CN112218303A - Signal conversion method based on Wi-Fi identification system - Google Patents

Signal conversion method based on Wi-Fi identification system Download PDF

Info

Publication number
CN112218303A
CN112218303A CN202011043349.5A CN202011043349A CN112218303A CN 112218303 A CN112218303 A CN 112218303A CN 202011043349 A CN202011043349 A CN 202011043349A CN 112218303 A CN112218303 A CN 112218303A
Authority
CN
China
Prior art keywords
signal
receiving end
arm
user
distance
Prior art date
Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
Granted
Application number
CN202011043349.5A
Other languages
Chinese (zh)
Other versions
CN112218303B (en
Inventor
龙水
卢立
俞嘉地
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
Shanghai Jiaotong University
Original Assignee
Shanghai Jiaotong University
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by Shanghai Jiaotong University filed Critical Shanghai Jiaotong University
Priority to CN202011043349.5A priority Critical patent/CN112218303B/en
Publication of CN112218303A publication Critical patent/CN112218303A/en
Application granted granted Critical
Publication of CN112218303B publication Critical patent/CN112218303B/en
Active legal-status Critical Current
Anticipated expiration legal-status Critical

Links

Images

Classifications

    • H—ELECTRICITY
    • H04—ELECTRIC COMMUNICATION TECHNIQUE
    • H04W—WIRELESS COMMUNICATION NETWORKS
    • H04W24/00—Supervisory, monitoring or testing arrangements
    • H—ELECTRICITY
    • H04—ELECTRIC COMMUNICATION TECHNIQUE
    • H04W—WIRELESS COMMUNICATION NETWORKS
    • H04W4/00—Services specially adapted for wireless communication networks; Facilities therefor
    • H04W4/80—Services using short range communication, e.g. near-field communication [NFC], radio-frequency identification [RFID] or low energy communication

Landscapes

  • Engineering & Computer Science (AREA)
  • Computer Networks & Wireless Communication (AREA)
  • Signal Processing (AREA)
  • User Interface Of Digital Computer (AREA)

Abstract

A signal conversion method based on a Wi-Fi identification system is characterized in that a geometrical relation of space position change of a receiving end and a propagation characteristic of Wi-Fi are modeled, so that a CSI signal acquired by the moving receiving end is directly converted into a signal received by Wi-Fi receiving equipment at an original position, and the equipment position independence of Wi-Fi action identification is achieved. According to the Wi-Fi gesture recognition method, propagation characteristics of Wi-Fi signals are utilized, through mathematical modeling and basic transformation, receivers placed at other positions can estimate the gesture of human body actions by means of collected signals, the signals are finally converted into signals which a user should receive when facing Wi-Fi equipment, and finally the effect that the user actions can be recognized by the equipment at almost any position is achieved.

Description

Signal conversion method based on Wi-Fi identification system
Technical Field
The invention relates to a technology in the field of wireless communication, in particular to a signal conversion method based on a Wi-Fi identification system.
Background
With the rapid development of internet of things (IoT) technology and smart home in recent years, researchers have conducted extensive research on the related work of identifying and authenticating human actions and further controlling smart home by using Wi-Fi devices. In a Wi-Fi based motion recognition system, a user station typically performs motion recognition and verification at a fixed location facing a transmitting end and a receiving end of a Wi-Fi device. However, due to changes in the home environment, the location of the Wi-Fi device may change accordingly, and although the user still stands at the same location for authentication, the received signal is completely different due to the change in the location of the Wi-Fi receiver.
The channel state information CSI refers to channel properties of a communication link, which describe the variation of signals on different subcarrier transmission paths, such as multipath effects, propagation attenuation, signal scattering, signal reflection, etc. Thus, changes in body posture can often result in changes in the CSI data received by the receiving end. Since different actions have different effects on the CSI, the human body action can be identified by using the change of the CSI. But at present, the user action recognition in most indoor environments needs the user and the Wi-Fi recognition equipment to be in fixed positions to achieve a good recognition effect. The need to retrain the motion recognition model if the device location changes greatly increases the user's use threshold, making commercialization of Wi-Fi recognition devices difficult. Therefore, a method for identifying a user action based on the independence of the Wi-Fi device position is particularly important.
Disclosure of Invention
Aiming at the defects of the existing Wi-Fi-based human body gesture recognition system, the invention provides a signal conversion method based on a Wi-Fi recognition system, which utilizes the propagation characteristics of Wi-Fi signals, enables receivers placed at other positions to estimate the human body action gesture by utilizing collected signals and finally converts the signals into signals which a user should receive when facing Wi-Fi equipment by utilizing the collected signals, and finally achieves the effect that the equipment can recognize the user action at almost any position.
The invention is realized by the following technical scheme:
the invention relates to a signal conversion method based on a Wi-Fi identification system, which directly converts a CSI signal acquired by a moving receiving end into a signal received by Wi-Fi receiving equipment at an original position by modeling the geometric relation of the space position change of the receiving end and the propagation characteristic of Wi-Fi, thereby realizing the equipment position independence of Wi-Fi action identification.
The invention relates to a system for realizing the method, which comprises the following steps: signal acquisition unit, signal conversion unit and signal action classification unit, wherein: the signal acquisition unit is responsible for collecting action signals reflected by a human body and transmitting the acquired signals into the signal conversion unit; the signal conversion unit outputs the converted signal to the signal action classification unit through signal differentiation, parameter solving through polynomial fitting and signal conversion, and the signal action classification unit carries out signal conversion and identification irrelevant to the position of equipment.
Technical effects
The invention integrally solves the problem that in the prior art, the position of the Wi-Fi equipment is possibly changed along with the position of a home, and at the moment, the signal from the same action received by a receiving end is greatly different from the signal at the original position due to the change of the position of the equipment, so that the action cannot be identified.
Compared with the prior art, the method utilizes the equipment position information and the Wi-Fi signal propagation characteristic to carry out modeling, and converts the signal actions of the receiving ends at different positions to ensure that the converted signals are consistent with the signals which the receiving end at the initial position should receive. The position of the signal receiving equipment can be changed at will theoretically, and the flexibility and the expansibility of the motion recognition system are greatly improved.
Drawings
FIG. 1 is a flow chart of the present invention;
FIG. 2 is a diagram illustrating an exemplary linear motion of a user;
FIG. 3 is a diagram of an embodiment of user non-linear behavior;
FIG. 4 is a diagram illustrating an exemplary signal conversion process.
Detailed Description
As shown in fig. 1, the present embodiment relates to a signal conversion method based on a Wi-Fi recognition system, which is based on a Wi-Fi signal conversion manner, senses a human motion by using Channel State Information (CSI) data collected by a Wi-Fi receiver, and achieves a purpose of signal conversion by using a propagation characteristic of a Wi-Fi signal, and the specific steps include:
step 1) modeling the reflection of the signal by the arm according to the propagation characteristic of the wireless signal, which specifically comprises the following steps:
1.1) when the Wi-Fi signal reaches the signal receiving end after being reflected by a unit object, the method comprises the following steps: y (f, t) ═ a (f, t) e-j2πftWherein: t is the propagation time of the signal from the unit object to the receiving end, f is the carrier frequency, a (f, t) is the signal attenuation, and Y (f, t) is the received signal;
1.2) when the propagation distance at the moment t is D (t), according to the inverse square law,
Figure BDA0002707282320000021
where k is the propagation coefficient, and a (f, 0) ═ x (f) is the signal at the object at the beginning, so that the received signal is reduced to:
Figure BDA0002707282320000022
Figure BDA0002707282320000023
and the channel frequency response is obtained as:
Figure BDA0002707282320000024
i.e., the expression of signal state information in an ideal state vacuum.
1.3) because the signal of the Wi-Fi identification system is reflected by the whole arm of the human body and then reaches a receiving end instead of an object with unit length, the movement of the linear object which is regarded as the process of arm movement needs to be modeled: take linear motion as an example. Fig. 2 is a schematic diagram illustrating the linear motion of the arm at the speed v until the arm moves for the time t, and the moving distance d of the arm is vt; by modeling the palm (i.e., the unit object). According to simple geometric relation calculation, the following results are obtained: d2sinα＝D1sinθ，D2cosα＝D1cos θ -d, wherein: d is the distance the arm moves within time t, D1Is the distance from the palm to the receiving end at the beginning, D2To the hand at time tThe distance between the palm and the receiving end, theta is an included angle between the movement direction of the arm and a connecting line between the palm and the receiving end at the initial time, and alpha is an included angle between the movement direction of the arm at the time t and the connecting line between the palm and the receiving end; elimination of alpha to give
Figure BDA0002707282320000031
Figure BDA0002707282320000032
Further obtain
Figure BDA0002707282320000033
I.e., the channel state equation for the palm motion.
1.4) to calculate the channel state equation of the whole arm, the channel state information reflected by the whole arm is superposed
Figure BDA0002707282320000034
Namely, the channel frequency response equation of the whole arm motion is specifically: first, t in the equation is derived:
Figure BDA0002707282320000035
further simplifying to:
Figure BDA0002707282320000036
Figure BDA0002707282320000037
because the existing Wi-Fi-based motion recognition model mostly only utilizes amplitude and does not consider the influence of phase, because | | e -j2πft1, so the cancellation of the influence of phase is followed by:
Figure BDA0002707282320000038
1.5) for most non-linear motions, it can be converted into several linear motion processes. Since most non-linear motion can be considered linear motion within a relatively small period of time. FIG. 3 is a schematic diagram of the nonlinear operation, which is obtained by selecting two time nodes i and i +1 with short intervals, and the classSimilarly, the following are obtained:
Figure BDA0002707282320000039
Figure BDA00027072823200000310
linear motion and non-linear motion can be handled in the same way.
Step 2) signal conversion: as shown in fig. 4, the user performing motion recognition usually stands on the midperpendicular between the signal transmitting end (Tx) and the signal receiving end (Rx), and the differentiated channel state information is as follows according to the geometrical relationship in the figure
Figure BDA00027072823200000311
Wherein: k is the propagation factor, v is the velocity of movement, DRxIs the distance, θ, of the user to the receiving end of the initial positionRxThe included angle between the moving direction of the user and the distance from the user to the receiving end when the receiving end is at the initial position; along with the change of the family environment, the receiving end moves to other positions, at the moment, the CSI signal received by the moving receiving end changes, and at the moment, the received CSI differential of the receiving end is obtained as follows:
Figure BDA00027072823200000312
wherein: dSpIs the distance, θ, of the user to the receiving end after the moving positionSpThe unknown parameter is k, and the angle between the moving direction of the user and the receiving end after the user arrives at the moving position is obtained by differentiating the signal of the receiving end after moving and then carrying out polynomial fitting, v, DSp，θSpCan be measured to obtain the position information D in the initial stateRx，θRxIt can also be obtained by measurements, i.e. converting the differential signal received at the new position into the differential signal in the initial state:
Figure BDA0002707282320000041
Figure BDA0002707282320000042
i.e. from dHSp(f, t) to dHRx(f, t); finally by mixing dHRxIntegration of (f, t) yields HRx(f，t)。
In this embodiment, two desktop computers equipped with Intel 5300 wireless network cards and running Ubuntu 12.04 operating systems are used as a sending end and a receiving end, and a Support Vector Machine (SVM) is used as a signal classification mode. Experiments show that even if the signal receiving equipment moves to change the action signals of the user, the signal sorting machine can still correctly identify the corresponding arm action types after the signal receiving equipment is processed by the method.
Compared with the existing recognition system, the recognition method has the advantages that the motion recognition irrelevant to the position of the equipment can be realized, and even if the receiving equipment moves, the accurate recognition of the motion can be still carried out by the aid of the conversion method classifier, so that the method is a method which is never proposed in documents before. Is a beneficial addition to and a great improvement over existing Wi-Fi device based identification systems.
The foregoing embodiments may be modified in many different ways by those skilled in the art without departing from the spirit and scope of the invention, which is defined by the appended claims and all changes that come within the meaning and range of equivalency of the claims are therefore intended to be embraced therein.

Claims (4)

1. A signal conversion method based on a Wi-Fi identification system is characterized in that a geometrical relation of space position change of a receiving end and a propagation characteristic of Wi-Fi are modeled, so that a CSI signal acquired by the moving receiving end is directly converted into a signal received by Wi-Fi receiving equipment at an original position, and the equipment position independence of Wi-Fi action identification is achieved.
2. The signal conversion method based on the Wi-Fi recognition system according to claim 1, further comprising:
step 1) modeling the reflection of the signal by the arm according to the propagation characteristic of the wireless signal;
step 2) signal conversion: user presence information for action recognitionOn the vertical line between the signal transmitting end (Tx) and the signal receiving end (Rx), the differentiated channel state information is
Figure FDA0002707282310000011
Wherein: k is the propagation factor, v is the velocity of movement, DRxIs the distance, θ, of the user to the receiving end of the initial positionRxThe included angle between the moving direction of the user and the distance from the user to the receiving end when the receiving end is at the initial position; along with the change of the family environment, the receiving end moves to other positions, at the moment, the CSI signal received by the moving receiving end changes, and at the moment, the received CSI differential of the receiving end is obtained as follows:
Figure FDA0002707282310000012
Figure FDA0002707282310000013
wherein: dSpIs the distance, θ, of the user to the receiving end after the moving positionSpThe unknown parameter is k, and the angle between the moving direction of the user and the receiving end after the user arrives at the moving position is obtained by differentiating the signal of the receiving end after moving and then carrying out polynomial fitting, v, DSp，θSpMeasured to obtain the position information D in the initial stateRx，θRxThe differential signal received at the new position is converted into the differential signal at the initial state through measurement:
Figure FDA0002707282310000014
Figure FDA0002707282310000015
i.e. from dHSp(f, t) to dHRx(f, t); finally by mixing dHRxIntegration of (f, t) yields HRx(f，t)。
3. The signal conversion method based on the Wi-Fi recognition system according to claim 1 or 2, wherein the modeling specifically includes:
1.1) when the Wi-Fi signal reaches the signal receiving end after being reflected by a unit object, the method comprises the following steps: y (f, t) ═ a (f, t) e-j2πftWherein: t is the propagation time of the signal from the unit object to the receiving end, f is the carrier frequency, a (f, t) is the signal attenuation, and Y (f, t) is the received signal;
1.2) according to the propagation distance D (t) at the moment t and the inverse square law,
Figure FDA0002707282310000016
where k is the propagation coefficient, and a (f, 0) ═ x (f) is the signal at the object at the beginning, so that the received signal is reduced to:
Figure FDA0002707282310000017
Figure FDA0002707282310000021
and the channel frequency response is obtained as:
Figure FDA0002707282310000022
namely an expression of signal state information in an ideal state vacuum;
1.3) because the signal of the Wi-Fi identification system is reflected by the whole arm of the human body and then reaches a receiving end instead of an object with unit length, the movement of the linear object which is regarded as the process of arm movement needs to be modeled: the arm does linear motion at a speed v until the arm moves for a time t, wherein the moving distance d is equal to vt; by modeling the palm, i.e. the unit object, we obtain: d2sinα＝D1sinθ，D2cosα＝D1cos θ -d, wherein: d is the distance the arm moves within time t, D1Is the distance from the palm to the receiving end at the beginning, D2The distance from the palm to the receiving end at the time t, theta is an included angle between the moving direction of the arm and a connecting line from the palm to the receiving end at the initial time, and alpha is an included angle between the moving direction of the arm at the time t and the connecting line from the palm to the receiving end; elimination of alpha to give
Figure FDA0002707282310000023
Figure FDA0002707282310000024
Further obtain
Figure FDA0002707282310000025
Namely the channel state equation of the palm motion;
1.4) superimposing the channel state information reflected by the whole arm
Figure FDA0002707282310000026
Namely, the channel frequency response equation of the whole arm motion is specifically: first, t in the equation is derived:
Figure FDA0002707282310000027
Figure FDA0002707282310000028
further simplifying to:
Figure FDA0002707282310000029
because the existing Wi-Fi-based motion recognition model mostly only utilizes amplitude and does not consider the influence of phase, because | | e-j2πft1, so the cancellation of the influence of phase is followed by:
Figure FDA00027072823100000210
1.5) selecting two time nodes i and i +1 with shorter intervals to obtain:
Figure FDA00027072823100000211
Figure FDA00027072823100000212
so that linear motion and non-linear motion are treated in the same way.
4. A system for implementing the method of any preceding claim, comprising: signal acquisition unit, signal conversion unit and signal action classification unit, wherein: the signal acquisition unit is responsible for collecting action signals reflected by a human body and transmitting the acquired signals into the signal conversion unit; the signal conversion unit outputs the converted signal to the signal action classification unit through signal differentiation, parameter solving through polynomial fitting and signal conversion, and the signal action classification unit carries out signal conversion and identification irrelevant to the position of equipment.
CN202011043349.5A 2020-09-28 2020-09-28 Signal conversion method based on Wi-Fi identification system Active CN112218303B (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN202011043349.5A CN112218303B (en) 2020-09-28 2020-09-28 Signal conversion method based on Wi-Fi identification system

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN202011043349.5A CN112218303B (en) 2020-09-28 2020-09-28 Signal conversion method based on Wi-Fi identification system

Publications (2)

Publication Number Publication Date
CN112218303A true CN112218303A (en) 2021-01-12
CN112218303B CN112218303B (en) 2022-02-18

Family

ID=74051312

Family Applications (1)

Application Number Title Priority Date Filing Date
CN202011043349.5A Active CN112218303B (en) 2020-09-28 2020-09-28 Signal conversion method based on Wi-Fi identification system

Country Status (1)

Country Link
CN (1) CN112218303B (en)

Citations (7)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US20130279488A1 (en) * 2012-04-18 2013-10-24 Telcom Ventures, L.L.C. Systems and methods for local-area-network-assisted location determination
CN105573498A (en) * 2015-12-15 2016-05-11 西北大学 Gesture recognition method based on Wi-Fi signal
US20160157201A1 (en) * 2014-12-02 2016-06-02 Cisco Technology, Inc. Wideband Angle-Of-Arrival Location Determination Using Bandwidth Partitioning
CN108901021A (en) * 2018-05-31 2018-11-27 大连理工大学 A kind of deep learning identification system and method based on channel state information of wireless network
CN110659598A (en) * 2019-09-12 2020-01-07 北京航空航天大学 Human body action layered analysis and identification method and device based on Wi-Fi signals
CN110929242A (en) * 2019-11-20 2020-03-27 上海交通大学 Method and system for carrying out attitude-independent continuous user authentication based on wireless signals
CN111123194A (en) * 2018-10-31 2020-05-08 慧与发展有限责任合伙企业 Accurately locating objects through network devices

Patent Citations (7)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US20130279488A1 (en) * 2012-04-18 2013-10-24 Telcom Ventures, L.L.C. Systems and methods for local-area-network-assisted location determination
US20160157201A1 (en) * 2014-12-02 2016-06-02 Cisco Technology, Inc. Wideband Angle-Of-Arrival Location Determination Using Bandwidth Partitioning
CN105573498A (en) * 2015-12-15 2016-05-11 西北大学 Gesture recognition method based on Wi-Fi signal
CN108901021A (en) * 2018-05-31 2018-11-27 大连理工大学 A kind of deep learning identification system and method based on channel state information of wireless network
CN111123194A (en) * 2018-10-31 2020-05-08 慧与发展有限责任合伙企业 Accurately locating objects through network devices
CN110659598A (en) * 2019-09-12 2020-01-07 北京航空航天大学 Human body action layered analysis and identification method and device based on Wi-Fi signals
CN110929242A (en) * 2019-11-20 2020-03-27 上海交通大学 Method and system for carrying out attitude-independent continuous user authentication based on wireless signals

Non-Patent Citations (2)

* Cited by examiner, † Cited by third party
Title
MOHAMMED A.A. AL-QANESS: "Indoor Micro-Activity Recognition Method Using Ubiquitous WiFi Devices", 《2018 UBIQUITOUS POSITIONING, INDOOR NAVIGATION AND LOCATION-BASED SERVICES (UPINLBS)》 *
陈泳希: "基于WIFI传播模型的室内定位技术研究", 《中国优秀硕士学位论文全文数据库 信息科技辑 I136-1175》 *

Also Published As

Publication number Publication date
CN112218303B (en) 2022-02-18

Similar Documents

Publication Publication Date Title
Virmani et al. Position and orientation agnostic gesture recognition using WiFi
Yun et al. Strata: Fine-grained acoustic-based device-free tracking
Wu et al. FingerDraw: Sub-wavelength level finger motion tracking with WiFi signals
He et al. WiG: WiFi-based gesture recognition system
Wang et al. Device-free simultaneous wireless localization and activity recognition with wavelet feature
CN101730851B (en) System and method for positioning
WO2020258106A1 (en) Gesture recognition method and device, and positioning and tracking method and device
CN113609976B (en) Direction-sensitive multi-gesture recognition system and method based on WiFi equipment
Zhang et al. WiFi-based cross-domain gesture recognition via modified prototypical networks
CN110069134B (en) Method for restoring aerial moving track of hand by using radio frequency signal
Wang et al. CSI-based human sensing using model-based approaches: a survey
CN112733609B (en) Domain-adaptive Wi-Fi gesture recognition method based on discrete wavelet transform
Meng et al. WiHGR: A robust WiFi-based human gesture recognition system via sparse recovery and modified attention-based BGRU
Chen et al. Dynamic gesture recognition using wireless signals with less disturbance
CN103971108A (en) Wireless communication-based human body posture recognition method and device
Ren et al. Intelligent contactless gesture recognition using WLAN physical layer information
CN112218303B (en) Signal conversion method based on Wi-Fi identification system
Regani et al. GWrite: Enabling through-the-wall gesture writing recognition using WiFi
Zhang et al. WiRITE: General and Practical Wi-Fi Based Hand-Writing Recognition
Guo et al. WiRoI: Spatial region of interest human sensing with commodity WiFi
Dang et al. Air gesture recognition using WLAN physical layer information
WO2024021379A1 (en) Wireless signal technology-based diaper humidity measurement method
CN116680597A (en) Omnidirectional gesture sensing and identifying method based on Wi-Fi
CN111257589A (en) Wind speed measuring method based on CRFID (cross-reference frequency identification) label
CN105738866A (en) 60GHz Non-Line-of-Sight identification and wireless fingerprint positioning method based on energy detection

Legal Events

Date Code Title Description
PB01 Publication
PB01 Publication
SE01 Entry into force of request for substantive examination
SE01 Entry into force of request for substantive examination
GR01 Patent grant
GR01 Patent grant