, 定位

Abstract: This paper proposes a fine-grained radio-map localization based information security boundary in sensor network. In off-line training phase, discrete Probability Density Functions(PDF) are used to characterize the Received Signal Strength(RSS), to build fine-grained radio-map and to define the security boundaries. In on-line positioning phase, Position Probability Matrix(PPM) is proposed to describe location estimations. The estimations of different observers, of different time can be fused by matrix operations. Bayesian filter is proposed to dynamically track positions of mobile targets. Experimental results show that high security determinant accuracy is achieved when the transition interval is set as 1 meter and 2 meters.

Key words: radio-map, security boundary, sensor network, localization

中图分类号: 

• N945