System and method for estimating signal time of arrival

• US 6,493,380 B1
• Filed: 05/28/1999
• Issued: 12/10/2002
• Est. Priority Date: 05/28/1999
• Status: Expired due to Term

• Current Assignee: Microsoft Technology Licensing LLC (Microsoft Corporation)
• Original Assignee: Nortel Networks Limited (Nortel Networks Corporation)
• Inventors: Wu, Shiquan, Zhang, Song, Chen, Xixian
• Primary Examiner(s): Pham, Chi
• Assistant Examiner(s): Nguyen, Dung X.
• Application Number:
  US09/321,819
• Time in Patent Office: 1,292 Days
• Field of Search: 375/224, 375/226, 375/130, 375/141, 375/147, 375/136, 342/450, 342/457, 342/350, 342/357.06, 342/357.01, 342/386, 342/417
• US Class Current: 375/224
• CPC Class Codes: G01S 2205/06 Emergency
• CPC Class Codes: G01S 5/011 Identifying the radio environment
• 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 7/005 Control of transmission; Equalising
• CPC Class Codes: H04L 25/0234 by non-linear interpolation

Abstract

A method for estimating a time delay including the steps of first, estimating a channel spectrum from a received signal and a known reference signal, second, extracting a virtual array structure from the estimated channel spectrum, and third, determining the signal time delay using the virtual array structure. Further, a system for estimating a time delay including a channel spectrum estimator, means for extracting a virtual array structure and means for determining the time delay is provided. The channel spectrum estimator is used for estimating a channel spectrum from a received signal and a known reference signal. The means for extracting a virtual array structure uses the estimated channel spectrum. The means for determining the time delay uses the virtual array structure.

Claims

1. A method for estimating a time delay, the method comprising the steps of:

a. estimating a channel spectrum from a received signal and a known reference signal;

b. formulating a unique data matrix using the estimated channel spectrum; and

c. determining a time delay estimate using the data matrix, wherein step (a) comprises the steps of

a. estimating a baseband channel response using the received signal and the known reference signal; and

b. estimating the continuous channel spectrum using a Discrete Fourier Transform of the baseband channel response.

2. A method for estimating a time delay, the method comprising the steps of:

a. estimating a channel spectrum from a received signal and a known reference signal;

b. formulating a unique data matrix using the estimated channel spectrum; and

c. determining a time delay estimate using the data matrix, wherein step (b) comprises the step of constructing a signal space using the estimated channel spectrum and one slot so that the signal space includes delay and multipath information.

3. The method recited in claim 2 wherein step (a) comprises the step of representing the signal space as a Hermitian Toeplitz matrix.

4. A method for estimating a time delay, the method comprising the steps of:

a. estimating a channel spectrum from a received signal and a known reference signal;

b. formulating a unique data matrix using the estimated channel spectrum; and

c. determining a time delay estimate using the data matrix, wherein step (c) comprises the steps of

a. performing singular value decomposition on the data matrix;

b. detecting automatically, based on the singular value decomposition results, the number of physical paths;

c. determining a noise space using the number of physical paths;

d. determining a projection polynomial using the noise space;

e. determining at least one root pair by solving the projection polynomial;

f. mapping the roots to determine a rough time delay estimate; and

g. estimating the time delay by removing bias in the rough time delay estimate.

5. The method recited in claim 4 wherein step (e) comprises the step of solving the projection polynomial using Kuhn'"'"'s sandwich method.

6. The method recited in claim 4 wherein step (f) comprises the steps of:

a. for each root pair, selecting the constituent root closest to the unit circle when plotted on the complex plane;

b. for each constituent root selected, determining a delay according to the formula;

$\mathrm{\tau <br><br>}=\min \left[\frac{<br><br>\arctan <br><br>\frac{\mathrm{Re}\left(r_{M+1-i}\right)}{\mathrm{im}\left(r_{M+1-i}\right)}<br><br>}{2<br><br>\text{\hspace{1em}<br><br>}<br><br>\mathrm{\pi <br><br>}<br><br>\text{\hspace{1em}<br><br>}<br><br>\mathrm{\Delta <br><br>}}\right]$

for each physical path i, where τ

=TOAF, r=roots, Re=real part, im=imaginary and Δ

is predetermined;

c. determining the rough time delay estimate by selecting the shortest delay.

7. The method recited in claim 4 wherein step (g) comprises the step of determining the time delay estimate by modifying the rough time delay using a predetermined, empirical multipath compensation factor.

8. A method for estimating a time delay, the method comprising the steps of:

a. estimating a channel spectrum from a received signal and a known reference signal;

b. formulating a unique data matrix using the estimated channel spectrum; and

c. determining a time delay estimate using the data matrix, wherein the known reference signal in step (a) is a shortened burst or sync word in norm burst in an IS-136 system.

9. A method for estimating a time of arrival, the method comprising the steps of:

a. matched filtering an input signal to produce a matched filtered signal;

b. buffering the matched filtered signal and adding global positioning system time stamps to produce a buffered signal;

c. compensating for the carrier offset in the buffered signal to produce a received signal;

d. determining the start of a shortened burst signal in the received signal using a framer;

e. at the start of the shortened burst, determining the inter-symbol delay by framing the shortened burst;

f. at the start of the shortened burst, estimating a channel response using the received signal and a known reference signal;

g. formulating a data matrix using the estimated channel spectrum;

h. determining an intra-symbol delay estimate using the data matrix; and

j. determining a time of arrival by adding the inter-symbol delay and the intra-symbol delay.

10. The method recited in claim 9 wherein step (f) comprises the steps of

a. estimating a baseband channel response using the received signal and the known reference signal;

b. estimating the continuous channel spectrum using a Discrete Fourier Transform of the baseband channel response.

11. The method recited in claim 9 wherein step (g) comprises the step of constructing a signal space using the estimated channel spectrum and one slot so that the signal space includes delay and multipath information.

12. The method recited in claim 9 wherein step (a) comprises the step of representing the signal space as a Hermitian Toeplitz matrix.

13. The method recited in claim 9 wherein step (h) comprises the steps of

a. performing singular value decomposition on the data matrix;

b. detecting automatically, based on the singular value decomposition results, the number of physical paths;

c. determining a noise space using the number of physical paths;

d. determining a projection polynomial using the noise space;

e. determining a least one root pair by solving the projection polynomial;

f. mapping the roots to determine a rough intra-symbol delay estimate; and

g. estimating the intra-symbol delay by removing bias in the rough intra-symbol delay estimate.

14. The method recited in claim 13 wherein step (e) comprises the step of solving the projection polynomial using Kuhn'"'"'s sandwich method.

15. The method recited in claim 13 wherein step (f) comprises the steps of:

a. for each root pair, selecting the constituent root closest to the unit circle when plotted on the complex plane;

b. for each constituent root selected, determine a delay according to the formula;

$\mathrm{\tau <br><br>}=\min \left[\frac{<br><br>\arctan <br><br>\frac{\mathrm{Re}\left(r_{M+1-i}\right)}{\mathrm{im}\left(r_{M+1-i}\right)}<br><br>}{2<br><br>\text{\hspace{1em}<br><br>}<br><br>\mathrm{\pi <br><br>}<br><br>\text{\hspace{1em}<br><br>}<br><br>\mathrm{\Delta <br><br>}}\right]$

 

for each physical path i, where τ

=TOAf, r=roots, Re=real part, im=imaginary and Δ

is predetermined.

c. determine the rough intra-symbol delay estimate by selecting the shortest delay.

16. The method recited in claim 13 wherein step (g) comprises the step of determining the intra-symbol delay estimate by modifying the rough intra-symbol delay using a predetermined, empirical multipath compensation factor.

17. A system for estimating a time delay, the system comprising:

a. a channel spectrum estimator for estimating a channel spectrum from a received signal and a known reference signal;

b. means for extracting a virtual array structure from the estimated channel spectrum; and

c. means for determining the time delay using the virtual array

structure, wherein the extracting means comprises means for formulating a data matrix representing a signal space, the formulating means using the estimated channel spectrum and one slot so that the data matrix representing the signal space includes delay and multipath information.

18. A system for estimating a time delay, the system comprising:

a. a channel spectrum estimator for estimating a channel spectrum from a received signal and a known reference signal;

b. means for extracting a virtual array structure from the estimated channel spectrum; and

c. means for determining the time delay using the virtual array structure, wherein the determining means comprises;

a. means for performing singular value decomposition on the data matrix;

b. means for detecting automatically, based on the singular value decomposition results, the number of physical paths;

c. means for determining a noise space using the number of physical paths;

d. means for determining a projection polynomial using the noise space;

e. means for determining at least one root pair by solving the projection polynomial;

f. means for mapping the roots to determine a rough intra-symbol delay estimate; and

g. means for estimating the intra-symbol delay by removing bias in the rough time delay estimate.

19. The system recited in claim 18,

a. wherein the mapping means comprises:

i. means for selecting, for each root pair, the constituent root closest to the unit circle when plotted on the complex plane;

ii. means for determining, for each constituent root selected, a delay according to the formula;

$\mathrm{\tau <br><br>}=\min \left[\frac{<br><br>\arctan <br><br>\frac{\mathrm{Re}\left(r_{M+1-i}\right)}{\mathrm{im}\left(r_{M+1-i}\right)}<br><br>}{2<br><br>\text{\hspace{1em}<br><br>}<br><br>\mathrm{\pi <br><br>}<br><br>\text{\hspace{1em}<br><br>}<br><br>\mathrm{\Delta <br><br>}}\right]$

 

for each physical path i, where τ

=TOAF, r=roots, Re=real part, im=imaginary and Δ

is predetermined.

iii. means for determining the rough time delay estimate by selecting the shortest delay; and

b. wherein the estimating means comprises;

i. means for determining the time delay estimate by modifying the rough time delay using a predetermined, empirical multipath compensation factor.

20. The system recited in claim 19, wherein the time delay is an intra-symbol delay.

21. The system recited in claim 20 further comprising:

a. a framer for estimating a inter-symbol delay using the received signal; and

b. a summer for combining the inter-symbol-delay and the intrasymbol delay to produce a time of arrival estimate.

22. The system recited in claim 21, the system further comprising:

a. a matched filter for matched filtering an input signal to produce a filtered signal;

b. a data buffer for buffering the filtered signal and for adding a global positioning system time stamp to the filtered signal to produce a buffered signal; and

c. a carrier frequency offset compensator for compensating for carrier offset in the buffered signal to produce the received signal.