Detection of Time of Arrival of CDMA Signals in a Wireless Location System

• US 20090149132A1
• Filed: 12/10/2007
• Published: 06/11/2009
• Est. Priority Date: 12/10/2007
• Status: Active Grant

• Current Assignee: Skyhook Holding, Inc. (Liberty Broadband Corp.)
• Original Assignee: TruePosition Incorporated
• Inventors: Anderson, Robert J., Mia, Rashidus S., LeFever, Ronald
• Granted Patent: US 7,667,649 B2
• Time in Patent Office: Days
• US Class Current: 455/67.110
• CPC Class Codes: G01S 5/0215 Interference
• CPC Class Codes: G01S 5/0218 Multipath in signal reception
• CPC Class Codes: G01S 5/06 Position of source determined by co-ordinating a plurality of position lines defined by path-difference measurements G01S5/12 takes precedence
• CPC Class Codes: H04B 1/7075 with code phase acquisition
• CPC Class Codes: H04B 17/27 for locating or positioning the transmitter
• CPC Class Codes: H04B 17/391 Modelling the propagation channel

Abstract

In a Wireless Location System (WLS) deployed in connection with a CDMA-based wireless communications system, Location Measurement Units are used to collect multi-path corrupted radio signaling for use in time difference of arrival (TDOA) and hybrid positioning methods. Signal processing techniques are used to enhance the WLS'"'"'s ability to determine the minimally time-delayed multi-path component and thus increase the accuracy of the TDOA location in CDMA-based wireless communications systems. The signal processing includes a filtering technique for reducing the leading sidelobes of the cross-correlation function as well as a leading edge discovery procedure.

Claims

1. In a wireless location system for use in locating a mobile transmitter, a method for improving a time difference of arrival (TDOA) estimate produced by cross-correlating a local signal with a reference signal, using either a time-domain or frequency-domain cross-correlation process, wherein the local signal is a copy of a signal transmitted by the mobile transmitter as received at a first antenna and the reference signal is a copy of the signal transmitted by the mobile transmitter as received at a second antenna, the method comprising:

receiving the local signal at a first location measurement unit (LMU);

processing the local and reference signals to produce a correlation function comprising a correlation-time-frequency map;

searching the correlation function for a global maximum above a detection assurance threshold, and identifying a valid global maximum correlation peak;

reducing the correlation-time-frequency map to a two-dimensional time-correlation timeslice centered on a frequency of the valid global maximum correlation peak;

searching the two-dimensional time-correlation timeslice for a local maximum above the detection assurance threshold occurring earlier in time than the global maximum correlation peak; and

performing a leading edge discovery procedure to find an earlier leading edge concealed within a correlation envelope of the two-dimensional time-correlation timeslice.

2. A method as recited in claim 1, wherein the processing includes correlation processing and filtering of the local and reference signals to produce a correlation correlation-time-frequency function comprising leading and trailing sidelobes, wherein said leading sidelobes are reduced in amplitude.

3. A method as recited in claim 2, wherein the filtering produces a reduction of at least eight Decibels (8 dB) in the amplitude of the leading sidelobes.

4. A method as recited in claim 3, wherein the filtering process comprises the use of a Finite Impulse Response (FIR) filtering process.

5. A method as recited in claim 4, wherein the FIR filtering process comprises an all-pass, phase shifting filtering process that produces a substantially flat frequency response.

6. A method as recited in claim 4, further comprising making a cross-correlation computation using a Fast Fourier Transform (FFT), wherein the FIR filtering process is performed by multiplying the product of reference and local signal FFTs by a complex frequency response of the FIR filtering process, and then performing an Inverse Fast Fourier Transform (IFFT) to obtain the cross-correlation function.

7. A method as recited in claim 2, further comprising determining limits for a constrained time-frequency correlation search.

8. A method as recited in claim 7, further comprising reporting the TDOA value of an earliest local maximum correlation peak after the leading edge discovery procedure fails to find an earlier leading edge.

9. A method as recited in claim 7, further comprising reporting the TDOA value calculated by a predefined time-of-arrival determination procedure (Tau) performed after the leading edge discovery procedure successfully finds an earlier leading edge.

10. A method as recited in claim 2, wherein the reference signal is obtained using a second LMU comprising a wideband receiver configured to collect and digitize a code division multiple access (CDMA)-modulated signal of interest for a sample period, and digital signal processing software within the second LMU models a rake receiver and demodulates the signal of interest.

11. A method as recited in claim 10, further comprising sending the demodulated signal of interest from the second LMU to the first LMU, and re-modulating the demodulated signal of interest at the first LMU to obtain the reference signal.

12. A method as recited in claim 11, further comprising, at the first LMU, recording the received local signal;

correlating the recorded signal with the reference signal shifted over time and frequency ranges to produce a correlation function; and

applying an asymmetric, linear phase Finite Impulse Response (FIR) filter in a timewise axis to suppress leading sidelobes of the correlation function.

13. A method as recited in claim 7, further comprising discovering a global maximum correlation peak;

confining the search space to a selected frequency corresponding to the global maximum correlation peak; and

then searching a timeslice corresponding to the selected frequency to identify an earlier, local maximum correlation peak.

14. A method as recited in claim 13, further comprising determining that the earlier, local maximum correlation peak is above the detection assurance threshold.

15. A method as recited in claim 14, comprising applying the leading edge discovery procedure to identify an earlier-arriving signal hidden in a correlation envelope created by the additive effects of multi-path reflections.

16. A method as recited in claim 14, wherein, in the leading edge discovery procedure, the timeslice centered on an earliest maximum correlation peak is divided into time bins in leading and trailing directions.

17. A method as recited in claim 16, wherein the leading edge discovery procedure comprises the creation of an additional threshold based on a correlation value of the earliest maximum correlation peak.

18. A method as recited in claim 17, wherein the leading edge discovery procedure includes the detection of a second correlation peak that occurs before the global maximum correlation peak and is a local maximum below the detection assurance threshold.

19. A method as recited in claim 18, wherein the second correlation peak is hidden as a ledge or ramp within a correlation envelope of an earliest maximum correlation peak.

20. A method as recited in claim 19, wherein the leading edge discovery procedure further comprises resampling the timeslice upon a failure to find an earlier leading edge and setting a plurality of correlation thresholds.

21. A method as recited in claim 20, wherein the earliest maximum correlation peak is set to a first value (B) and the thresholds are based on the value of the earliest maximum correlation peak and a value of a sample at a later interval.

22. A method as recited in claim 21, wherein said thresholds are determined in accordance with the following time-of-arrival determination procedure (Tau):

W>

0.7*correlation value at time T=B+1,

Y>

0.2*earliest maximum correlation peak at time T=B, and

Z>

detection assurance threshold or 0.01;

wherein, starting from the earliest maximum correlation peak, the time-of-arrival determination procedure steps timewise, comparing earlier samples against the thresholds (W, Y, Z) and, once the correlation value of the sample drops below any of the thresholds, the procedure counts the samples from the earliest maximum correlation peak to the earliest sample above all thresholds; and

then the sample number (m) is used to find the earliest time-of-arrival according to the following rule set;

if m=0, then the time of a previous earliest maximum correlation peak remains the time-of-arrival for the locally received signal;

for all values of m greater than zero, the following rule is used to establish the time-of-arrival;

Tau=¼

*(chip rate)−

m/2*(chip rate)+(0.7−

c)*w,

wherein m is the number of bins between an earliest valid bin and the bin of the maximum correlation peak;

chip rate is determined based on a table defining chip rates for different spread spectrum technologies;

c is the ratio of the correlation value of the main correlation peak to the correlation value one bin later than the main correlation peak bin; and

w is an empirically-determined offset.

23. A method as recited in claim 22, wherein, once the time-of-arrival (Tau) is computed, Tau is subtracted from the time corresponding to the maximum correlation peak resulting in a new time-of-arrival that is used in a computation of the time-difference-of arrival.

24. A wireless location system (WLS) for use in locating a mobile transmitter, comprising:

a network of location measurement units (LMUs) including at least a first LMU and a second LMU;

wherein said first LMU includes a processor and a computer readable storage medium coupled to said processor, said computer readable storage medium comprising instructions for configuring said processor to compute a time difference of arrival (TDOA) estimate in accordance with a prescribed method, including by cross-correlating a local signal with a reference signal, using either a time-domain or frequency-domain cross-correlation process, wherein the local signal is a copy of a signal transmitted by the mobile transmitter as received at a first antenna associated with the first LMU and the reference signal is a copy of the signal transmitted by the mobile transmitter as received at a second antenna associated with the second LMU, said prescribed method comprising;

receiving the local signal at said first location measurement unit (LMU);

processing the local and reference signals to produce a correlation function comprising a correlation-time-frequency map;

searching the correlation function for a global maximum above a detection assurance threshold, and identifying a valid global maximum correlation peak;

reducing the correlation-time-frequency map to a two-dimensional time-correlation timeslice centered on a frequency of the valid global maximum correlation peak;

searching the two-dimensional time-correlation timeslice for a local maximum above the detection assurance threshold occurring earlier in time than the global maximum correlation peak; and

performing a leading edge discovery procedure to find an earlier leading edge concealed within a correlation envelope of the two-dimensional time-correlation timeslice.

25. A WLS as recited in claim 24, wherein the processing includes correlation processing and filtering of the local and reference signals to produce a correlation correlation-time-frequency function comprising leading and trailing sidelobes, wherein said leading sidelobes are reduced in amplitude.

26. A WLS as recited in claim 25, wherein the filtering produces a reduction of at least eight Decibels (8 dB) in the amplitude of the leading sidelobes.

27. A WLS as recited in claim 26, wherein the filtering process comprises the use of a Finite Impulse Response (FIR) filtering process.

28. A WLS as recited in claim 27, wherein the FIR filtering process comprises an all-pass, phase shifting filtering process that produces a substantially flat frequency response.

29. A WLS as recited in claim 27, said method further comprising making a cross-correlation computation using a Fast Fourier Transform (FFT), wherein the FIR filtering process is performed by multiplying the product of reference and local signal FFTs by a complex frequency response of the FIR filtering process, and then performing an Inverse Fast Fourier Transform (IFFT) to obtain the cross-correlation function.

30. A WLS as recited in claim 25, said method further comprising determining limits for a constrained time-frequency correlation search.

31. A WLS as recited in claim 30, said method further comprising reporting the TDOA value of an earliest local maximum correlation peak after the leading edge discovery procedure fails to find an earlier leading edge.

32. A WLS as recited in claim 31, said method further comprising reporting the TDOA value calculated by a predefined time-of-arrival determination procedure (Tau) performed after the leading edge discovery procedure successfully finds an earlier leading edge.

33. A WLS as recited in claim 25, wherein the reference signal is obtained using a second LMU comprising a wideband receiver configured to collect and digitize a code division multiple access (CDMA)-modulated signal of interest for a sample period, and digital signal processing software within the second LMU models a rake receiver and demodulates the signal of interest.

34. A WLS as recited in claim 33, said method further comprising sending the demodulated signal of interest from the second LMU to the first LMU, and re-modulating the demodulated signal of interest at the first LMU to obtain the reference signal.

35. A WLS as recited in claim 34, said method further comprising, at the first LMU, recording the received local signal;

correlating the recorded signal with the reference signal shifted over time and frequency ranges to produce a correlation function; and

applying an asymmetric, linear phase Finite Impulse Response (FIR) filter in a timewise axis to suppress leading sidelobes of the correlation function.

36. A WLS as recited in claim 31, said method further comprising discovering a global maximum correlation peak;

confining the search space to a selected frequency corresponding to the global maximum correlation peak; and

then searching a timeslice corresponding to the selected frequency to identify an earlier, local maximum correlation peak.

37. A WLS as recited in claim 36, said method further comprising determining that the earlier, local maximum correlation peak is above the detection assurance threshold.

38. A WLS as recited in claim 37, comprising applying the leading edge discovery procedure to identify an earlier-arriving signal hidden in a correlation envelope created by the additive effects of multi-path reflections.

39. A WLS as recited in claim 37, wherein, in the leading edge discovery procedure, the timeslice centered on an earliest maximum correlation peak is divided into time bins in leading and trailing directions.

40. A WLS as recited in claim 39, wherein the leading edge discovery procedure comprises the creation of an additional threshold based on a correlation value of the earliest maximum correlation peak.

41. A WLS as recited in claim 40, wherein the leading edge discovery procedure includes the detection of a second correlation peak that occurs before the global maximum correlation peak and is a local maximum below the detection assurance threshold.

42. A WLS as recited in claim 41, wherein the second correlation peak is hidden as a ledge or ramp within a correlation envelope of an earliest maximum correlation peak.

43. A WLS as recited in claim 42, wherein the leading edge discovery procedure further comprises resampling the timeslice upon a failure to find an earlier leading edge and setting a plurality of correlation thresholds.

44. A WLS as recited in claim 43, wherein the earliest maximum correlation peak is set to a first value (B) and the thresholds are based on the value of the earliest maximum correlation peak and a value of a sample at a later interval.

45. A WLS as recited in claim 44, wherein said thresholds are determined in accordance with the following time-of-arrival determination procedure (Tau):

W>

0.7*correlation value at time T=B+1,

Y>

0.2*earliest maximum correlation peak at time T=B, and

Z>

detection assurance threshold or 0.01;

wherein, starting from the earliest maximum correlation peak, the time-of-arrival determination procedure steps timewise, comparing earlier samples against the thresholds (W, Y, Z) and, once the correlation value of the sample drops below any of the thresholds, the procedure counts the samples from the earliest maximum correlation peak to the earliest sample above all thresholds; and

then the sample number (m) is used to find the earliest time-of-arrival according to the following rule set;

if m=0, then the time of a previous earliest maximum correlation peak remains the time-of-arrival for the locally received signal;

for all values of m greater than zero, the following rule is used to establish the time-of-arrival;

Tau=¼

*(chip rate)−

m/2*(chip rate)+(0.7−

c)*w,

wherein m is the number of bins between an earliest valid bin and the bin of the maximum correlation peak;

chip rate is determined based on a table defining chip rates for different spread spectrum technologies;

c is the ratio of the correlation value of the main correlation peak to the correlation value one bin later than the main correlation peak bin; and

w is an empirically-determined offset.

46. A WLS as recited in claim 45, wherein, once the time-of-arrival (Tau) is computed, Tau is subtracted from the time corresponding to the maximum correlation peak resulting in a new time-of-arrival that is used in a computation of the time-difference-of arrival.