融合ROF模型的高斯滤波RSSI测距算法
|
摘要
针对WiFi室内定位时由于室内环境的复杂性、不确定性导致所采集的RSSI信号存在奇异值、波动性的问题,提出一种基于高斯滤波的改进算法,以期提高WiFi定位的测距精度:在原有高斯滤波算法基础上与全变分去噪算法相结合,引入ROF模型;并利用采集的RSSI矩阵数据梯度控制扩散系数,当扩散系数较小时保留有效信息,扩散系数较大时去除噪声;接着对去噪后的RSSI矩阵进行自适应校正,同时抑制阶梯效应引起的RSSI矩阵平滑不均匀现象,实现对强度信号的滤波及平滑处理。实验结果表明,该方法对提高WiFi定位的测距精度有较好效果。
Aiming at the problem of the singularity and the fluctuation of the collected RSSI signal in WiFi indoor positioning due to the complexity and uncertainty of the indoor environment,the paper proposed an improved Gaussian filter algorithm for promoting the ranging accuracy of WiFi positioning:the total variational denoising algorithmth was comined with original Gaussian filtering algorithm,and Rudin Osher Fatemi model was introduced;the diffusion coefficient was controlled by the gradient of the collected RSSI matrix data,that is,when the diffusion coefficient was smaller,the effective information was retained,while when the coefficient was larger,the noise was removed;then the denoised RSSI matrix was adaptively corrected with suppressing the unevenness of RSSI matrix caused by the staircase effect,so that the filtering and smoothing of the intensity signal was realized finally.Experimental result showed that the proposed method could well improve the ranging accuracy of WiFi positioning.
Aiming at the problem of the singularity and the fluctuation of the collected RSSI signal in WiFi indoor positioning due to the complexity and uncertainty of the indoor environment,the paper proposed an improved Gaussian filter algorithm for promoting the ranging accuracy of WiFi positioning:the total variational denoising algorithmth was comined with original Gaussian filtering algorithm,and Rudin Osher Fatemi model was introduced;the diffusion coefficient was controlled by the gradient of the collected RSSI matrix data,that is,when the diffusion coefficient was smaller,the effective information was retained,while when the coefficient was larger,the noise was removed;then the denoised RSSI matrix was adaptively corrected with suppressing the unevenness of RSSI matrix caused by the staircase effect,so that the filtering and smoothing of the intensity signal was realized finally.Experimental result showed that the proposed method could well improve the ranging accuracy of WiFi positioning.
引文
[1]彭宇,王丹.无线传感器网络定位技术综述[J].电子测量与仪器学报,2011,25(5):389-399.
[2]陈国良,张言哲,汪云甲,等.WiFi-PDR室内组合定位的无迹卡尔曼滤波算法[J].测绘学报,2015,44(12):1314-1321.DOI:10.11947/j.AGCS.2015.20140691.
[3]陈群,高星伟,郭杭,等.一种基于RSSI的三维加权质心定位算法[J].测绘科学,2013,38(4):91-93.DOI:10.16251/j.cnki.1009-2307.2013.04.029.
[4]赵永翔,周怀北,陈淼,等.卡尔曼滤波在室内定位系统实时跟踪中的应用[J].武汉大学学报(理学版),2009,55(6):696-700.
[5]吴东金,夏林元.面向室内WLAN定位的动态自适应模型[J].测绘学报,2015,44(12):1322-1330.DOI:10.11947/j.AGCS.2015.20130780.
[6]HUA J S,XIAO P Z,ZHI W.Efficient closed-form algorithms for AOA based self-localization of sensor nodes using auxiliary variables[J].IEEE Transaction on Signal Processing,2014,62(10):2580-2594.
[7]CHAN E C L,BACIU G,MAK S C.Using WiFi signal strength to localize in wireless sensor networks[J].IEEE Computer Society,2009:538-542.DOI:10.1109/CMC.2009.233.
[8]王彦斐,赵言,黎慕韩.高斯滤波在Zigbee室内定位中的应用研究[J].测绘地理信息系统,2016,45(5):56-59.
[9]陶为戈,朱华,贾子彦.基于RSSI混合滤波和最小二乘参数估计的测距算法[J].传感技术学报,2012,25(12):1748-1753.
[10]贾银山,曾帅斌,王璟.一种基于优化RSSI精度的加权质心算法[J].计算机应用与软件,2017,34(11):162-165.
[11]李桢,黄劲松.基于RSSI抗差滤波的WiFi定位[J].武汉大学学报(信息科学版),2016,41(3):361-366.
[12]万群,郭贤生,陈章鑫.室内定位理论、方法和应用[M].北京:电子工业出版社,2012:80-81.
[13]薛卫星,邱卫宁,花向红.一种用于RSS提取的改进粒子滤波算法[J].测绘科学,2017,42(11):20-24.
[14]陈丽,王学东,孙晶晶,等.基于改进高斯滤波的室内无线定位算法[J].计算机技术及应用,2014,36(3):31-34.
[15]RUDIN L I,OSHER S,FATEMI E.Nonlinear total variation based noise removal algorithms[J].Physica D,1992,60:259-268.
[2]陈国良,张言哲,汪云甲,等.WiFi-PDR室内组合定位的无迹卡尔曼滤波算法[J].测绘学报,2015,44(12):1314-1321.DOI:10.11947/j.AGCS.2015.20140691.
[3]陈群,高星伟,郭杭,等.一种基于RSSI的三维加权质心定位算法[J].测绘科学,2013,38(4):91-93.DOI:10.16251/j.cnki.1009-2307.2013.04.029.
[4]赵永翔,周怀北,陈淼,等.卡尔曼滤波在室内定位系统实时跟踪中的应用[J].武汉大学学报(理学版),2009,55(6):696-700.
[5]吴东金,夏林元.面向室内WLAN定位的动态自适应模型[J].测绘学报,2015,44(12):1322-1330.DOI:10.11947/j.AGCS.2015.20130780.
[6]HUA J S,XIAO P Z,ZHI W.Efficient closed-form algorithms for AOA based self-localization of sensor nodes using auxiliary variables[J].IEEE Transaction on Signal Processing,2014,62(10):2580-2594.
[7]CHAN E C L,BACIU G,MAK S C.Using WiFi signal strength to localize in wireless sensor networks[J].IEEE Computer Society,2009:538-542.DOI:10.1109/CMC.2009.233.
[8]王彦斐,赵言,黎慕韩.高斯滤波在Zigbee室内定位中的应用研究[J].测绘地理信息系统,2016,45(5):56-59.
[9]陶为戈,朱华,贾子彦.基于RSSI混合滤波和最小二乘参数估计的测距算法[J].传感技术学报,2012,25(12):1748-1753.
[10]贾银山,曾帅斌,王璟.一种基于优化RSSI精度的加权质心算法[J].计算机应用与软件,2017,34(11):162-165.
[11]李桢,黄劲松.基于RSSI抗差滤波的WiFi定位[J].武汉大学学报(信息科学版),2016,41(3):361-366.
[12]万群,郭贤生,陈章鑫.室内定位理论、方法和应用[M].北京:电子工业出版社,2012:80-81.
[13]薛卫星,邱卫宁,花向红.一种用于RSS提取的改进粒子滤波算法[J].测绘科学,2017,42(11):20-24.
[14]陈丽,王学东,孙晶晶,等.基于改进高斯滤波的室内无线定位算法[J].计算机技术及应用,2014,36(3):31-34.
[15]RUDIN L I,OSHER S,FATEMI E.Nonlinear total variation based noise removal algorithms[J].Physica D,1992,60:259-268.