Rutgers University Chenren Xu Joint work with Bernhard Firner , Yanyong Zhang
description
Transcript of Rutgers University Chenren Xu Joint work with Bernhard Firner , Yanyong Zhang
![Page 1: Rutgers University Chenren Xu Joint work with Bernhard Firner , Yanyong Zhang](https://reader035.fdocuments.net/reader035/viewer/2022062423/568144fc550346895db1c743/html5/thumbnails/1.jpg)
WINLAB
Improving RF-Based Device-Free Passive Localization In Cluttered Indoor Environments Through Probabilistic Classification Methods
Rutgers University
Chenren Xu
Joint work with
Bernhard Firner, Yanyong Zhang
Richard Howard, Jun Li, Xiaodong Lin
![Page 2: Rutgers University Chenren Xu Joint work with Bernhard Firner , Yanyong Zhang](https://reader035.fdocuments.net/reader035/viewer/2022062423/568144fc550346895db1c743/html5/thumbnails/2.jpg)
WINLAB2
Passive Localization
Motivation
Indoor challenge
Proposed solution
Experimental methodology
Performance evaluation
Conclusion and future work
![Page 3: Rutgers University Chenren Xu Joint work with Bernhard Firner , Yanyong Zhang](https://reader035.fdocuments.net/reader035/viewer/2022062423/568144fc550346895db1c743/html5/thumbnails/3.jpg)
WINLAB3
RF-Based Localization
Active Localization
![Page 4: Rutgers University Chenren Xu Joint work with Bernhard Firner , Yanyong Zhang](https://reader035.fdocuments.net/reader035/viewer/2022062423/568144fc550346895db1c743/html5/thumbnails/4.jpg)
WINLAB4
RF-Based Localization
![Page 5: Rutgers University Chenren Xu Joint work with Bernhard Firner , Yanyong Zhang](https://reader035.fdocuments.net/reader035/viewer/2022062423/568144fc550346895db1c743/html5/thumbnails/5.jpg)
WINLAB5
RF-Based Localization
Passive Localization
![Page 6: Rutgers University Chenren Xu Joint work with Bernhard Firner , Yanyong Zhang](https://reader035.fdocuments.net/reader035/viewer/2022062423/568144fc550346895db1c743/html5/thumbnails/6.jpg)
WINLAB6
Passive Localization
Motivation
Indoor challenge
Proposed solution
Experimental methodology
Performance evaluation
Conclusion and future work
![Page 7: Rutgers University Chenren Xu Joint work with Bernhard Firner , Yanyong Zhang](https://reader035.fdocuments.net/reader035/viewer/2022062423/568144fc550346895db1c743/html5/thumbnails/7.jpg)
WINLAB7
Why Passive Localization?
Monitor indoor human mobility
Elder/health care
![Page 8: Rutgers University Chenren Xu Joint work with Bernhard Firner , Yanyong Zhang](https://reader035.fdocuments.net/reader035/viewer/2022062423/568144fc550346895db1c743/html5/thumbnails/8.jpg)
WINLAB8
Why Passive Localization?
Monitor indoor human mobility
Detect traffic flow
![Page 9: Rutgers University Chenren Xu Joint work with Bernhard Firner , Yanyong Zhang](https://reader035.fdocuments.net/reader035/viewer/2022062423/568144fc550346895db1c743/html5/thumbnails/9.jpg)
WINLAB9
Why Passive Localization?
Monitor indoor human mobility
Health/elder care, safety
Detect traffic flow
Provides privacy protection
No identification
Use existing wireless infrastructure
![Page 10: Rutgers University Chenren Xu Joint work with Bernhard Firner , Yanyong Zhang](https://reader035.fdocuments.net/reader035/viewer/2022062423/568144fc550346895db1c743/html5/thumbnails/10.jpg)
WINLAB10
Passive Localization
Motivation
Indoor challenge
Proposed solution
Experimental methodology
Performance evaluation
Conclusion and future work
![Page 11: Rutgers University Chenren Xu Joint work with Bernhard Firner , Yanyong Zhang](https://reader035.fdocuments.net/reader035/viewer/2022062423/568144fc550346895db1c743/html5/thumbnails/11.jpg)
WINLAB11
Multipath Effect
![Page 12: Rutgers University Chenren Xu Joint work with Bernhard Firner , Yanyong Zhang](https://reader035.fdocuments.net/reader035/viewer/2022062423/568144fc550346895db1c743/html5/thumbnails/12.jpg)
WINLAB12
Multipath Effect
![Page 13: Rutgers University Chenren Xu Joint work with Bernhard Firner , Yanyong Zhang](https://reader035.fdocuments.net/reader035/viewer/2022062423/568144fc550346895db1c743/html5/thumbnails/13.jpg)
WINLAB13
Multipath Effect
![Page 14: Rutgers University Chenren Xu Joint work with Bernhard Firner , Yanyong Zhang](https://reader035.fdocuments.net/reader035/viewer/2022062423/568144fc550346895db1c743/html5/thumbnails/14.jpg)
WINLAB14
Cluttered Indoor Scenario
![Page 15: Rutgers University Chenren Xu Joint work with Bernhard Firner , Yanyong Zhang](https://reader035.fdocuments.net/reader035/viewer/2022062423/568144fc550346895db1c743/html5/thumbnails/15.jpg)
WINLAB15
Cluttered Indoor Scenario
A user steps across one Line-of-Sight
![Page 16: Rutgers University Chenren Xu Joint work with Bernhard Firner , Yanyong Zhang](https://reader035.fdocuments.net/reader035/viewer/2022062423/568144fc550346895db1c743/html5/thumbnails/16.jpg)
WINLAB16
Cluttered Indoor Scenario
A user steps across one Line-of-Sight
RSS fluctuates in a unpredictable fashion
![Page 17: Rutgers University Chenren Xu Joint work with Bernhard Firner , Yanyong Zhang](https://reader035.fdocuments.net/reader035/viewer/2022062423/568144fc550346895db1c743/html5/thumbnails/17.jpg)
WINLAB17
Cluttered Indoor Scenario
The RSS change can either go up to 12 dBm
![Page 18: Rutgers University Chenren Xu Joint work with Bernhard Firner , Yanyong Zhang](https://reader035.fdocuments.net/reader035/viewer/2022062423/568144fc550346895db1c743/html5/thumbnails/18.jpg)
WINLAB18
Cluttered Indoor Scenario
Or go down to -12 dBm
![Page 19: Rutgers University Chenren Xu Joint work with Bernhard Firner , Yanyong Zhang](https://reader035.fdocuments.net/reader035/viewer/2022062423/568144fc550346895db1c743/html5/thumbnails/19.jpg)
WINLAB19
Cluttered Indoor Scenario
These two peak points can have 24 dB difference in energy within only 2 meters.
![Page 20: Rutgers University Chenren Xu Joint work with Bernhard Firner , Yanyong Zhang](https://reader035.fdocuments.net/reader035/viewer/2022062423/568144fc550346895db1c743/html5/thumbnails/20.jpg)
WINLAB20
Cluttered Indoor Scenario
We also observe that these two points within 0.2 m can have 15 dB difference.
Deep fade
![Page 21: Rutgers University Chenren Xu Joint work with Bernhard Firner , Yanyong Zhang](https://reader035.fdocuments.net/reader035/viewer/2022062423/568144fc550346895db1c743/html5/thumbnails/21.jpg)
WINLAB21
Cluttered Indoor Scenario
![Page 22: Rutgers University Chenren Xu Joint work with Bernhard Firner , Yanyong Zhang](https://reader035.fdocuments.net/reader035/viewer/2022062423/568144fc550346895db1c743/html5/thumbnails/22.jpg)
WINLAB22
Cluttered Indoor Scenario
![Page 23: Rutgers University Chenren Xu Joint work with Bernhard Firner , Yanyong Zhang](https://reader035.fdocuments.net/reader035/viewer/2022062423/568144fc550346895db1c743/html5/thumbnails/23.jpg)
WINLAB23
Cluttered Indoor Scenario
![Page 24: Rutgers University Chenren Xu Joint work with Bernhard Firner , Yanyong Zhang](https://reader035.fdocuments.net/reader035/viewer/2022062423/568144fc550346895db1c743/html5/thumbnails/24.jpg)
WINLAB24
Passive Localization
Motivation
Indoor challenge
Proposed solution
Experimental methodology
Performance evaluation
Conclusion and future work
![Page 25: Rutgers University Chenren Xu Joint work with Bernhard Firner , Yanyong Zhang](https://reader035.fdocuments.net/reader035/viewer/2022062423/568144fc550346895db1c743/html5/thumbnails/25.jpg)
WINLAB25
Proposed Solution
High dimensional space
Measure radio signal strength (RSS) changes in
multiple transmitter and receiver links.
