Method and system for multipath fingerprinting by maximum discrimination techniques

• US 9,143,176 B2
• Filed: 04/08/2014
• Issued: 09/22/2015
• Est. Priority Date: 04/08/2013
• Status: Active Grant

• Current Assignee: Wavion limited (Alvarion Limited)
• Original Assignee: Wavion limited (Alvarion Limited)
• Inventors: YAFFE, Ariel, Wax, Mati
• Primary Examiner(s): Vo, Don N
• Application Number:
  US14/247,513
• Publication Number:
  US 20140301511A1
• Time in Patent Office: 532 Days
• Field of Search: 375/267, 375/348, 455/456.1
• US Class Current: 1/1
• CPC Class Codes: G01S 5/0218 Multipath in signal reception
• CPC Class Codes: G01S 5/02521 using a radio-map
• CPC Class Codes: G01S 5/0273 using multipath or indirect path propagation signals in position determination
• CPC Class Codes: H04B 1/1081 Reduction of multipath noise by equalising H04B7/005
• CPC Class Codes: H04W 64/00 Locating users or terminals or network equipment for network management purposes, e.g. mobility management

Abstract

Discriminative methods for multipath fingerprinting, enabling single-site localization in rich multipath environments, are provided herein. In contrast to the existing descriptive method that uses only the data of each location for its fingerprint extraction, these methods use also the data of the other locations in the database, and leverage it to make the fingerprint as different as possible from the other fingerprints in the database. The performance of these methods, validated with both simulated and real data, is superior to the existing method, demonstrating localization accuracy of about 1 m in typical indoor environments.

Claims

1. A method of determining a location of a wireless emitter comprising:

receiving by an antenna array and multi-channel receiver, a signal from the wireless emitter;

extracting a fingerprint from the signal;

matching the fingerprint with a fingerprint in a fingerprints database, to identify a matching fingerprint having a best match; and

selecting as an estimated location of the emitter, a location corresponding to the matching fingerprint,

wherein the fingerprint database stores fingerprints and corresponding locations, wherein the fingerprints are extracted from wireless signals received by said antenna array, from emitters at known K locations in a geographical coverage region,

wherein the fingerprint extracted from the received signal is a single coherent entity implicitly including directions-of-arrival and differential-delays information of multipath components of the signal, and wherein the fingerprint is extracted from a spatio-temporal covariance matrix of the received signal,

wherein a projection of the covariance matrix of an i-th location out of the K locations on the subspace is maximized, and

wherein the projection of the covariance matrix of the i-th location is calculated so as to minimize a sum of projections of the covariance matrices of part or all of other K−

1 locations in the fingerprint database.

2. The method according to claim 1, wherein the fingerprints database is generated in an off-line phase, prior to the receiving, by pre-collecting and storing the fingerprints and the corresponding K locations.

3. The method according to claim 1 wherein the projection of the covariance matrix of an i-th location is defined as P_i, wherein P_i is calculated by solving

4. The method according to claim 3, wherein solving is achieved by calculating Ŵ

_i=ortho{eigv{{circumflex over (R)}_T<sup>−

1</sup>{circumflex over (R)}_i}}.

5. A method of determining a location of a wireless emitter comprising:

receiving by an antenna array and multi-channel receiver, a signal from the wireless emitter;

extracting a fingerprint from the signal;

matching the fingerprint with a in a fingerprint database to identify a matching fingerprint having a best match; and

selecting as an estimated location of the emitter, a location corresponding to the matching fingerprint,

wherein the fingerprint database stores fingerprints and corresponding locations, wherein the fingerprints are extracted from wireless signals received by said antenna array from emitters at known K locations in a geographical coverage region,

wherein the fingerprint extracted from the received signal is a single coherent entity implicitly including directions-of-arrival and differential-delays information of multipath components of the signal and wherein the fingerprint is extracted from a spatio-temporal covariance matrix of the received signal,

wherein a transformation of the covariance matrix of an i-th location out of the K locations on the subspace is maximized, and

wherein the transformation of the covariance matrix of the i-th location is calculated so as to minimize a norm of the covariance matrices of part or all of other K−

1 locations in the fingerprint database and allows an arbitrarily structured transformation.

6. The method according to claim 5, wherein the fingerprints database is generated in an off-line phase, prior to the receiving, by pre-collecting and storing the fingerprints and the corresponding K locations.

7. The method according to claim 5, wherein the transformation of the covariance matrix of an i-th location is defined as P_i, wherein P_i is calculated by solving:

8. The method according to claim 7, wherein a solution for:

9. A system of determining a location of a wireless emitter comprising:

an antenna array and multi-channel receiver configured to receive a signal from the wireless emitter;

a fingerprint database configured to store fingerprints and corresponding locations, wherein the fingerprints are extracted from wireless signals received by said antenna array, from emitters at known K locations in a geographical coverage region; and

a processor configured to;

extract a fingerprint from the signal;

match the fingerprint with a fingerprint in a fingerprints database, to identify a matching fingerprint having a best match; and

select as an estimated location of the emitter, a location corresponding to the matching fingerprint,

wherein the fingerprint extracted from the received signal is a single coherent entity implicitly including directions-of-arrival and differential-delays information of multipath components of the signal, and wherein the fingerprint is extracted from a spatio-temporal covariance matrix of the received signal,

wherein a projection of the covariance matrix of an i-th location out of the K locations on the subspace is maximized, and

wherein the projection of the covariance matrix of the i-th location is calculated so as to minimize a sum of projections of the covariance matrices of part or all of other K−

1 locations in the fingerprint database.

10. The system according to claim 9, wherein the fingerprints database is generated in an off-line phase, prior to the receiving, by pre-collecting and storing the fingerprints and the corresponding K locations.

11. The system according to claim 9 wherein the a projection of the covariance matrix of an i-th location is defined as P_i, wherein P_i is calculated by solving

12. The system according to claim 11, wherein solving is achieved by calculating Ŵ

_i=ortho{eigv{{circumflex over (R)}_T<sup>−

1</sup>{circumflex over (R)}_i}}.

13. A system of determining a location of a wireless emitter comprising:

an antenna array and multi-channel receiver configured to receive a signal from the wireless emitter;

a fingerprint database configured to store fingerprints and corresponding locations, wherein the fingerprints are extracted from wireless signals received by said antenna array, from emitters at known K locations in a geographical coverage region; and

a processor configured to;

extract a fingerprint from the signal;

match the fingerprint with a fingerprint in a fingerprints database, to identify a matching fingerprint having a best match; and

select as an estimated location of the emitter, a location corresponding to the matching fingerprint,

wherein the fingerprint extracted from the received signal is a single coherent entity implicitly including directions-of-arrival and differential-delays information of multipath components of the signal and wherein the fingerprint is extracted from a spatio-temporal covariance matrix of the received signal,

wherein a transformation of the covariance matrix of an i-th location out of the K locations on the subspace is maximized, and

wherein the transformation of the covariance matrix of the i-th location is calculated so as to minimize a norm of the covariance matrices of part or all of other K−

1 locations in the fingerprint database and allows an arbitrarily structured transformation.

14. The system according to claim 13, wherein the fingerprints database is generated in an off-line phase, prior to the receiving, by pre-collecting and storing the fingerprints and the corresponding K locations.

15. The system according to claim 13, wherein the transformation of the covariance matrix of an i-th location is defined as P_i, wherein P_i is calculated by solving:

16. The system according to claim 15, wherein a solution for: