多径效应下变分贝叶斯压缩感知无源定位方法  

Variational Bayesian compressive sensing device-free localization method under multiple path effect

作　　者：李华静 李宁 郭艳 盛金锋 陈承 LI Hua-jing;LI Ning;GUO Yan;SHENG Jin-feng;CHEN Cheng(College of Communications Engineering,Army Engineering University of PLA,Nanjing 210007,China)

机构地区：[1]中国人民解放军陆军工程大学通信工程学院,江苏南京210007

出　　处：《计算机工程与设计》2023年第8期2241-2247,共7页Computer Engineering and Design

基　　金：国家自然科学基金面上基金项目(61871400);江苏省自然科学基金面上基金项目(BK20211227)。

摘　　要：无源定位技术常受信号多径效应的干扰,严重影响定位精度。现有定位技术大多直接忽略多径效应,然而其信号强度高,可提供定位信息,故忽略其影响不够贴合实际。基于此,提出基于多径反射的变分贝叶斯压缩感知无源定位方法。利用智能反射面以及虚拟节点模型构建所需的通信链路,在不额外增加硬件资源的情况下增加通信链路,节省通信资源。利用改进的椭圆权重模型,量化目标阴影效应,同时在贝叶斯稀疏重构中引入网格裁剪机制,在保证定位精度的条件下降低计算复杂度,节省定位时间。仿真结果表明,在节省通信资源的情况下,该方法实现了快速高精度定位。Device-free localization techniques are often disturbed by signal multiple path effects,which seriously affect the localization accuracy.Most of the existing localization techniques directly ignore the multiple path effect,but the signal strength is high and can provide positioning information,so the ignored influence is not relevant to the reality.Based on this,a variational Bayesian compression-aware passive localization method based on multiple path reflections was proposed.The required communica-tion links were constructed using intelligent reflective surfaces as well as virtual node models,which increased the communication links without additional hardware resources,and communication resources were saved.An improved elliptical weight model was used to quantify the target shadowing effect,while a grid cropping mechanism was introduced in the Bayesian sparse reconstruction,which also reduced the computational complexity and saved the localization time under the condition of ensuring the localization accuracy.The final simulation results show that fast and high-precision localization is achieved on the premise of saving communication resources.

关 键 词：无源定位 多径效应 虚拟节点模型 压缩感知 通信链路 贝叶斯稀疏重构 网格裁剪 

分 类 号：TN911.7[电子电信—通信与信息系统]