Variable suppression of multipath signal effects

0Associated
Cases 
0Associated
Defendants 
0Accused
Products 
110Forward
Citations 
0
Petitions 
2
Assignments
First Claim
1. A method of formation of an autocorrelation difference function of an incoming signal that reduces effects of presence of a multipath signal in the incoming signal, the method comprising the steps of:
 (1) receiving an incoming digital signal s(t) that can vary with the time t and that has a digital signal bit period with a selected length Δ
τ
_{chip} ;
(2) forming a first signal product difference
space="preserve" listingtype="equation">Δ
s.sub.1 (t;
τ
;
t.sub.E ;
t.sub.L ;
qE;
qL)=s(t)S.sub.d (t+τ
t.sub.E)w1(t+τ
t.sub.E ;
qE)s(t)S.sub.d (t+τ
t.sub.L)w1(t+τ
t.sub.L ;
qL),where S_{d} (t) is a selected reference signal, w1(t;
q) is a first selected, nonconstant weighting signal that may depend upon one or more parameters q, τ
is a selected time shift, t_{E} and t_{L} are first and second selected time values satisfying 0<
t_{L} t_{E} <
2Δ
τ
_{chip} ;
(3) forming a second signal product difference
space="preserve" listingtype="equation">Δ
s.sub.2 (t;
τ
;
t.sub.E ;
t.sub.L ;
qE;
qL)=s(t)S.sub.d (t+τ
t.sub.E)w1 (t+τ
t.sub.E ;
qE)s(t)S.sub.d (t+τ
t.sub.L)w1 (t+τ
t.sub.L ;
qL),where w1 (t;
q) is a second selected weighting signal that may depend upon one or more parameters q;
(4) selecting an integer n≧
2 and setting an accumulation A_{n1} =0;
(5) examining the incoming digital signal bit value b_{n}, for the time interval defined by nΔ
τ
_{chip} ≦
t<
(n+1)Δ
τ
_{chip}, and the immediately preceding digital bit value b_{n1} of the reference signal S_{d} (t);
(6) when b_{n1} ≠
b_{n}, computing the contribution of the first signal product difference over a time interval I_{n} ={t'"'"'t_{n1} +Δ
<
t'"'"'≦
t_{n} +Δ
}, where Δ
is a selected time value satisfying 0≦
Δ
<
Δ
τ
_{chip}, to an autocorrelation difference function Δ
AC#(τ
;
q) for the incoming digital signal s(t), and adding this contribution to A_{n1} to form A_{n} ;
(7) when b_{n1} =b_{n}, computing the contribution of the second signal product difference over the time interval I_{n} to the autocorrelation difference function Δ
AC#(τ
;
q) for the incoming digital signal s(t), and adding this integral or sum to A_{n1} to form A_{n} ;
(8) replacing the integer n by the integer n+1 and repeating steps (5), (6) and (7) at least once;
(9) interpreting the accumulation A_{N} for a selected positive integer N as the autocorrelation difference function Δ
AC#(τ
;
q) for the incoming digital signal s(t);
(10) determining at least one value t_{0} of the time shift variable τ
for which the autocorretation difference function Δ
AC#(τ
;
q) changes sign; and
(11) interpreting the time value t=t_{0} as an estimate of the time at which a signal, which is substantially free of the presence of a multipath signal, was received.
2 Assignments
0 Petitions
Accused Products
Abstract
Method and apparatus for formation of an autocorrelation difference function of an incoming digital signal that reduces the effects of presence of a multipath signal or of noise in an incoming digital composite signal. An incoming digital composite signal, including direct and multipath signals, is received that has a bit value transition interval of length Δτ_{chip}. Two or three consecutive bit values b_{n2}, b_{n1} and b_{n} of the direct (ideal) signal are examined. If a test condition for these bit values is satisfied, a first nonuniform weighting function w1(t) is used to compute the contribution of a time interval I_{n} ={t'"'"'t_{n1} +Δ<t'"'"'≦t_{n} +Δ}, where Δ is a selected time value satisfying 0≦Δ<Δτ_{chip}, to first and second autocorrelation functions AC#(τ;E) and AC#(τ;L) with respective selected first and second time shifts τ=t_{E} and τ=t_{L} (>t_{E}) If the test condition is not satisfied, a second weighting function w1 (t) is used to compute the contribution of the time interval I_{n} to AC#(τ;E) and AC#(τ;L). An autocorrelation difference function ΔAC#(τ)=AC#(τ;E) and AC#(τ;L) is formed in which the effects of noise or of multipath signals is suppressed, relative these effects in a conventionally computed autocorrelation difference function ΔAC(τ)=AC(τ;E)AC(τ;L).
128 Citations
Method for synchronizing a radio network using end user radio terminals  
Patent #
US 7,877,104 B2
Filed 05/21/2002

Current Assignee
CSR Technology Incorporated

Sponsoring Entity
CSR Technology Incorporated

Cancellation system and method for a wireless positioning system  
Patent #
US 7,885,314 B1
Filed 05/02/2000

Current Assignee
CSR Technology Incorporated

Sponsoring Entity
CSR Technology Incorporated

Frequency phase correction system  
Patent #
US 7,929,928 B2
Filed 11/24/2004

Current Assignee
CSR Technology Incorporated

Sponsoring Entity
CSR Technology Incorporated

Mode determination for mobile GPS terminals  
Patent #
US 7,969,351 B2
Filed 12/04/2007

Current Assignee
CSR Technology Incorporated

Sponsoring Entity
CSR Technology Incorporated

GPS RF front end IC with frequency plan for improved integrability  
Patent #
US 7,933,627 B2
Filed 03/06/2006

Current Assignee
CSR Technology Incorporated

Sponsoring Entity
CSR Technology Incorporated

Synchronizing a radio network with end user radio terminals  
Patent #
US 7,925,210 B2
Filed 08/16/2005

Current Assignee
CSR Technology Incorporated

Sponsoring Entity
CSR Technology Incorporated

Satellite positioning aided communication system selection  
Patent #
US 7,949,362 B2
Filed 12/22/2004

