An NLoS-based Enhanced Sensing Method for MmWave Communication System Shiwen Heyz Kangli Cai Shiyue Huang Zhenyu Anz Wei Huangx Ning Gao

2025-04-27 0 0 1.85MB 7 页 10玖币

侵权投诉

An NLoS-based Enhanced Sensing Method for

MmWave Communication System

Shiwen He∗†‡, Kangli Cai∗, Shiyue Huang∗, Zhenyu An‡, Wei Huang§, Ning Gao¶

∗The School of Computer Science and Engineering, Central South University, Changsha 410083, China.

†The National Mobile Communications Research Laboratory, Southeast University, Nanjing 210096, China.

‡The Purple Mountain Laboratories, Nanjing 211111, China.

§The School of Computer Science and Information Engineering, Hefei University of Technology, Hefei 230601, China.

¶Department of Standardization, OPPO Research Institute, Beijing, 100020, China.

Email: {shiwen.he.hn, caikangli, huangsy}@csu.edu.cn, anzhenyu@pmlabs.com.cn, huangwei@hfut.edu.cn, gaoning1@oppo.com

Abstract—The millimeter-wave (mmWave)-based Wi-Fi sens-

ing technology has recently attracted extensive attention since

it provides a possibility to realize higher sensing accuracy.

However, current works mainly concentrate on sensing scenarios

where the line-of-sight (LoS) path exists, which signiﬁcantly

limits their applications. To address the problem, we propose

an enhanced mmWave sensing algorithm in the 3D non-line-

of-sight environment (mm3NLoS), aiming to sense the direction

and distance of the target when the LoS path is weak or blocked.

Speciﬁcally, we ﬁrst adopt the directional beam to estimate the

azimuth/elevation angle of arrival (AoA) and angle of departure

(AoD) of the reﬂection path. Then, the distance of the related path

is measured by the ﬁne timing measurement protocol. Finally,

we transform the AoA and AoD of the multiple non-line-of-

sight (NLoS) paths into the direction vector and then obtain

the information of targets based on the geometric relationship.

The simulation results demonstrate that mm3NLoS can achieve a

centimeter-level error with a 2m spacing. Compared to the prior

work, it can signiﬁcantly reduce the performance degradation

under the NLoS condition.

Index Terms—mmWave sensing, Wi-Fi, NLoS path

I. INTRODUCTION

In recent years, Wi-Fi has been widely deployed in most

public and private spaces due to its simplicity, reliability, and

ﬂexibility. The extremely dense Wi-Fi devices not only provide

convenience for people, but also create a perfect opportunity

to sense the environment. Therefore, by extracting appropriate

signal features of Wi-Fi signals, e.g., phase differences [1] or

doppler shifts [2], we can effectively detect the presence of

targets and further track them.

Target sensing based on Wi-Fi signals has been widely

studied for lower frequencies, e.g., the ﬁngerprint-based [3]

and geometry-based methods [4]. These works achieved con-

siderable performance due to the rich multipath signals in the

environment and their weak attenuation characteristics. How-

ever, they critically depended on the channel state information

(CSI), and the accuracy was limited by the antenna numbers

and bandwidth. Moreover, these systems were designed for

communication and did not consider the sensing function. To

this end, the IEEE 802.11bf task group (TGbf) is working

on making appropriate modiﬁcations to the Wi-Fi standard to

utilize the existing 802.11-compatible waveforms for Wi-Fi

sensing or integrated sensing and communication (ISAC) [5].

Speciﬁcally, IEEE 802.11bf deﬁnes the support of 802.11ad

and 802.11ay protocols, which signiﬁcantly operate in the

millimeter wave (mmWave) band. Therefore, a higher sensing

performance can be expected in the future.

Although mmWave sensing is attractive, the short wave-

length of mmWave leads to high path loss, and the propagation

path is easily blocked by obstacles. To compensate for the at-

tenuation, phased-array antennas and beamforming techniques

for directional transmission are usually adopted. It means that

one can estimate the angle of departure (AoD) and angle of

arrival (AoA) from the directionally transmitted and received

signals. Besides, the large bandwidth of mmWave provides a

high distance resolution. Therefore, it is possible to realize

accurate target sensing geometrically.

Prior work has demonstrated that mmWave could provide

sub-decimeter accuracy in short-range sensing scenes, such as

gesture tracking [6], mainly realized by leveraging two Wi-

Fi links to detect the phase changes of CSI values due to the

variation of propagation path length. However, the transceivers

signiﬁcantly required a speciﬁc placement. The authors of

[7] proposed a passive target sensing algorithm POLAR for

IEEE 802.11ad devices, which used the AoD and time of

ﬂight (ToF) of the multi-path components estimated from

channel impulse response (CIR) corresponding to different

beam patterns to sense the target. Still, it could only locate

the object in 2D space. Furthermore, these systems were

signiﬁcantly designed based on the premise that the line-of-

sight (LoS) path always existed. For non-line-of-sight (NLoS)

conditions, the propagations of the signals were signiﬁcantly

affected, e.g., increased ToF and changed AoA. If the NLoS

measurements were utilized directly as the LoS measurement,

it would result in a large sensing error [8]. To address this

problem, the monostatic radar for sensing was proposed in

[9], which could directly estimate the range and relative radial

speed using the received echo signal. Nevertheless, it is more

attractive to realize mmWave sensing that can be applied for

multi-device scenes, since the multi-angle detection for the

target can remarkably improve sensing accuracy.

Based on the above consideration, this paper explores a

arXiv:2210.04747v1 [cs.IT] 10 Oct 2022

sensing method between two devices in the 3D space. It is

more challenging compared to the previous approach. On the

one hand, the baseline information is signiﬁcantly unknown

due to the blocked LoS path, which makes solving bistatic

triangles infeasible. On the other hand, diverse targets poten-

tially lie on different planes. The complex geometry between

targets, transmitter, and receiver further increases the difﬁculty.

To address the above challenges, we propose an enhanced

mmWave sensing algorithm in the 3D NLoS environment

(mm3NLoS), which tries to sense the target by exploiting

the geometric relationship of multiple NLoS paths. The main

contributions of this work can be summarized as follows:

•We design an enhanced sensing approach in a 3D NLoS

environment so as to mitigate performance degradation

when the LoS path is weak or blocked.

•We introduce the projection operation to simplify the

problem and derive an analytical expression about the

direction and distance between the target and receiver

with the AoD, AoA, and ToF of two propagation paths.

•We compare the proposed method with the POLAR

algorithm using simulated data. The result shows that our

method performs better regardless of whether the LoS

path exists.

II. SYSTEM MODEL

In this paper, we exploit an mmWave MIMO system with

an analog transceiver structure to sense a target. As shown in

Fig. 1, it consists of one access point (AP) and one station

(STA), where AP and STA are the transmitter and receiver,

respectively. We further assume that the LoS path is blocked

by obstacles, so the NLoS path plays a dominant role in

sensing. Generally, an NLoS path may be a multiple-bounce

or single-bounce reﬂection. In this paper, we only consider the

strongest propagation path and assume it to be single-bounce.

In particular, utilizing multiple NLoS paths (the current and

at least one historical path), the parameters of the target,

including the distance and direction, can be estimated uniquely

via geometric relations. To realize this, we record the AoD,

AoA, ToF, and signal-to-noise ratio (SNR) of the sensing path

in a historical measurement table and choose it based on the

SNR.

Consider the uniform planar array (UPA) antennas and the

single-path extended Saleh-Valenzuela geometric model for

the mmWave system [10]. Then, the channel matrix can be

expressed as

H=pNtNrgar(ϕr, θr)aH

t(ϕt, θt)(1)

where gis the complex path gain with g∼ CN(0,1).Ntand

Nrare the antenna numbers of the AP and STA. ϕtand θtare

the azimuth and elevation of AoD. ϕrand θrare the azimuth

and elevation of AoA. For convenience, we assume that the

Fig. 1. An illustration of target sensing in a 3D NLoS scene.

UPA is placed in the YOZ plane, then the array response

vectors are given by

at(ϕt,θt) = 1

√Nth1,··· , e−jkd(psin ϕtsin θt+qcos θt),

··· , e−jkd((Nt,h −1) sin ϕtsin θt+(Nt,v −1) cos θt)i

(2)

ar(ϕr,θr) = 1

√Nrh1,··· , e−jkd(psin ϕrsin θr+qcos θr),

··· , e−jkd((Nr,h−1) sin ϕrsin θr+(Nr,v −1) cos θr)i

(3)

where λdenotes the wavelength, and k=2π

λ.Nt,h and

Nt,v respectively denote the numbers of transmit antennas

in the horizontal and vertical directions, which satisfy Nt=

Nt,hNt,v. Similarly, Nr,h and Nr,v respectively denote the

numbers of receive antennas in the horizontal and vertical

directions, which satisfy Nr=Nr,hNr,v. d is the inner-

element spacing. pand qare the indices of elements, and

p= 0,1, . . . , Na,h −1,q= 0,1, . . . , Na,v −1,a∈ {t, r}.

Beneﬁting from the directional transmission and receive, the

AoD and AoA can be obtained by a beam training procedure

with a codebook. Speciﬁcally, the sensing signal is transmitted

and received by different beams, in which the combination

of the highest received SNR is used to estimate the angle

information. For convenience, we assume that the i-th unit-

norm codeword fiand the j-th unit-norm codeword wjof the

Kronecker-Product codebook Care selected by AP and STA

at time t, respectively. Then, the signal received by STA can

be expressed as

yt=√ptwH

jHfixt+wH

jnt(4)

where ptand xtrepresent the transmit power and signal of

AP at time t.ntis the independent and identically distributed

noise vector, and each element has 0mean and σ2variance.

Then, the SNR can be calculated as

SNRt=pt

wH

jHfi



σ2(5)

After obtaining the AoA and AoD, the ToF of the propaga-

tion path could be further estimated by FTM with centimeter

文档加载中……请稍候！
如果长时间未打开，您也可以点击刷新试试。

下载文档到电脑，查找使用更方便

10 玖币 0人已下载

立即下载

摘要：

AnNLoS-basedEnhancedSensingMethodforMmWaveCommunicationSystemShiwenHeyz,KangliCai,ShiyueHuang,ZhenyuAnz,WeiHuangx,NingGao{TheSchoolofComputerScienceandEngineering,CentralSouthUniversity,Changsha410083,China.yTheNationalMobileCommunicationsResearchLaboratory,SoutheastUniversity,Nanjing210096,Chin...

展开>> 收起<<

An NLoS-based Enhanced Sensing Method for MmWave Communication System Shiwen Heyz Kangli Cai Shiyue Huang Zhenyu Anz Wei Huangx Ning Gao.pdf

共7页,预览2页

还剩页未读，继续阅读

声明：本站为文档C2C交易模式，即用户上传的文档直接被用户下载，本站只是中间服务平台，本站所有文档下载所得的收益归上传人(含作者)所有。玖贝云文库仅提供信息存储空间，仅对用户上传内容的表现方式做保护处理，对上载内容本身不做任何修改或编辑。若文档所含内容侵犯了您的版权或隐私，请立即通知玖贝云文库，我们立即给予删除！

An NLoS-based Enhanced Sensing Method for MmWave Communication System Shiwen Heyz Kangli Cai Shiyue Huang Zhenyu Anz Wei Huangx Ning Gao

相关推荐

开通VIP享超值会员特权

作者详情

相关内容

热门标签

举报选择: