WINLAB Chenren Xu Joint work with Bernhard Firner, Yanyong Zhang - - PowerPoint PPT Presentation

▶
winlab chenren xu joint work with bernhard firner yanyong
SMART_READER_LITE
LIVE PREVIEW

WINLAB Chenren Xu Joint work with Bernhard Firner, Yanyong Zhang - - PowerPoint PPT Presentation

Feb 19, 2023 447 likes •1.13k views

Improving RF-Based Device-Free Passive Localization Through Probabilistic Classification Methods Fall 2012 Research Review WINLAB Chenren Xu Joint work with Bernhard Firner, Yanyong Zhang Richard Howard, Jun Li, Xiaodong Lin Pervasive Radio

slide-1
SLIDE 1

WINLAB Improving RF-Based Device-Free Passive Localization Through Probabilistic Classification Methods

Fall 2012 Research Review Chenren Xu

Joint work with Bernhard Firner, Yanyong Zhang Richard Howard, Jun Li, Xiaodong Lin

slide-2
SLIDE 2

WINLAB

Pervasive Radio Space

2

slide-3
SLIDE 3

WINLAB

RF-Based Localization

3

Active Localization

slide-4
SLIDE 4

WINLAB

RF-Based Localization

4

slide-5
SLIDE 5

WINLAB

RF-Based Localization

5

Passive Localization

slide-6
SLIDE 6

WINLAB

Passive Localization

 Motivation  Indoor challenge  Proposed solution  Experimental methodology  Performance evaluation  Conclusion and future work

6

slide-7
SLIDE 7

WINLAB

Why Passive Localization?

 Monitor indoor human mobility

7

slide-8
SLIDE 8

WINLAB

Why Passive Localization?

 Monitor indoor human mobility

Elder/health care

8

slide-9
SLIDE 9

WINLAB

Why Passive Localization?

 Monitor indoor human mobility

Detect traffic flow

9

slide-10
SLIDE 10

WINLAB

Why Passive Localization?

 Monitor indoor human mobility

 Health/elder care, safety  Detect traffic flow

10

slide-11
SLIDE 11

WINLAB

Why Passive Localization?

 Monitor indoor human mobility

 Health/elder care, safety  Detect traffic flow

 Provides privacy protection

 No identification

11

slide-12
SLIDE 12

WINLAB

Why Passive Localization?

 Monitor indoor human mobility

 Health/elder care, safety  Detect traffic flow

 Provides privacy protection

 No identification

 Use existing wireless infrastructure

12

slide-13
SLIDE 13

WINLAB

Passive Localization

 Motivation  Indoor challenge  Proposed solution  Experimental methodology  Performance evaluation  Conclusion and future work

13

slide-14
SLIDE 14

WINLAB

Multipath Effect

14

Tx: Radio transmitter Rx: Radio receiver

slide-15
SLIDE 15

WINLAB

Multipath Effect

15

slide-16
SLIDE 16

WINLAB

Multipath Effect

16

slide-17
SLIDE 17

WINLAB

Multipath Effect

17

slide-18
SLIDE 18

WINLAB

Multipath Effect

18

slide-19
SLIDE 19

WINLAB

Multipath Effect

19

slide-20
SLIDE 20

WINLAB

Multipath Effect

20

slide-21
SLIDE 21

WINLAB

Multipath Effect

21

slide-22
SLIDE 22

WINLAB

Cluttered Indoor Scenario

22

 Find a cluttered indoor environments…

slide-23
SLIDE 23

WINLAB

Cluttered Indoor Scenario

23

slide-24
SLIDE 24

WINLAB

Cluttered Indoor Scenario

24

A user steps across one Line-of-Sight

slide-25
SLIDE 25

WINLAB

Cluttered Indoor Scenario

25

A user steps across the Line-of-Sight

slide-26
SLIDE 26

WINLAB

Cluttered Indoor Scenario

26

A user steps across

  • ne Line-of-Sight

RSS fluctuates in a unpredictable fashion

slide-27
SLIDE 27

WINLAB

Cluttered Indoor Scenario

27

The RSS change can either go up to 12 dBm

slide-28
SLIDE 28

WINLAB

Cluttered Indoor Scenario

28

Or go down to -12 dBm

slide-29
SLIDE 29

WINLAB

Cluttered Indoor Scenario

29

These two peak points can have 24 dB difference in energy within only 2 meters.

slide-30
SLIDE 30

WINLAB

Cluttered Indoor Scenario

30

We also observe that another two points within 0.2 m can have 15 dB difference.

slide-31
SLIDE 31

WINLAB

Cluttered Indoor Scenario

31

We also observe that these two points within 0.2 m can have 15 dB difference. Deep fade

slide-32
SLIDE 32

WINLAB

Cluttered Indoor Scenario

32

slide-33
SLIDE 33

WINLAB

Cluttered Indoor Scenario

33

slide-34
SLIDE 34

WINLAB

Cluttered Indoor Scenario

34

slide-35
SLIDE 35

WINLAB

Passive Localization

 Motivation  Indoor challenge  Proposed solution  Experimental methodology  Performance evaluation  Conclusion and future work

35

slide-36
SLIDE 36

WINLAB

Proposed Solution

 High dimensional space

 Measure radio signal strength (RSS) changes in

multiple transmitter and receiver links.

36

slide-37
SLIDE 37

WINLAB

Proposed Solution

 High dimensional space

 Measure radio signal strength (RSS) changes in

multiple transmitter and receiver links.

37 Link T1 – R1 Link T2 – R2

slide-38
SLIDE 38

WINLAB

Proposed Solution

 High dimensional space  Cell-based localization

 Flexible precision  Classification approach

38

slide-39
SLIDE 39

WINLAB

Linear Discriminant Analysis

 RSS measurements with user’s presence in each cell

is treated as a class k

 Each class k is Multivariate Gaussian with common

covariance

 Linear discriminant function:

39

Link 1 RSS (dBm) Link 2 RSS (dBm) k = 1 k = 2 k = 3

slide-40
SLIDE 40

WINLAB

Proposed Solution

 High dimensional space  Cell-based localization  Lower radio frequency

 Smooth the spatial variation

40

slide-41
SLIDE 41

WINLAB

Frequency Impact

41

RSS changes smoother on 433.1 MHz than on 909.1 MHz

slide-42
SLIDE 42

WINLAB

Frequency Impact

42

Less deep fading points!

slide-43
SLIDE 43

WINLAB

Proposed Solution

 High dimensional space

 Find features with fewer deep fading points

 Cell-based localization

 Average the deep fading effect

 Lower radio frequency

 Reduce the deep fading points

43

slide-44
SLIDE 44

WINLAB

Proposed Solution

 High dimensional space

 Find features with fewer deep fading points

 Cell-based localization

 Average the deep fading effect

 Lower radio frequency

 Reduce the deep fading points

44

Mitigate the error caused by the multipath effect!

slide-45
SLIDE 45

WINLAB

Passive Localization

 Motivation  Indoor challenge  Proposed solution  Experimental methodology  Performance evaluation  Conclusion and future work

45

slide-46
SLIDE 46

WINLAB

Experimental Deployment

46

Total Size: 5 × 8 m

slide-47
SLIDE 47

WINLAB

Experimental Deployment

47

slide-48
SLIDE 48

WINLAB

System Parameters

48

Parameter Default value Meaning K 32 Number of cells P 64 Number of pair-wise radio links Ntrn 100 Number of training data per cell Ntst 100 Number of testing data per cell

slide-49
SLIDE 49

WINLAB

System Description

 Hardware: PIP tag

 Microprocessor: C8051F321  Radio chip: CC1100  Power: Lithium coin cell battery (~1 year)

 Protocol: Unidirectional heartbeat (Uni-HB)

 Packet size: 10 bytes  Beacon interval: 100 millisecond

49

slide-50
SLIDE 50

WINLAB

Training Methodology

 Case A: stand still at the each cell center

 Measurement only involves center of the cell  Ignore the deep fade points

 Case B: random walk within each cell

 Measurement includes all the space  Average the multi-path effects

50

slide-51
SLIDE 51

WINLAB

Training Methodology

 Case A: stand still at the each cell center

 Measurement only involves center of the cell  Ignore the deep fade points

 Case B: random walk within each cell

 Measurement includes all the space  Average the multi-path effects

51

Training only takes 15 mins!

slide-52
SLIDE 52

WINLAB

Passive Localization

 Motivation  Indoor challenge  Proposed solution  Experimental methodology  Performance evaluation  Conclusion and future work

52

slide-53
SLIDE 53

WINLAB

Metrics

 Cell estimation accuracy

 The ratio of successful cell estimations with

respect to the total number of estimations.

 Average error distance

 Average distance between the actual location and

the estimated cell’s center.

53

slide-54
SLIDE 54

WINLAB

Localization Accuracy

 Cell estimation accuracy:

54

Stand still at each cell center Random walk with in each cell 433.1 MHz 90.1% 97.2% 909.1 MHz 82.9% 93.8%

97.2 % cell estimation accuracy with 0.36 m average error distance

slide-55
SLIDE 55

WINLAB

Reducing Training Dataset

55

100

8

Only 8 samples are good enough

slide-56
SLIDE 56

WINLAB

Robust to Link Failure

56

5 transmitter + 3 receivers = 90% cell estimation accuracy

slide-57
SLIDE 57

WINLAB

Multiple Subjects Localization

57

slide-58
SLIDE 58

WINLAB

Larger Deployment

58

Total Size: 10 × 15 m Cell Size: 2 × 2 m 13 transmitters and 9 receivers

slide-59
SLIDE 59

WINLAB

Larger Deployment

59

Cell estimation accuracy: 93.8% Average error distance: 1.3 m

slide-60
SLIDE 60

WINLAB

Passive Localization

 Motivation  Indoor challenge  Proposed solution  Experimental methodology  Performance evaluation  Conclusion and future work

60

slide-61
SLIDE 61

WINLAB

Conclusion and Future Work

 Conclusion

 We propose a general probabilistic classification framework

to solve the passive localization problem with:

 High accuracy, low cost, and robust  Multiple subjects tracking generalization

 Future work

 Improving multiple people tracking  Passively detect the number of people

61

slide-62
SLIDE 62

WINLAB

Q & A

Thank you

62

slide-63
SLIDE 63

WINLAB

Classification algorithms

63

MED ignores deep fading QDA overfits training data

slide-64
SLIDE 64

WINLAB

Gaussian Approximation

64

(a) (b) (c)

RSS change (dBm)

slide-65
SLIDE 65

WINLAB

Principal Components

65

slide-66
SLIDE 66

WINLAB

Long-term Stability

66