无传感器运动感应与自更新指纹技术提升室内围栏定位精度

30 浏览量更新于2024-07-14 收藏 3.22MB PDF 举报

本文主要探讨了"基于无传感器运动感应和指纹自更新的室内围墙"这一主题，发表在《移动网络与应用》(Mobile Networks and Applications)上，DOI为10.1007/s11036-019-01329-0。作者是Kun Zhao、Wei Xi、Zhiping Jiang等人，该研究发表于2019年，由Springer Science+Business Media, LLC, part of Springer Nature出版。室内定位是现代技术中的关键特性，尤其是在地理围栏应用中具有巨大的价值。传统的接收信号强度（RSS）指纹识别方法广泛用于判断用户接近特定区域的程度，然而这种方法的动态性和维护成本一直是主要挑战。论文提出了一种结合自更新RSS指纹和信道状态信息（CSI）运动检测的混合电子围栏解决方案。通过多维度匹配和过滤原理，实现RSS指纹的自我更新，从而提高定位精度。 CSI（信道状态信息）被用于速度估计，这有助于减少定位频率和总体开销。作者强调，他们的实验结果显示，提出的室内地理围栏方法在实际环境中表现优异，对于超过3个维度的复杂环境，其性能稳定且有效。这种方法不仅提高了位置确定的准确性，还降低了系统的实时监控需求，对于物联网(IoT)设备的能耗管理和用户体验优化具有重要意义。此研究对于无线通信、智能家居、智能建筑等领域具有潜在的应用前景，它展示了如何利用先进的信号处理技术和机器学习算法来解决室内定位问题，提高系统效率和用户感知。未来的研究可能进一步探索如何将这种自适应围栏技术与其他传感器融合，或者在更大规模的室内环境中进行优化。这项工作对提升室内定位技术的实用性和可靠性有着积极的推动作用。

Mobile Netw Appl

Table 1 A glance of unpacked frames

Record time(HH:MM:SS) Frequency RSS BSSID Source address Destination address

06:53:31.148619 5500 MHz -−57 dB 30:5a:3a:cb:16:a0 00:16:ea:12:34:56 74:e5:43:5e:fd:a9

06:53:31.153621 5500 MHz −57 dB 30:5a:3a:cb:16:a0 00:16:ea:12:34:56 74:e5:43:5e:fd:a9

06:53:31.158509 5500 MHz −-57 dB 30:5a:3a:cb:16:a0 00:16:ea:12:34:56 74:e5:43:5e:fd:a9

··· ··· ··· ··· ··· ···

06:53:38.141203 5500 MHz −58 dB 30:5a:3a:cb:16:a0 00:16:ea:12:34:56 74:e5:43:5e:fd:a9

··· ··· ··· ··· ··· ···

06:53:45.722784 5500 MHz −57 dB 30:5a:3a:cb:16:a0 00:16:ea:12:34:56 74:e5:43:5e:fd:a9

that have been added to the kernel recently. For instance,

we can use command iw phy phy0 interface add

mon0 type monitor to create a new interface mon0

to set phy0 into monitor mode. In addition, we can use

command iw dev mon0 set freq f

to switch the

center frequency of mon0 to f

MHz. These operations will

only spend several milliseconds. In motion detection phase

(see Fig. 2), the client keeps normal communication to APs

and obtain CSIs from preambles. When the system detects

object’s abnormal moving (see details in Section 4.2),

the client’s NIC card will change to monitor mode and

record all the packets heard from ambient APs. Table 1

lists the unpacked packets information using single channel

(Channel 100 whose f

= 5500 MHz). The table lists

the packets information as follow, time stamp indicated

by the Timing synchronization function,

channel center

frequency, RSS value, the identifier of basic service set,

transmitter’s MAC (source) address, and receiver’s MAC

(destination) address, respectively.

As shown in Table 1, SDF-based solutions expose two

main drawbacks: 1) In training phase, we have no need to

record such amount of RSS values in a single channel,

whereas we should leverage this opportunity to obtain

more abundant, and meaningful fingerprints; 2) In matching

phase, since users cannot endure long waiting cycles to

obtain a stable RSS numerical characteristics, e.g., mean or

median, calculating the distance between collected samples

average or median and the mean or median in fingerprint

database would result some localization errors. Hence, we,

in this paper, first propose to leverage multiple channel

fingerprint to improve location accuracy (detailed in

Section 4.1) and exploit sample hypothesis testing to verify

Timing synchronization function (TSF) is specified in IEEE 802.11

wireless local area network (WLAN) standard which is based on a

1-MHz clock and “tick” in microseconds [49].

In IEEE 802.11 wireless local area networking standards, basic

service sets (BSS) are units of devices operating with the same medium

access characteristics.

users’ location. In this way, we can not merely leverage all

the sampled RSS values, but also obtain the confidence of

localization precision (detailed in Section 4.6).

3.2Resolutiondiversity

Wireless devices, e.g., APs, have a discrimination capability

when fingerprinting a specific location. This capability,

however, is subject to inherent constraints of radio signal

propagation.

Discrimination capability is referred to as the ability

that an AP distinguishes a specific location when using

its RSS observations as fingerprints. Ideally, following the

propagation law of wireless signals, the RSS of a certain

wireless signal decays logarithmically with its propaga-

tion distance d. In other words, an identical ΔRSS can

imply a smaller distance change Δd at the location closer

to the antenna of AP, or a larger Δd at the location far

away from the AP’s antenna. As shown in Fig. 1,theRSS

variance shows vast difference in different physical posi-

tions, depending on the specific distance d. Hence, using

the RSS reading from a long-distance AP may cause a large

error in the location estimation, while using those from near

APs can conversely mitigate the error. In fact, such a RSS

variance also indicates a diversity of AP resolution across

different transmitter-receiver distances.

Fig. 1 AP resolution

剩余18页未读，继续阅读

weixin_38741540

粉丝: 6
资源: 960

无传感器运动感应与自更新指纹技术提升室内围栏定位精度

基于机器人的室内指纹地图更新.pdf

基于地磁传感器的室内定位.zip

三星将自己生产指纹传感器芯片

描述1000字AS608指纹传感器的特点和引脚功能和驱动原理

MG200指纹传感器工作原理

指纹传感器as608工作原理

在门禁系统中，指纹传感器、射频传感器工作原理？

指纹室内定位代码 stm32

arduino指纹传感器

最新资源