Current Assignee
CSR Technology Incorporated

Sponsoring Entity
CSR Technology Incorporated

Aided location communication system  
Patent #
US 7,970,412 B2
Filed 07/20/2005

Current Assignee
Samsung SDI Company Limited

Sponsoring Entity
CSR Technology Incorporated

Signal Processing System for Satellite Positioning Signals  
Patent #
US 20110102258A1
Filed 09/02/2004

Current Assignee
CSR Technology Incorporated

Sponsoring Entity
CSR Technology Incorporated

Aided location communication system  
Patent #
US 7,970,411 B2
Filed 07/03/2004

Current Assignee
Samsung SDI Company Limited

Sponsoring Entity
CSR Technology Incorporated

Control and features for satellite positioning system receivers  
Patent #
US 8,013,787 B2
Filed 09/02/2004

Current Assignee
CSR Technology Incorporated

Sponsoring Entity
CSR Technology Incorporated

Method and apparatus for transmitting position data using control channels in wireless networks  
Patent #
US 7,986,952 B2
Filed 05/22/2006

Current Assignee
CSR Technology Incorporated

Sponsoring Entity
CSR Technology Incorporated

Fast reacquisition of a GPS signal  
Patent #
US 7,999,733 B2
Filed 02/19/2007

Current Assignee
CSR Technology Incorporated

Sponsoring Entity
CSR Technology Incorporated

Communications systems that reduces autocorrelation or crosscorrelation in weak signals  
Patent #
US 7,719,466 B2
Filed 02/19/2007

Current Assignee
CSR Technology Incorporated

Sponsoring Entity
SiRF Technology Holdings Incorporated

System and method for despreading in a spread spectrum matched filter  
Patent #
US 7,852,905 B2
Filed 06/16/2004

Current Assignee
CSR Technology Incorporated

Sponsoring Entity
CSR Technology Incorporated

Method and apparatus for estimating local oscillator frequency for GPS receivers  
Patent #
US 7,747,236 B1
Filed 12/11/2001

Current Assignee
CSR Technology Incorporated

Sponsoring Entity
CSR Technology Incorporated

Method and apparatus for selectively maintaining circuit power when higher voltages are present  
Patent #
US 7,671,489 B1
Filed 01/17/2002

Current Assignee
CSR Technology Incorporated

Sponsoring Entity
SiRF Technology

Method for reducing autocorrelation or crosscorrelation in weak signals  
Patent #
US 7,724,807 B2
Filed 08/02/2006

Current Assignee
CSR Technology Incorporated

Sponsoring Entity
CSR Technology Incorporated

WAVEFORM CORRELATION RESULT PROCESSING METHODS AND APPARATUSES  
Patent #
US 20100153040A1
Filed 12/12/2008

Current Assignee
Qualcomm Inc.

Sponsoring Entity
Qualcomm Inc.

Crosscorrelation detection and elimination in a receiver  
Patent #
US 7,680,178 B2
Filed 12/07/2005

Current Assignee
CSR Technology Incorporated

Sponsoring Entity
SiRF Technology

Network system for aided GPS broadcast positioning  
Patent #
US 7,668,554 B2
Filed 12/16/2005

Current Assignee
CSR Technology Incorporated

Sponsoring Entity
SiRF Technology

System and method for despreading in a spread spectrum matched filter  
Patent #
US 7,711,038 B1
Filed 06/27/2000

Current Assignee
CSR Technology Incorporated

Sponsoring Entity
SiRF Technology

GPSBASED POSITIONING SYSTEM FOR MOBILE GPS TERMINALS  
Patent #
US 20090033553A1
Filed 02/18/2008

Current Assignee
Panasonic Corporation

Sponsoring Entity
Panasonic Corporation

GPS RF front end IC with programmable frequency synthesizer for use in wireless phones  
Patent #
US 7,512,385 B2
Filed 06/06/2005

Current Assignee
CSR Technology Incorporated

Sponsoring Entity
SiRF Technology

Doppler corrected spread spectrum matched filter  
Patent #
US 7,545,854 B1
Filed 02/07/2000

Current Assignee
CSR Technology Incorporated

Sponsoring Entity
SiRF Technology

Multimode global positioning system for use with wireless networks  
Patent #
US 7,577,448 B2
Filed 07/12/2002

Current Assignee
Samsung SDI Company Limited

Sponsoring Entity
SiRF Technology

Monolithic GPS RF front end integrated circuit  
Patent #
US 7,616,705 B1
Filed 07/26/2001

Current Assignee
CSR Technology Incorporated

Sponsoring Entity
SiRF Technology

Communications Systems That Reduces AutoCorrelation or CrossCorrelation In Weak Signals  
Patent #
US 20080001817A1
Filed 02/19/2007

Current Assignee
CSR Technology Incorporated

Sponsoring Entity
CSR Technology Incorporated

Location services system that reduces autocorrelation or crosscorrelation in weak signals  
Patent #
US 7,365,680 B2
Filed 01/31/2006

Current Assignee
CSR Technology Incorporated

Sponsoring Entity
SiRF Technology

Monolithic GPS RF front end integrated circuit  
Patent #
US 7,369,830 B2
Filed 09/13/2004

Current Assignee
CSR Technology Incorporated

Sponsoring Entity
SiRF Technology

Mode Determination for Mobile GPS Terminals  
Patent #
US 20080180316A1
Filed 12/04/2007

Current Assignee
CSR Technology Incorporated

Sponsoring Entity
CSR Technology Incorporated

Phase multipath mitigation  
Patent #
US 7,453,925 B2
Filed 10/18/2004

Current Assignee
Deere Company

Sponsoring Entity
Navcom Technology Inc.

Tracker architecture for GPS systems  
Patent #
US 20070008218A1
Filed 06/26/2006

Current Assignee
CSR Technology Incorporated

Sponsoring Entity
CSR Technology Incorporated

Communications system that reduces autocorrelation or crosscorrelation in weak signals  
Patent #
US 7,183,972 B2
Filed 11/12/2003

Current Assignee
CSR Technology Incorporated

Sponsoring Entity
SiRF Technology

Method for reducing autocorrelation or crosscorrelation in weak signals  
Patent #
US 20070030888A1
Filed 08/02/2006

Current Assignee
CSR Technology Incorporated

Sponsoring Entity
CSR Technology Incorporated

Apparatus for reducing autocorrelation or crosscorrelation in weak CDMA signals  
Patent #
US 7,197,305 B2
Filed 07/20/2001

Current Assignee
CSR Technology Incorporated

Sponsoring Entity
SiRF Technology

Aiding in a satellite positioning system  
Patent #
US 7,236,883 B2
Filed 12/04/2003

Current Assignee
CSR plc

Sponsoring Entity
SiRF Technology

Method and system for data detection in a global positioning system satellite receiver  
Patent #
US 7,269,511 B2
Filed 07/06/2005

Current Assignee
CSR Technology Incorporated

Sponsoring Entity
SiRF Technology

Fast acquisition of GPS signal  
Patent #
US 6,778,136 B2
Filed 12/13/2001

Current Assignee
CSR Technology Incorporated

Sponsoring Entity
SiRF Technology

Adaptive wireless communication receiver  
Patent #
US 6,693,953 B2
Filed 09/30/1998

Current Assignee
Intel Corporation

Sponsoring Entity
Skyworks Solutions Incorporated

Personal communications device with ratio counter  
Patent #
US 6,788,655 B1
Filed 04/18/2000

Current Assignee
CSR Technology Incorporated

Sponsoring Entity
SiRF Technology

Signal detector employing correlation analysis of nonuniform and disjoint sample segments  
Patent #
US 7,002,516 B2
Filed 08/19/2003

Current Assignee
CSR Technology Incorporated

Sponsoring Entity
SiRF Technology

Signal detector employing coherent integration  
Patent #
US 6,496,145 B2
Filed 10/04/2001

Current Assignee
CSR Technology Holdings Inc.

Sponsoring Entity
SiRF Technology Holdings Incorporated

Signal detector employing correlation analysis of nonuniform and disjoint sample segments  
Patent #
US 6,636,178 B2
Filed 10/04/2001

Current Assignee
Samsung SDI Company Limited

Sponsoring Entity
SiRF Technology Holdings Incorporated

Jamming detection and blanking for GPS receivers  
Patent #
US 6,448,925 B1
Filed 02/04/1999

Current Assignee
CSR Technology Incorporated

Sponsoring Entity
SiRF Technology

Method and an apparatus for positioning system assisted cellular radiotelephone handoff and dropoff  
Patent #
US 6,327,471 B1
Filed 02/19/1998

Current Assignee
Skyworks Solutions Incorporated

Sponsoring Entity
Conexant Systems Incorporated

Spread spectrum receiver performance improvement  
Patent #
US 6,606,349 B1
Filed 02/04/1999

Current Assignee
CSR Technology Incorporated

Sponsoring Entity
SiRF Technology

Method and system for data detection in a global positioning system satellite receiver  
Patent #
US 6,714,158 B1
Filed 04/18/2000

Current Assignee
CSR Technology Incorporated

Sponsoring Entity
SiRF Technology

Voice activated circuit and radio using same  
Patent #
US 6,633,847 B1
Filed 01/05/2000

Current Assignee
Motorola Inc.

Sponsoring Entity
Motorola Inc.

Communications network with wireless gateways for mobile terminal access  
Patent #
US 7,092,369 B2
Filed 07/19/2001

Current Assignee
Symbol Technologies LLC

Sponsoring Entity
Symbol Technologies Inc.

Signal detector employing correlation analysis of nonuniform and disjoint sample segments  
Patent #
US 6,304,216 B1
Filed 03/30/1999

Current Assignee
CSR Technology Incorporated

Sponsoring Entity
SiRF Technology Holdings Incorporated

GPS RF front end IC with frequency plan for improved integrability  
Patent #
US 7,047,023 B1
Filed 12/03/2001

Current Assignee
CSR Technology Incorporated

Sponsoring Entity
SiRF Technology

Integrated power management module  
Patent #
US 6,348,744 B1
Filed 04/14/1998

Current Assignee
Alpha Industries Incorporated

Sponsoring Entity
Skyworks Solutions Incorporated

Method for reducing autocorrelation or crosscorrelation in weak signals  
Patent #
US 7,106,786 B2
Filed 07/20/2001

Current Assignee
CSR Technology Incorporated

Sponsoring Entity
SiRF Technology

Signal detector employing a Doppler phase correction system  
Patent #
US 6,952,440 B1
Filed 04/18/2000

Current Assignee
CSR Technology Incorporated

Sponsoring Entity
SiRF Technology

Accelerated selection of a base station in a wireless communication system  
Patent #
US 6,519,277 B2
Filed 10/16/2001

Current Assignee
CSR Technology Incorporated

Sponsoring Entity
SiRF Technology

Signal detector employing a doppler phase correction system  
Patent #
US 6,931,055 B1
Filed 04/18/2000

Current Assignee
CSR Technology Incorporated

Sponsoring Entity
SiRF Technology

Field unit for use in a GPS system  
Patent #
US 6,531,982 B1
Filed 09/30/1997

Current Assignee
CSR Technology Incorporated

Sponsoring Entity
SiRF Technology

Signal detector employing coherent integration  
Patent #
US 6,577,271 B1
Filed 03/30/1999

Current Assignee
CSR Technology Incorporated

Sponsoring Entity
SiRF Technology

Method and system for data detection in a global positioning system satellite receiver  
Patent #
US 6,961,660 B2
Filed 03/03/2004

Current Assignee
Samsung SDI Company Limited

Sponsoring Entity
SiRF Technology

GPS RF front end IC with programmable frequency synthesizer for use in wireless phones  
Patent #
US 20060040631A1
Filed 06/06/2005

Current Assignee
CSR Technology Incorporated

Sponsoring Entity
CSR Technology Incorporated

GPSbased positioning system for mobile GPS terminals  
Patent #
US 7,009,555 B2
Filed 08/07/2002

Current Assignee
CSR Technology Incorporated

Sponsoring Entity
Matsushita Electric Industrial Company Limited

Phase multipath mitigation  
Patent #
US 20060083293A1
Filed 10/18/2004

Current Assignee
Deere Company

Sponsoring Entity
Deere Company

Aiding in a satellite positioning system  
Patent #
US 20060095206A1
Filed 05/22/2003

Current Assignee
CSR plc

Sponsoring Entity
CSR plc

Method and apparatus for transmitting position data using control channels in wireless networks  
Patent #
US 7,076,256 B1
Filed 04/15/2002

Current Assignee
CSR Technology Incorporated

Sponsoring Entity
SiRF Technology

Location services system that reduces autocorrelation or crosscorrelation in weak signals  
Patent #
US 20060214846A1
Filed 01/31/2006

Current Assignee
CSR Technology Incorporated

Sponsoring Entity
CSR Technology Incorporated

Phase sampling techniques using amplitude bits for digital receivers  
Patent #
US 7,113,552 B1
Filed 12/21/2001

Current Assignee
CSR Technology Incorporated

Sponsoring Entity
SiRF Technology

Network system for aided GPS broadcast positioning  
Patent #
US 20060223549A1
Filed 12/16/2005

Current Assignee
CSR Technology Incorporated

Sponsoring Entity
CSR Technology Incorporated

Signal detector employing correlation analysis of nonuniform and disjoint sample segments  
Patent #
US 20050035905A1
Filed 08/19/2003

Current Assignee
CSR Technology Incorporated

Sponsoring Entity
CSR Technology Incorporated

Aided location communication system  
Patent #
US 20050062643A1
Filed 07/03/2004

Current Assignee
Samsung SDI Company Limited

Sponsoring Entity
Samsung SDI Company Limited

Monolithic GPS RF front end integrated circuit  
Patent #
US 20050096004A1
Filed 09/13/2004

Current Assignee
CSR Technology Incorporated

Sponsoring Entity
CSR Technology Incorporated

Distributed data collection of satellite data  
Patent #
US 20050130590A1
Filed 10/20/2004

Current Assignee
CSR Technology Incorporated

Sponsoring Entity
CSR Technology Incorporated

GPS RF front end IC with programmable frequency synthesizer for use in wireless phones  
Patent #
US 6,931,233 B1
Filed 08/31/2001

Current Assignee
CSR Technology Incorporated

Sponsoring Entity
SiRF Technology

Shared memory architecture in GPS signal processing  
Patent #
US 6,930,634 B2
Filed 12/04/2002

Current Assignee
CSR plc

Sponsoring Entity
SiRF Technology

METHOD AND APPARATUS FOR CONTROLLING A VEHICLE USING AN OBJECT DETECTION SYSTEM AND BRAKESTEER  
Patent #
US 20050209762A1
Filed 03/18/2004

Current Assignee
Ford Global Technologies LLC

Sponsoring Entity
Ford Global Technologies LLC

Method and system for data detection in a global positioning system satellite receiver  
Patent #
US 20050264446A1
Filed 07/06/2005

Current Assignee
CSR Technology Incorporated

Sponsoring Entity
CSR Technology Incorporated

Communications system that reduces autocorrelation or crosscorrelation in weak signals  
Patent #
US 6,680,695 B2
Filed 07/20/2001

Current Assignee
CSR Technology Incorporated

Sponsoring Entity
SiRF Technology

Method and apparatus for optimally tuning a circularly polarized patch antenna after installation  
Patent #
US 6,680,703 B1
Filed 02/14/2002

Current Assignee
CSR Technology Incorporated

Sponsoring Entity
SiRF Technology

Mode determination for mobile GPS terminals  
Patent #
US 6,703,971 B2
Filed 02/21/2002

Current Assignee
CSR Technology Incorporated

Sponsoring Entity
SiRF Technology

Location services system that reduces autocorrelation or crosscorrelation in weak signals  
Patent #
US 6,707,423 B2
Filed 09/16/2002

Current Assignee
CSR Technology Incorporated

Sponsoring Entity
SiRF Technology

Communications system that reduces autocorrelation or crosscorrelation in weak signals  
Patent #
US 20040137914A1
Filed 11/12/2003

Current Assignee
CSR Technology Incorporated

Sponsoring Entity
CSR Technology Incorporated

Method and system for data detection in a global positioning system satellite receiver  
Patent #
US 20040172195A1
Filed 03/03/2004

Current Assignee
Samsung SDI Company Limited

Sponsoring Entity
Samsung SDI Company Limited

Shared memory architecture in GPS signal processing  
Patent #
US 6,526,322 B1
Filed 12/16/1999

Current Assignee
Samsung SDI Company Limited

Sponsoring Entity
SiRF Technology

Multimode global positioning system for use with wireless networks  
Patent #
US 6,519,466 B2
Filed 02/05/2002

Current Assignee
Samsung SDI Company Limited

Sponsoring Entity
SiRF Technology

Information transfer in a multimode global positioning system used with wireless networks  
Patent #
US 6,542,823 B2
Filed 04/19/2002

Current Assignee
Samsung SDI Company Limited

Sponsoring Entity
SiRF Technology

Method and apparatus for satellite positioning system (SPS) time measurement  
Patent #
US 6,583,757 B2
Filed 11/19/2001

Current Assignee
Snaptrack Incorporated

Sponsoring Entity
Snaptrack Incorporated

Method and apparatus for determining global position using almanac information  
Patent #
US 6,671,620 B1
Filed 05/18/2000

Current Assignee
CSR Technology Incorporated

Sponsoring Entity
SiRF Technology

Apparatus for reducing autocorrelation or crosscorrelation in weak CDMA signals  
Patent #
US 20020025828A1
Filed 07/20/2001

Current Assignee
CSR Technology Incorporated

Sponsoring Entity
CSR Technology Incorporated

Method and apparatus for satellite positioning system (SPS) time measurement  
Patent #
US 6,377,209 B1
Filed 03/21/2000

Current Assignee
Snaptrack Incorporated

Sponsoring Entity
Snaptrack Incorporated

Method for reducing autocorrelation or crosscorrelation in weak signals  
Patent #
US 20020064209A1
Filed 07/20/2001

Current Assignee
CSR Technology Incorporated

Sponsoring Entity
CSR Technology Incorporated

Multimode global positioning system for use with wireless networks  
Patent #
US 6,389,291 B1
Filed 02/08/2001

Current Assignee
Samsung SDI Company Limited

Sponsoring Entity
SiRF Technology

GPSbased positioning system for mobile GPS terminals  
Patent #
US 6,462,708 B1
Filed 04/05/2001

Current Assignee
CSR Technology Incorporated

Sponsoring Entity
SiRF Technology

Location services system that reduces autocorrelation or crosscorrelation in weak signals  
Patent #
US 6,466,161 B2
Filed 07/20/2001

Current Assignee
CSR Technology Incorporated

Sponsoring Entity
SiRF Technology

Method for synchronizing a radio network using end user radio terminals  
Patent #
US 20020173322A1
Filed 05/21/2002

Current Assignee
CSR Technology Incorporated

Sponsoring Entity
CSR Technology Incorporated

Autonomous hardwired tracking loop coprocessor for GPS and WAAS receiver  
Patent #
US 6,480,150 B2
Filed 05/31/2001

Current Assignee
CSR plc

Sponsoring Entity
SiRF Technology

Autonomous hardwired tracking loop coprocessor for GPS and WAAS receiver  
Patent #
US 6,278,403 B1
Filed 09/17/1999

Current Assignee
CSR plc

Sponsoring Entity
SiRF Technology

Satellite based positioning method and system for coarse location positioning  
Patent #
US 8,116,976 B2
Filed 08/07/2007

Current Assignee
CSR Technology Incorporated

Sponsoring Entity
CSR Technology Incorporated

System and method for providing location based services over a network  
Patent #
US 8,078,189 B2
Filed 03/24/2005

Current Assignee
CSR Technology Incorporated

Sponsoring Entity
CSR Technology Incorporated

Signal processing system for satellite positioning signals  
Patent #
US 8,138,972 B2
Filed 09/02/2004

Current Assignee
CSR Technology Incorporated

Sponsoring Entity
CSR Technology Incorporated

GPSbased positioning system for mobile GPS terminals  
Patent #
US 8,164,516 B2
Filed 02/18/2008

Current Assignee
Panasonic Corporation

Sponsoring Entity
Panasonic Corporation

Global positioning system receiver timeline management  
Patent #
US 8,164,517 B2
Filed 10/28/2006

Current Assignee
CSR Technology Incorporated

Sponsoring Entity
CSR Technology Incorporated

Waveform correlation result processing methods and apparatuses  
Patent #
US 8,200,446 B2
Filed 12/12/2008

Current Assignee
Qualcomm Inc.

Sponsoring Entity
Qualcomm Inc.

Distributed data collection of satellite data  
Patent #
US 8,244,271 B2
Filed 10/20/2004

Current Assignee
CSR Technology Incorporated

Sponsoring Entity
CSR Technology Incorporated

Satellite based positioning method and system for coarse location positioning  
Patent #
US 8,260,548 B2
Filed 07/30/2010

Current Assignee
CSR Technology Incorporated

Sponsoring Entity
CSR Technology Incorporated

Systems and methods for computing vertical position  
Patent #
US 8,395,542 B2
Filed 08/27/2010

Current Assignee
Trimble Navigation Limited

Sponsoring Entity
Trimble Navigation Limited

Synchronizing a radio network with end user radio terminals  
Patent #
US 8,437,693 B2
Filed 03/24/2011

Current Assignee
CSR Technology Incorporated

Sponsoring Entity
CSR Technology Incorporated

Signal processing system for satellite positioning signals  
Patent #
US 8,593,345 B2
Filed 02/20/2012

Current Assignee
CSR Technology Incorporated

Sponsoring Entity
CSR Technology Incorporated

Global positioning system receiver timeline management  
Patent #
US 8,810,450 B2
Filed 03/30/2012

Current Assignee
CSR Technology Incorporated

Sponsoring Entity
CSR Technology Incorporated

Control and features for satellite positioning system receivers  
Patent #
US 8,947,300 B2
Filed 07/29/2011

Current Assignee
CSR Technology Incorporated

Sponsoring Entity
CSR Technology Incorporated

Control and features for satellite positioning system receivers  
Patent #
US 9,869,770 B2
Filed 12/12/2014

Current Assignee
CSR Technology Incorporated

Sponsoring Entity
CSR Technology Incorporated

Spread spectrum communications with resistance to multipath at differential delays both larger and smaller than a chip width  
Patent #
US 4,862,478 A
Filed 11/06/1987

Current Assignee
DIGITAL WIRELESS CORPORATION

Sponsoring Entity
GAMBATTE INC. 1438 TULLIE ROAD ATLANTA GEORGIA A CORP. OF GA

System and method for identifying an arrival time of a communicated signal  
Patent #
US 5,488,662 A
Filed 03/10/1994

Current Assignee
Motorola Solutions Inc.

Sponsoring Entity
Motorola Inc.

Apparatus for and method of adaptively processing sonar data  
Patent #
US 5,481,503 A
Filed 12/12/1984

Current Assignee
Lockheed Martin Corporation

Sponsoring Entity
Lockheed Martin Corporation

Multipath compensation for code phase signals  
Patent #
US 5,493,588 A
Filed 12/09/1994

Current Assignee
Trimble Navigation Limited

Sponsoring Entity
Trimble Navigation Limited

Pseudorandom noise ranging receiver which compensates for multipath distortion by dynamically adjusting the time delay spacing between early and late correlators  
Patent #
US 5,495,499 A
Filed 02/03/1995

Current Assignee
Novatel Communications Incorporated

Sponsoring Entity
Novatel Communications Incorporated

Pseudorandom noise ranging receiver which compensates for multipath distortion by dynamically adjusting the time delay spacing between early and late correlators  
Patent #
US 5,390,207 A
Filed 03/24/1994

Current Assignee
Novatel Incorporated

Sponsoring Entity
Novatel Communications Incorporated

Signal timing synchronizer  
Patent #
US 5,402,450 A
Filed 01/22/1992

Current Assignee
Trimble Navigation Limited

Sponsoring Entity
Trimble Navigation Limited

Passive means for single site radio location  
Patent #
US 5,444,451 A
Filed 05/13/1994

Current Assignee
Southwest Research Institute

Sponsoring Entity
Southwest Research Institute

Timing and automatic frequency control of digital receiver using the cyclic properties of a nonlinear operation  
Patent #
US 5,282,228 A
Filed 12/09/1991

Current Assignee
HARRIS STRATEX NETWORKS CANADA ULC

Sponsoring Entity
Novatel Communications Incorporated

Equalizers  
Patent #
US 5,091,918 A
Filed 09/18/1990

Current Assignee
Orbitel Mobile Communications Ltd.

Sponsoring Entity
Plessey Overseas Limited

Multichannel digital receiver for global positioning system  
Patent #
US 5,101,416 A
Filed 11/28/1990

Current Assignee
Novatel Incorporated

Sponsoring Entity
Novatel Communications Incorporated

Digital equalization method and apparatus  
Patent #
US 5,164,959 A
Filed 01/22/1991

Current Assignee
Microelectronics Technology Inc

Sponsoring Entity
Hughes Aircraft Company

Multiplexed digital correlator  
Patent #
US 4,660,164 A
Filed 12/05/1983

Current Assignee
The United States of America As Represented By The Secretary of Agriculture

Sponsoring Entity
The United States of America As Represented By The Secretary of Agriculture

Adaptive signal processor for interference cancellation  
Patent #
US 4,608,569 A
Filed 09/09/1983

Current Assignee
Lockheed Martin Corporation

Sponsoring Entity
General Electric Company

Pseudorandomnumber code detection and tracking system  
Patent #
US 4,203,070 A
Filed 08/08/1978

Current Assignee
Charles Stark Draper Laboratory Incorporated

Sponsoring Entity
Charles Stark Draper Laboratory Incorporated

Pseudorandomnumbercodedetection and tracking system  
Patent #
US 4,203,071 A
Filed 08/08/1978

Current Assignee
Charles Stark Draper Laboratory Incorporated

Sponsoring Entity
Charles Stark Draper Laboratory Incorporated

Code synchronizing apparatus  
Patent #
US 4,168,529 A
Filed 03/29/1978

Current Assignee
The secretary of state for defence in her Britannic Majestys Government of the United Kingdom of Great Britain And Northern Ireland

Sponsoring Entity
The secretary of state for defence in her Britannic Majestys Government of the United Kingdom of Great Britain And Northern Ireland

Method and apparatus for demodulation of relative phase modulated binary data  
Patent #
US 4,007,330 A
Filed 08/13/1975

Current Assignee
Allied Corp.

