Real-time RFID localization using uniform, high-performance tags and related method thereof

• US 10,401,488 B2
• Filed: 06/24/2014
• Issued: 09/03/2019
• Est. Priority Date: 06/26/2013
• Status: Active Grant

• Current Assignee: University of Virginia Patent Foundation (University of Virginia)
• Original Assignee: University of Virginia Patent Foundation (University of Virginia)
• Inventors: Chawla, Kirti, Robins, Gabriel
• Primary Examiner(s): Liu, Harry K
• Application Number:
  US14/899,399
• Publication Number:
  US 20160146934A1
• Time in Patent Office: 1,897 Days
• Field of Search: 342458
• CPC Class Codes: G01S 11/06 using intensity measurements
• CPC Class Codes: G01S 13/74 Systems using reradiation of radio waves, e.g. secondary radar systems; Analogous systems
• CPC Class Codes: G01S 5/021 Calibration, monitoring or correction G01S5/0252 takes precedence
• CPC Class Codes: G01S 5/14 Determining absolute distances from a plurality of spaced points of known location

Abstract

Disclosed herein are systems, methods, and machine readable media that enable object localization with an RFID infrastructure. Localization is performed by modeling the distance-decaying behavior of received signal strength. Selection of optimally performing tag types and the use of uniformly sensitive tags of those optimal tag types further enhances localization accuracy. When these components are combined, RFID becomes a feasible way to locate moving and stationary objects in a 3D environment that includes other objects, as well as noise and signal pollution.

Claims

1. A method of selecting a type of radio-frequency identification (RFID) tag, the method comprising:

providing a plurality of RFID tags, wherein the plurality of RFID tags is comprised of at least two different RFID tag types;

measuring the received signal strength (RSS) of respective ones of the plurality of RFID tags for at least one RFID reader transmit power level of an RFID reader, and at least one reader-tag distance, wherein the reader-tag distance is the distance between the RFID reader and a respective RFID tag amongst the plurality of RFID tags; and

analyzing the measured RSS of the respective ones of the plurality of RFID tags by evaluating the measured RSS for each of the respective ones of the plurality of RFID tags with respect to at least one metric, and in response, selecting an RFID tag type from amongst the at least two different RFID tag types.

2. The method of claim 1, wherein said analyzing includes determining which of said different RFID tag types has the least variation between measured RSS over the different RFID reader output power levels.

3. The method of claim 1, wherein said analyzing includes determining which of the at least two different RFID tag types has the highest correlation between increases in reader-tag distance and decreases in measured RSS.

4. The method of claim 1, wherein said analyzing includes determining which of the at least two different RFID tag types has the longest read range.

5. The method of claim 1, wherein said analyzing includes determining which of the at least two different RFID tag types has the highest read count.

6. The method of claim 1, wherein the plurality of RFID tags are passive.

7. The method of claim 1, wherein the plurality of RFID tags are active.

8. The method of claims 1, wherein the plurality of RFID tags are semi-passive.

9. The method of claim 1, wherein the plurality of RFID tags are at least one of passive, active, semi-passive, or any combination thereof.

10. The method of claim 1, wherein said analyzing comprises selecting two or more optimal RFID tag types.

11. The method of claim 1, further comprising:

identifying a subset of RFID tags of the selected RFID tag type that perform closest to the mean on a particular one of the at least one metric.

12. The method of claim 11, wherein said particular one of the at least one metric is measured RSS.

13. The method of claim 11, wherein said particular one of the at least one metric is measured read-count.

14. The method of claim 12, wherein the subset of RFID tags are the RFID tags that have measured RSS within a predetermined number of standard deviations from the mean.

15. The method of claim 12, wherein the subset of RFID tags are the RFID tags that have measured RSS within a user-defined number of standard deviations from the mean.

16. The method of claim 15, wherein said user-defined number of standard deviations comprises two standard deviations from the mean.

17. The method of claim 11, further comprising:

developing an RSS decay model that relates increasing tag-reader distance to decreasing RSS, using the measured RSS of said subset of RFID tags at a plurality of tag-reader distances.

18. The method of claim 17, wherein said RSS decay model is an exponential decay formula created with a least-squares fitting.

19. The method of claim 18, wherein goodness of the fit of said least-squares fitting is determined using an R² metric.

20. The method of claim 18, wherein goodness of the fit of said least-squares fitting is determined using a normalized root-mean-square error (NRMSE) metric.

21. The method of claim 17, wherein said exponential decay formula is:

RSS=C*D{circumflex over ( 

)}E

wherein;

C is a coefficient,

D is a tag-reader distance, and

E is an exponent.

22. The method of claim 17, wherein the RSS of said subset of RFID tags is also measured at a plurality of tag orientations with respect to the RFID reader.

23. The method of claim 22, wherein said plurality of tag orientations are axial, radial, or both.

24. The method of claim 1, further comprising:

developing an RSS decay model that relates increasing tag-reader distance to decreasing RSS, using the measured RSS of one or more RFID tags of said optimal RFID tag type at a plurality of tag-reader distances.

25. The method of claim 24, wherein said RSS decay model is an exponential decay formula created with a least-squares fitting.

26. The method of claim 25, wherein goodness of the fit of said least-squares fitting is determined using an R² metric.

27. The method of claim 25, wherein goodness of the fit of said least-squares fitting is determined using a normalized root-mean-square error (NRMSE) metric.

28. The method of claim 24, wherein said exponential decay formula is:

RSS=C*D{circumflex over ( 

)}E

wherein;

C is a coefficient,

D is a tag-reader distance, and

E is an exponent.

29. The method of claim 24, wherein the RSS is also measured at a plurality of tag orientations with respect to the RFID reader.

30. The method of claim 29, wherein said plurality of tag orientations are axial, radial, or both.

31. A method of developing a received signal strength (RSS) decay model for a radio-frequency identification (RFID) system, the method comprising:

measuring RSS of one or more RFID tags at a plurality of tag-reader distances and a plurality of orientations with respect to an RFID reader, wherein the tag-reader distance is the distance between the RFID reader and a respective RFID tag of the one or more RFID tags; and

fitting said RSS measurements to an exponential decay formula to provide said RSS decay model, wherein said exponential decay formula relates increasing tag-reader distance to decreasing RSS.

32. The method of claim 31, wherein said one or more RFID tags measured are passive.

33. The method of claim 31, wherein said one or more RFID tags measured are active.

34. The method of claim 31, wherein said one or more RFID tags measured are semi-passive.

35. The method of claim 31, wherein said fitting is a least-squares fitting.

36. The method of claim 35, wherein goodness of the fit of said least-squares fitting is determined using an R² metric.

37. The method of claim 35, wherein goodness of the fit of said least-squares fitting is determined using a normalized root-mean-square error (NRMSE) metric.

38. The method of claim 31, wherein said exponential decay formula is:

RSS=C*D{circumflex over ( 

)}E

wherein;

C is a coefficient,

D is a tag-reader distance, and

E is an exponent.

39. The method of claim 31, wherein the orientation of said one or more RFID tags is axial.

40. The method of claim 31, wherein the orientation of said one or more RFID tags is radial.

41. The method of claim 31, wherein the orientation of said one or more RFID tags is axial and radial.

42. The method of claim 31, wherein said fitting separates RSS measurements at different orientations, creates a preliminary decay model for each orientation, and averages the preliminary decay models to create said RSS decay model.

43. The method of claim 31, wherein said fitting of said RSS measurements comprises averaging all RSS measurements at each tag-reader distance and creating said RSS decay model using the average RSS measurements.

44. A system comprising:

a tag testing circuit configured to measure the received signal strength (RSS) of respective ones of a plurality of RFID tags for at least one RFID reader transmit power level of an RFID reader and at least one reader-tag distance, wherein the plurality of RFID tags is comprised of at least two different RFID tag types; and

a tag selection circuit configured to analyze the measured RSS of the respective ones of the plurality of RFID tags by evaluating the measured RSS for each of the plurality of RFID tags with respect to at least one metric, and in response, selecting an RFID tag type amongst the at least two different RFID tag types.

45. The system of claim 44, wherein said tag selection circuit is configured to determine which of at least two different RFID tag types has the least variation between measured RSS over the different RFID reader output power levels.

46. The system of claim 44, wherein said tag selection circuit is configured to determine which of at least two different RFID tag types has the highest correlation between increases in reader-tag distance and decreases in measured RSS.

47. The system of claim 44, wherein said tag selection circuit is configured to determine which of at least two different RFID tag types has the longest read range.

48. The system of claim 44, wherein said tag selection circuit is configured to determine which of at least two different RFID tag types has the highest read count.

49. The system of claim 44, wherein the plurality of RFID tags are passive.

50. The system of claim 44, wherein the plurality of RFID tags are active.

51. The system of claims 44, wherein the plurality of RFID tags are semi-passive.

52. The system of claim 44, wherein the plurality of RFID tags are at least one of passive, active, semi-passive, or any combination thereof.

53. The system of claim 44, wherein said tag selection circuit is configured to select two or more optimal RFID tag types.

54. The system of claim 44, further comprising:

a binning circuit configured to identify a subset of RFID tags of said optimal RFID tag type that perform closest to the mean on a particular one of the at least one metric.

55. The system of claim 54, wherein said particular one of the at least one metric is measured RSS.

56. The system of claim 54, wherein said particular one of the at least one metric is measured read-count.

57. The system of claim 55, wherein the subset of RFID tags are the RFID tags that have measured RSS within a predetermined number of standard deviations from the mean.

58. The system of claim 55, wherein the subset of RFID tags are the RFID tags that have measured RSS within a user-defined number of standard deviations from the mean.

59. The system of claim 55, wherein said user-defined number of standard deviations comprises two standard deviations from the mean.

60. The system of claim 54, further comprising:

an RSS measurement circuit configured to measure the RSS of said subset of RFID tags at a plurality of tag-reader distances; and

a calibration circuit configured to develop an RSS decay model that relates increasing tag-reader distance to decreasing RSS, using the measurements from said RSS measurement circuit.

61. The system of claim 60, wherein said RSS decay model is an exponential decay formula developed through a least-squares fitting.

62. The system of claim 61, wherein goodness of the fit of said least-squares fitting is determined using an R² metric.

63. The system of claim 61, wherein goodness of the fit of said least-squares fitting is determined using a normalized root-mean-square error (NRMSE) metric.

64. The system of claim 61, wherein said exponential decay formula is:

RSS=C*D{circumflex over ( 

)}E

wherein;

C is a coefficient,

D is a tag-reader distance, and

E is an exponent.

65. The system of claim 60, wherein said RSS measurement circuit is configured to measure the RSS of said subset of RFID tags at a plurality of tag orientations with respect to the RFID reader.

66. The system of claim 65, wherein said plurality of tag orientations are axial, radial, or both.

67. The system of claim 44, further comprising:

an RSS measurement circuit configured to measure the RSS of said subset of RFID tags at a plurality of tag-reader distances; and

a calibration circuit configured to develop an RSS decay model that relates increasing tag-reader distance to decreasing RSS, using the measurements from said RSS measurement circuit.

68. The system of claim 67, wherein said RSS decay model is an exponential decay formula created with a least-squares fitting.

69. The system of claim 68, wherein goodness of the fit of said least-squares fitting is determined using an R² metric.

70. The system of claim 68, wherein goodness of the fit of said least-squares fitting is determined using a normalized root-mean-square error (NRMSE) metric.

71. The system of claim 67, wherein said exponential decay formula is:

RSS=C*D{circumflex over ( 

)}E

wherein;

C is a coefficient,

D is a tag-reader distance, and

E is an exponent.

72. The system of claim 67, wherein the RSS is also measured at a plurality of tag orientations with respect to the RFID reader.

73. The system of claim 72, wherein said plurality of tag orientations are axial, radial, or both.

74. A system comprising:

a received signal strength (RSS) measurement circuit configured to measure the RSS of one or more RFID tags at a plurality of tag-reader distances and a plurality of orientations with respect to an RFID reader, wherein the tag-reader distance is the distance between the RFID reader and a respective RFID tag of the one or more RFID tags; and

a calibration circuit configured to fit said RSS measurements to an exponential decay formula to provide an RSS decay model, wherein said exponential decay formula relates increasing tag-reader distance to decreasing RSS.

75. The system of claim 74, wherein said one or more RFID tags are passive.

76. The system of claim 74, wherein said one or more RFID tags are active.

77. The system of claim 74, wherein said one or more RFID tags are semi-passive.

78. The system of claim 74, wherein said fitting is a least-squares fitting.

79. The system of claim 78, wherein goodness of the fit of said least-squares fitting is determined using an R² metric.

80. The system of claim 78, wherein goodness of the fit of said least-squares fitting is determined using a normalized root-mean-square error (NRMSE) metric.

81. The system of claim 74, wherein said exponential decay formula is:

RSS=C*D{circumflex over ( 

)}E

wherein;

C is a coefficient,

D is a tag-reader distance, and

E is an exponent.

82. The system of claim 74, wherein the orientation of said one or more RFID tags is axial.

83. The system of claim 74, wherein the orientation of said one or more RFID tags is radial.

84. The system of claim 74, wherein the orientation of said one or more RFID tags is axial and radial.

85. The system of claim 74, wherein said fitting separates RSS measurements at different orientations, creates a preliminary decay model for each orientation, and averages the preliminary decay models to create said RSS decay model.

86. The system of claim 74, wherein said fitting comprises averaging all RSS measurements at each tag-reader distance and creating said RSS decay model using the average RSS measurements.