基于BP神经网络的LQI测距研究  

Research of LQI Ranging Based on BP Neural Network

作　　者：赵罡 盛林 王军 ZHAO Gang;SHENG Lin;WANG Jun(CETHIK Group Co.,Ltd.,Hangzhou 310012,China)

机构地区：[1]中电海康集团有限公司,浙江杭州310012

出　　处：《智能物联技术》2019年第3期19-25,共7页Technology of Io T& AI

摘　　要：接收信号强度(RSSI)测距是无线传感器网络定位技术中较常采用的方法,但使用RSSI测距时,反射、散射和障碍物会产生极大影响。本文采用LQI进行测距,并通过实验表明LQI值和距离之间存在衰减和一一对应关系。为了实现LQI测距,本文将高斯滤波预处理后的LQI值作为输入,通过BP神经网络训练后进行距离求解。通过CC2530的Zig Bee平台实验表明,与传统的方法相比,基于BP神经网络模型的LQI测距具有更高的测距精度和稳定性。Received Signal Strength Indicator(RSSI)ranging is a commonly used method in wireless sensor network positioning technology,but when using RSSI ranging,reflection,scattering,and obstacles can have a significant impact.In this paper,the LQI is used for ranging,and experiments showed that there was attenuation and one-to-one correspondence between LQI and distance.In order to realize the LQI value ranging,the LQI value preprocessed by Gaussian filter was taken as input,and the distance was solved by BP neural network after training.Experiments on the ZigBee platform of CC2530 showed that LQI ranging based on BP neural network model has higher ranging accuracy and stability than traditional methods.

关 键 词：LQI 高斯滤波 BP神经网络 测距 

分 类 号：TP391[自动化与计算机技术—计算机应用技术]