Sponsoring Entity
Bunker Ramo Corporation

18 Claims
 1. A method of formation of an autocorrelation difference function of an incoming signal that reduces effects of presence of a multipath signal in the incoming signal, the method comprising the steps of:
(1) receiving an incoming digital signal s(t) that can vary with the time t and that has a digital signal bit period with a selected length Δ
τ
_{chip} ;(2) forming a first signal product difference
space="preserve" listingtype="equation">Δ
s.sub.1 (t;
τ
;
t.sub.E ;
t.sub.L ;
qE;
qL)=s(t)S.sub.d (t+τ
t.sub.E)w1(t+τ
t.sub.E ;
qE)s(t)S.sub.d (t+τ
t.sub.L)w1(t+τ
t.sub.L ;
qL),where S_{d} (t) is a selected reference signal, w1(t;
q) is a first selected, nonconstant weighting signal that may depend upon one or more parameters q, τ
is a selected time shift, t_{E} and t_{L} are first and second selected time values satisfying 0<
t_{L} t_{E} <
2Δ
τ
_{chip} ;(3) forming a second signal product difference
space="preserve" listingtype="equation">Δ
s.sub.2 (t;
τ
;
t.sub.E ;
t.sub.L ;
qE;
qL)=s(t)S.sub.d (t+τ
t.sub.E)w1 (t+τ
t.sub.E ;
qE)s(t)S.sub.d (t+τ
t.sub.L)w1 (t+τ
t.sub.L ;
qL),where w1 (t;
q) is a second selected weighting signal that may depend upon one or more parameters q;(4) selecting an integer n≧
2 and setting an accumulation A_{n1} =0;(5) examining the incoming digital signal bit value b_{n}, for the time interval defined by nΔ
τ
_{chip} ≦
t<
(n+1)Δ
τ
_{chip}, and the immediately preceding digital bit value b_{n1} of the reference signal S_{d} (t);(6) when b_{n1} ≠
b_{n}, computing the contribution of the first signal product difference over a time interval I_{n} ={t'"'"'t_{n1} +Δ
<
t'"'"'≦
t_{n} +Δ
}, where Δ
is a selected time value satisfying 0≦
Δ
<
Δ
τ
_{chip}, to an autocorrelation difference function Δ
AC#(τ
;
q) for the incoming digital signal s(t), and adding this contribution to A_{n1} to form A_{n} ;(7) when b_{n1} =b_{n}, computing the contribution of the second signal product difference over the time interval I_{n} to the autocorrelation difference function Δ
AC#(τ
;
q) for the incoming digital signal s(t), and adding this integral or sum to A_{n1} to form A_{n} ;(8) replacing the integer n by the integer n+1 and repeating steps (5), (6) and (7) at least once; (9) interpreting the accumulation A_{N} for a selected positive integer N as the autocorrelation difference function Δ
AC#(τ
;
q) for the incoming digital signal s(t);(10) determining at least one value t_{0} of the time shift variable τ
for which the autocorretation difference function Δ
AC#(τ
;
q) changes sign; and(11) interpreting the time value t=t_{0} as an estimate of the time at which a signal, which is substantially free of the presence of a multipath signal, was received.  View Dependent Claims (2, 3, 4)
 5. A method of formation of an autocorrelation difference function of an incoming signal that reduces effects of presence of a multipath signal in the incoming signal, the method comprising the steps of:
(1) receiving an incoming digital signal s(t) that can vary with the time t and that has a digital signal bit period with a selected length Δ
τ
_{chip} ;(2) forming a first signal product difference
space="preserve" listingtype="equation">Δ
s.sub.1 (t;
τ
;
t.sub.E ;
t.sub.L ;
qE;
qL)=s(t)S.sub.d (t+τ
t.sub.E)w1(t+τ
t.sub.E ;
qE)s(t)S.sub.d (tτ
t.sub.L)w1(t+τ
t.sub.L ;
qL),where S_{d} (t) is a selected reference signal, w1(t;
q) is a first selected, nonconstant weighting signal that may depend upon one or more parameters q, τ
is a selected time shift, t_{E} and t_{L} are first and second selected time values satisfying 0<
t_{L} t_{E} <
2Δ
τ
_{chip} ;(3) forming a second signal product difference
space="preserve" listingtype="equation">Δ
s.sub.2 (t;
τ
;
t.sub.E ;
t.sub.L ;
qE;
qL)=s(t)S.sub.d (t+τ
t.sub.E)w1 (t+τ
t.sub.E ;
qE)s(t)S.sub.d (t+τ
t.sub.L)w1 (t+τ
t.sub.L ;
qL),where w1 (t;
q) is a second selected weighting signal that may depend upon one or more parameters q;(4) selecting an integer n≧
3 and setting an accumulation A_{n1} =0;(5) examining the incoming digital signal bit value b_{n}, for the time interval defined by nΔ
τ
_{chip} ≦
t<
(n+1)Δ
τ
_{chip}, and the two immediately preceding digital bit values b_{n1} and b_{n2} of the reference signal S_{d} (t);(6) when b_{n2} =b_{n1} and b_{n1} ≠
b_{n}, computing the contribution of the first signal product difference over a time interval I_{n} ={t'"'"'t_{n1} +Δ
<
t'"'"' ≦
t_{n} +Δ
}, where Δ
is a selected time value satisfying 0≦
Δ
<
Δ
τ
_{chip}, to an autocorrelation difference function Δ
AC#(τ
;
q) for the incoming digital signal s(t), and adding this contribution to A_{n1} to form A_{n} ;(7) when b_{n2} ≠
b_{n1} or b_{n1} =b_{n}, computing the contribution of the second signal product difference over the time interval I_{n} to the autocorrelation difference function Δ
AC#(τ
;
q) for the incoming digital signal s(t), and adding this integral or sum to A_{n1} to form A_{n} ;(8) replacing the integer n by the integer n+1 and repeating steps (5), (6) and (7) at least once; (9) interpreting the accumulation A_{N} for a selected positive integer N as the autocorrelation difference function Δ
AC#(τ
;
q) for the incoming digital signal s(t); and(10) determining at least one value t_{0} of the time shift variable τ
for which the autocorrelation difference function Δ
AC#(τ
;
q) changes sign; and(11) interpreting the time value t=t_{0} as an estimate of the time at which a signal, which is substantially free of the presence of a multipath signal, was received.  View Dependent Claims (6, 7, 8)
 9. Apparatus for formation of an autocorrelation difference function of an incoming signal that reduces effects of presence of a multipath signal in the incoming signal, the apparatus comprising:
a signal antenna that receives an incoming digital signal s(t) that can vary with the time t and that has a digital signal bit period with a selected length Δ
τ
_{chip} ;a signal receiver/processor, including a computer, that generates a selected digital reference signal S_{d} (t), that examines a signal bit value b_{n} and the immediately preceding digital bit value b_{n1} of the reference signal S_{d} (t), and that is programmed to; (1) receive an incoming digital signal s(t) that can vary with the time t and that has a digital signal bit period with a selected length Δ
t_{chip} ;(2) form a first signal product difference
space="preserve" listingtype="equation">Δ
s.sub.1 (t;
τ
t.sub.E ;
t.sub.L ;
qE;
qL)=s(t)S.sub.d (t+τ
t.sub.E)w1(t+τ
t.sub.E ;
qE)s(t)S.sub.d (t+τ
t.sub.L)w1(t+τ
t.sub.L ;
qL),where w1(t;
q) is a first selected, nonconstant weighting signal that may depend upon one or more parameters q, τ
is a selected time shift, t_{E} and t_{L} are first and second selected time values satisfying 0<