Link T1 – R1
Link T2 – R2
![Page 26: Rutgers University Chenren Xu Joint work with Bernhard Firner , Yanyong Zhang](https://reader035.fdocuments.net/reader035/viewer/2022062423/568144fc550346895db1c743/html5/thumbnails/26.jpg)
WINLAB26
Proposed Solution
High dimensional space
Cell-based localization
Flexible precision
Classification approach
![Page 27: Rutgers University Chenren Xu Joint work with Bernhard Firner , Yanyong Zhang](https://reader035.fdocuments.net/reader035/viewer/2022062423/568144fc550346895db1c743/html5/thumbnails/27.jpg)
WINLAB27
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:
Link 1 RSS (dBm)L
ink
2 R
SS
(d
Bm
)
k = 1k = 2
k = 3
![Page 28: Rutgers University Chenren Xu Joint work with Bernhard Firner , Yanyong Zhang](https://reader035.fdocuments.net/reader035/viewer/2022062423/568144fc550346895db1c743/html5/thumbnails/28.jpg)
WINLAB28
Proposed Solution
High dimensional space
Cell-based localization
Lower radio frequency
Smooth the spatial variation
![Page 29: Rutgers University Chenren Xu Joint work with Bernhard Firner , Yanyong Zhang](https://reader035.fdocuments.net/reader035/viewer/2022062423/568144fc550346895db1c743/html5/thumbnails/29.jpg)
WINLAB29
Frequency Impact
RSS changes smoother on 433.1 MHz than on 909.1 MHz
![Page 30: Rutgers University Chenren Xu Joint work with Bernhard Firner , Yanyong Zhang](https://reader035.fdocuments.net/reader035/viewer/2022062423/568144fc550346895db1c743/html5/thumbnails/30.jpg)
WINLAB30
Frequency Impact
Less deep fading points!
![Page 31: Rutgers University Chenren Xu Joint work with Bernhard Firner , Yanyong Zhang](https://reader035.fdocuments.net/reader035/viewer/2022062423/568144fc550346895db1c743/html5/thumbnails/31.jpg)
WINLAB31
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
Mitigate the error caused by the multipath effect!
![Page 32: Rutgers University Chenren Xu Joint work with Bernhard Firner , Yanyong Zhang](https://reader035.fdocuments.net/reader035/viewer/2022062423/568144fc550346895db1c743/html5/thumbnails/32.jpg)
WINLAB32
Passive Localization
Motivation
Indoor challenge
Proposed solution
Experimental methodology
Performance evaluation
Conclusion and future work
![Page 33: Rutgers University Chenren Xu Joint work with Bernhard Firner , Yanyong Zhang](https://reader035.fdocuments.net/reader035/viewer/2022062423/568144fc550346895db1c743/html5/thumbnails/33.jpg)
WINLAB33
Experimental Deployment
Total Size:5 × 8 m
![Page 34: Rutgers University Chenren Xu Joint work with Bernhard Firner , Yanyong Zhang](https://reader035.fdocuments.net/reader035/viewer/2022062423/568144fc550346895db1c743/html5/thumbnails/34.jpg)
WINLAB34
Experimental Deployment
![Page 35: Rutgers University Chenren Xu Joint work with Bernhard Firner , Yanyong Zhang](https://reader035.fdocuments.net/reader035/viewer/2022062423/568144fc550346895db1c743/html5/thumbnails/35.jpg)
WINLAB35
System Parameters
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
![Page 36: Rutgers University Chenren Xu Joint work with Bernhard Firner , Yanyong Zhang](https://reader035.fdocuments.net/reader035/viewer/2022062423/568144fc550346895db1c743/html5/thumbnails/36.jpg)
WINLAB36
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
![Page 37: Rutgers University Chenren Xu Joint work with Bernhard Firner , Yanyong Zhang](https://reader035.fdocuments.net/reader035/viewer/2022062423/568144fc550346895db1c743/html5/thumbnails/37.jpg)
WINLAB37
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
Training only takes 15 mins!
![Page 38: Rutgers University Chenren Xu Joint work with Bernhard Firner , Yanyong Zhang](https://reader035.fdocuments.net/reader035/viewer/2022062423/568144fc550346895db1c743/html5/thumbnails/38.jpg)
WINLAB38
Passive Localization
Motivation
Indoor challenge
Proposed solution
Experimental methodology
Performance evaluation
Conclusion and future work
![Page 39: Rutgers University Chenren Xu Joint work with Bernhard Firner , Yanyong Zhang](https://reader035.fdocuments.net/reader035/viewer/2022062423/568144fc550346895db1c743/html5/thumbnails/39.jpg)
WINLAB39
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.
![Page 40: Rutgers University Chenren Xu Joint work with Bernhard Firner , Yanyong Zhang](https://reader035.fdocuments.net/reader035/viewer/2022062423/568144fc550346895db1c743/html5/thumbnails/40.jpg)
WINLAB40
Localization Accuracy
Cell estimation accuracy:
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
![Page 41: Rutgers University Chenren Xu Joint work with Bernhard Firner , Yanyong Zhang](https://reader035.fdocuments.net/reader035/viewer/2022062423/568144fc550346895db1c743/html5/thumbnails/41.jpg)
WINLAB41
Reducing Training Dataset
1008
Only 8 samples are good enough
![Page 42: Rutgers University Chenren Xu Joint work with Bernhard Firner , Yanyong Zhang](https://reader035.fdocuments.net/reader035/viewer/2022062423/568144fc550346895db1c743/html5/thumbnails/42.jpg)
WINLAB42
Robust to Link Failure
5 transmitter + 3 receivers =
90% cell estimation accuracy
![Page 43: Rutgers University Chenren Xu Joint work with Bernhard Firner , Yanyong Zhang](https://reader035.fdocuments.net/reader035/viewer/2022062423/568144fc550346895db1c743/html5/thumbnails/43.jpg)
WINLAB43
Long-term Stability
![Page 44: Rutgers University Chenren Xu Joint work with Bernhard Firner , Yanyong Zhang](https://reader035.fdocuments.net/reader035/viewer/2022062423/568144fc550346895db1c743/html5/thumbnails/44.jpg)
WINLAB44
Multiple Subjects Localization
![Page 45: Rutgers University Chenren Xu Joint work with Bernhard Firner , Yanyong Zhang](https://reader035.fdocuments.net/reader035/viewer/2022062423/568144fc550346895db1c743/html5/thumbnails/45.jpg)
WINLAB45
Larger Deployment
Total Size: 10 × 15 m Cell Size: 2 × 2 m13 transmitters and 9 receivers
![Page 46: Rutgers University Chenren Xu Joint work with Bernhard Firner , Yanyong Zhang](https://reader035.fdocuments.net/reader035/viewer/2022062423/568144fc550346895db1c743/html5/thumbnails/46.jpg)
WINLAB46
Larger Deployment
Cell estimation accuracy: 93.8%Average error distance: 1.3 m
![Page 47: Rutgers University Chenren Xu Joint work with Bernhard Firner , Yanyong Zhang](https://reader035.fdocuments.net/reader035/viewer/2022062423/568144fc550346895db1c743/html5/thumbnails/47.jpg)
WINLAB47
Passive Localization
Motivation
Indoor challenge
Proposed solution
Experimental methodology
Performance evaluation
Conclusion and future work
![Page 48: Rutgers University Chenren Xu Joint work with Bernhard Firner , Yanyong Zhang](https://reader035.fdocuments.net/reader035/viewer/2022062423/568144fc550346895db1c743/html5/thumbnails/48.jpg)
WINLAB48
Conclusion and Future Work
Conclusion
We propose a general probabilistic classification framework
to solve the passive localization problem with:
High accuracy, low cost, robust and stable
Multiple subjects tracking generalization
Future work
Improving multiple people tracking
Passively detect the number of people
![Page 49: Rutgers University Chenren Xu Joint work with Bernhard Firner , Yanyong Zhang](https://reader035.fdocuments.net/reader035/viewer/2022062423/568144fc550346895db1c743/html5/thumbnails/49.jpg)
WINLAB49
Q & A
Thank you