SCPL: Indoor Device-Free Multi-Subject Counting and Localization Using Radio Signal Strength

Transcription

1 SCPL: Indoor Device-Free Multi-Subject Counting and Localization Using Radio Signal Strength Rutgers University Chenren Xu Joint work with Bernhard Firner, Robert S. Moore, Yanyong Zhang Wade Trappe, Richard Howard, Feixiong Zhang, Ning An

2 Device-free Localization 2

3 Device-free Localization 3

4 Why Device-free Localization? q Monitor indoor human mobility 4

5 Why Device-free Localization? q Monitor indoor human mobility Elder/health care 5

6 Why Device-free Localization? q Monitor indoor human mobility Traffic flow statistics 6

7 Why Device-free Localization? q Monitor indoor human mobility Traffic flow statistics 7

8 Why Device-free Localization? q Monitor indoor human mobility q Health/elder care, safety q Detect traffic flow q Provides privacy protection q No identification 8

9 Why Device-free Localization? q Monitor indoor human mobility q Health/elder care, safety q Detect traffic flow q Provides privacy protection q No identification q Use existing wireless infrastructure 9

10 Previous Work q Single subject localization q Fingerprinting-based approach 10

11 Fingerprinting N Subjects q Multiple subjects localization q Needs to take calibration data from N people for localizing N people 11

12 Fingerprinting N Subjects 9 trials in total for 1 person 12

13 Fingerprinting N Subjects 13

14 Fingerprinting N Subjects 14

15 Fingerprinting N Subjects 36 trials in total for 2 people! 15

16 Fingerprinting N Subjects 1 person 9 cells min = 9 min 16

17 Fingerprinting N Subjects 1 person 2 people 9 cells cells min = 10.5 hr 17

18 Fingerprinting N Subjects 1 person 2 people 3 people 9 cells cells cells min = 112 days The calibration effort is prohibitive! 18

19 SCPL q Input q Collecting calibration data only from 1 subject (D1) q Observed RSSI change caused by n subjects q Output q count and localize N subjects. q Main Insight: q If the number n is known, localizing n subjects is strightforward 19

20 No Subjects 20

21 One Subject 21

22 Two Subjects 22

23 Measurement N = 0 N = 1 N = 2 Link Link Link Total ( N) N N? N / 1 = N? 23

24 Measurement 24

25 Measurement 1.6 Nonlinear problem! N / 1 N 25

26 Measurement 4 db 5 db 26

27 Measurement 6 db 5 db 27

28 Measurement 4 db 4 db 5 db + 6 db = 7 db? 5 db 5 db 4 db + 0 db = 4 db 5 db + 6 db = 11 db 7 db X 0 db + 5 db = 5 db 28

29 Measurement 5 db + 6 db 7 db! 5 db + 6 db 7 db X Shared links observe nonlinear fading effect from multiple people 29

30 SCPL Part I Sequential Counting (SC) 30

31 Counting algorithm 31

32 Phase 1: Detection 4 db N = = 16 db 7 db 5 db Measurement in 1 st round N > 1 Subject Count ++ 32

33 Phase 2: Localization 4 db PC-DfP: 7 db 5 db Measurement in 1 st round Find this guy C. Xu, B. Firner, Y. Zhang, R. Howard, J. Li, and X. Lin. Improving rf-based device-free passive localization in cluttered indoor environments through probabilistic classification methods. In Proceedings of the 11th international conference on Information Processing in Sensor Networks, IPSN 12 33

34 Phase 3: Subtraction 6 db 5 db Calibration data 34

35 Phase 3: Subtraction 4 db 4 db 7 db - 6 db = 1 db 5 db 5 db Measurement in 1 st round Calibration data Measurement In 2 nd round Subject count ++ Go to the next iteration 35

36 Phase 3: Subtraction 4 db 4 db 7 db - 6 db = 1 db 5 db 5 db Measurement in 1 st round Calibration data Measurement In 2 nd round Subject count ++ Go to the next iteration Hold on 36

37 Phase 3: Subtraction 4 db 1 db Measurement In 2 nd round 37

38 Phase 3: Subtraction 4 db 4 db 1 db 5 db Measurement In 2 nd round Calibration data 38

39 Phase 3: Subtraction 4 db 4 db 1 db - = 5 db -4 db Measurement In 2 nd round Calibration data We over-subtracted its impact on shared link! 39

40 Measurement 40

41 Measurement 1 st round 41

42 Measurement 1 st round 42

43 Measurement 1 st round 2 st round 43

44 Phase 3: Subtraction 4 db 4 db 7 db - 6 db = 1 db 5 db 5 db Measurement in 1 st round Calibration data Measurement In 2 nd round We need to multiply a coefficient β ϵ [0, 1] when subtracting each link 44

45 Location-Link Correlation q To mitigate the error caused by this oversubtraction problem, we propose to multiply a location-link correlation coefficient before successive subtracting: 45

46 Phase 3: Subtraction 4 db 4 db 7 db db = 4.6 db 5 db Measurement in 1 st round db Calibration Data 1 db Measurement in 2 nd round Subject count ++ Go to the next iteration 46

47 Phase 3: Subtraction 4 db db 1 db 4.6 db db = 1 db 1 db 1 db Measurement in 2 nd round Calibration data Measurement in 3 rd round We are done! 47

48 SCPL Part II Parallel Localization (PL) 48

49 Localization q Cell-based localization q Allows use of context information q Reduce calibration overhead q Classification problem formulation C. Xu, B. Firner, Y. Zhang, R. Howard, J. Li, and X. Lin. Improving rf-based device-free passive localization in cluttered indoor environments through probabilistic classification methods. In Proceedings of the 11th international conference on Information Processing in Sensor Networks, IPSN 12 49

50 Linear Discriminant Analysis q RSS measurements with person s presence in each cell is treated as a class/state k q Each class k is Multivariate Gaussian with common covariance q Linear discriminant function: Link 2 RSS (dbm) k = 1 k = 2 k = 3 Link 1 RSS (dbm) 50

51 Localization q Cell-based localization q Trajectory-assisted localization q Improve accuracy by using human mobility constraints 51

52 Human Mobility Constraints You are free to go anywhere with limited step size inside a ring in free space 52

53 Human Mobility Constraints In a building, your next step is constrained by cubicles, walls, etc. 53

54 Phase 1: Data Likelihood Map 54

55 Impossible movements 55

56 Impossible movements 56

57 Phase 2: Trajectory Ring Filter 57

58 Phase 3: Refinement 58

59 Here you are! 59

60 Viterbi optimal trajectory q Single subject localization q Multiple subjects localization ViterbiScore = 60

61 System Description q Hardware: PIP tag q Microprocessor: C8051F321 q Radio chip: CC1100 q Power: Lithium coin cell battery q Protocol: Unidirectional heartbeat (Uni-HB) q Packet size: 10 bytes q Beacon interval: 100 msec 61

62 Office deployment Total Size: m 62

63 Office deployment 37 cells of cubicles, aisle segments 63

64 Office deployment 13 transmitters and 9 receivers 64

65 Office deployment Four subjects testing paths 65

66 Counting results 66

67 Counting results 67

68 Localization results 68

69 Open floor deployment Total Size: m 69

70 Open floor deployment 56 cells, 12 transmitters and 8 receivers 70

71 Open floor deployment Four subjects testing paths 71

72 Counting results 72

73 Localization results 73

74 Conclusion and Future Work q Conclusion q Calibration data collected from one subject can be used to count and localize multiple subjects. q Though indoor spaces have complex radio propagation characteristics, the increased mobility constraints can be leveraged to improve accuracy. q Future work q Count and localize more than 4 subjects 74

75 Q & A Thank you 75