t_{L} t_{E} <
2Δ
τ
_{chip} ;(3) form a second signal product difference
space="preserve" listingtype="equation">Δ
s.sub.2 (t;
τ
;
t.sub.E ;
t.sub.L ;
qE;
qL)=s(t)S.sub.d (t+τ
t.sub.E)w1 (t+τ
t.sub.E ;
qE)s(t)S.sub.d (t+τ
t.sub.L)w1 (t+τ
t.sub.L ;
qL),where w1 (t;
q) is a second selected weighting signal that may depend upon one or more parameters q;(4) select an integer n≧
2 and set an accumulation A_{n1} =0;(5) examine the bit value b_{n}, for a time interval defined by nΔ
τ
_{chip} ≦
t<
(n+1)Δ
τ
_{chip}, and the immediately preceding bit value b_{n1} of the reference signal S_{d} (t);(6) when b_{n1} ≠
b_{n}, compute the contribution of the first signal product difference over the time interval I_{n} ={t'"'"'t_{n1} +Δ
<
t'"'"'≦
t_{n} +Δ
}, where Δ
is a selected time value satisfying 0≦
Δ
<
Δ
τ
_{chip} to an autocorrelation difference function Δ
AC#(τ
;
q) for the incoming digital signal s(t), and adding this contribution to A_{n1} to form A_{n} ;(7) when b_{n1} =b_{n}, compute the contribution of the second signal product difference over the time interval I_{n} to the autocorrelation difference function Δ
AC#(τ
;
q) for the incoming digital signal s(t), and adding this integral or sum to A_{n1} to form A_{n} ;(8) replace the integer n by the integer n+1 and repeat steps (5), (6) and (7) at least once; (9) interpret the accumulation A_{N} for a selected positive integer N as the autocorrelation difference function Δ
AC#(τ
;
q) for the incoming digital signal s(t);(10) determine at least one value t_{0} of the time shift variable τ
for which the autocorrelation difference function Δ
AC#(τ
;
q) changes sign; and(11) interpret the time value t=t_{0} as an estimate of the time at which a signal, which is substantially free of the presence of a multipath signal, was received.  View Dependent Claims (10, 11, 12)
 13. Apparatus for formation of an autocorrelation difference function of an incoming signal that reduces effects of presence of a multipath signal in the incoming signal, the apparatus comprising:
a signal antenna that receives an incoming digital signal s(t) that can vary with the time t and that has a digital signal bit period with a selected length Δ
t_{chip} ;a signal receiver/processor, including a computer, that generates a selected digital reference signal S_{d} (t), that examines a bit value b_{n} and the two immediately preceding digital bit values b_{n1} and b_{n2} of the reference signal S_{d} (t), and that is programmed to; (1) receive an incoming digital signal s(t) that can vary with the time t and that has a digital signal bit period with a selected length Δ
τ
_{chip} ;(2) form a first signal product difference
space="preserve" listingtype="equation">Δ
s.sub.1 (t;
τ
;
t.sub.E ;
t.sub.L ;
qE;
qL)=s(t)S.sub.d (t+τ
t.sub.E)w1(t+τ
t.sub.E ;
qE)s(t)S.sub.d (t+τ
t.sub.L)w1(t+τ
t.sub.L ;
qL),where w1(t;
q) is a first selected, nonconstant weighting signal that may depend upon one or more parameters q, τ
is a selected time shift, t_{E} and t_{L} are first and second selected time values satisfying 0<
t_{L} t_{E} <
2Δ
τ
_{chip} ;(3) form a second signal product difference
space="preserve" listingtype="equation">Δ
s.sub.2 (t;
τ
;
t.sub.E ;
t.sub.L ;
qE;
qL)=s(t)S.sub.d (t+τ
t.sub.E)w1 (t+τ
t.sub.E ;
qE)s(t)S.sub.d (t+τ
t.sub.L)w1 (t+τ
t.sub.L ;
qL),where w1'"'"'(t;
q) is a second selected weighting signal that may depend upon one or more parameters q;(4) select an integer n≧
3 and set an accumulation A_{n1} =0;(5) examine the bit value b_{n}, for the time interval defined by nΔ
τ
_{chip} ≦
t≦
(n+1)Δ
τ
_{chip}, and the two immediately preceding digital bit values b_{n1} and b_{n2} of the reference signal S_{d} (t);(6) when b_{n2} =b_{n1} and b_{n1} ≠
b_{n}, compute the contribution of the first signal product difference over a time interval I_{n} ={t'"'"'t_{n1} +Δ
<
T'"'"'≦
t_{n} +Δ
}, where Δ
is a selected time value satisfying 0≦
Δ
<
Δ
τ
_{chip}, to an autocorrelation difference function Δ
AC#(τ
;
q) for the incoming digital signal s(t), and adding this contribution to A_{n1} to form A_{n} ;(7) when b_{n2} ≠
b_{n1} or b_{n1} =b_{n}, compute the contribution of the second signal product difference over the time interval I_{n} to the autocorrelation difference function Δ
AC#(τ
;
q) for the incoming digital signal s(t), and adding this integral or sum to A_{n1} to form A_{n} ;(8) replace the integer n by the integer n+1 and repeat steps (5), (6) and (7) at least once; (9) interpret the accumulation A_{N} for a selected positive integer N as the autocorrelation difference function Δ
AC#(τ
;
q) for the incoming digital signal s(t); and(10) determine at least one value t_{0} of the time shift variable T for which the autocorrelation difference function Δ
AC#(τ
;
q) changes sign; and(11) interpret the time value t=t_{0} as an estimate of the time at which a signal, which is substantially free of the presence of a multipath signal, was received.  View Dependent Claims (14, 15, 16)
 17. A method of formation of an autocorrelation difference function of an incoming signal that reduces effects of presence of a multipath signal in the incoming signal, the method comprising the steps of:
(1) receiving an incoming digital signal s(t) that can vary with the time t and that has a digital signal bit period with a selected length Δ
τ
_{chip} ;(2) forming a first signal product difference
space="preserve" listingtype="equation">Δ
s.sub.1'"'"' (t;
τ
;
t.sub.E ;
t.sub.L ;
qE;
qL)=s(t)S.sub.d (t+τ
t.sub.E)w1(t+τ
t.sub.E ;
qE)s(t)S.sub.d (t+τ
t.sub.L)w1(t+τ
t.sub.L ;
qL),where S_{d} (t) is a selected reference signal, w1(t;
q) is a first selected, nonconstant weighting signal that may depend upon one or more parameters q and that is periodic with period equal to Δ
τ
_{chip}, τ
is a selected time shift, t_{E} and t_{L} are first and second selected time values satisfying 0<
t_{L} t_{E} <
2Δ
τ
_{chip} ;(3) forming a second signal product difference
space="preserve" listingtype="equation">Δ
s.sub.2 (t;
τ
;
t.sub.E ;
t.sub.L ;
qE;
qL)=s(t)S.sub.d (t+τ
t.sub.E)w1 (t+τ
t.sub.E ;
qE)s(t)S.sub.d (t+τ
t.sub.L)w1 (t+τ
t.sub.L ;
qL),where w1 (t;
q;
k) is a second selected weighting signal that may depend upon one or more parameters q and that is periodic with period kΔ
τ
_{chip}, where k is a selected integer ≧
2;(4) selecting an integer n≧
2 and setting an accumulation A_{n1} =0;(5) examining the incoming digital signal bit value b_{n}, for the time interval defined by nΔ
τ
_{chip} ≦
t≦
(n+1)Δ
τ
_{chip}, and the immediately preceding digital bit value b_{n1} of the reference signal S_{d} (t);(6) when b_{n1} ≠
b_{n}, computing the contribution of the first signal product difference over a time interval I_{n} ={t'"'"'t_{n1} +Δ
<
t'"'"'≦
t_{n} +Δ
}, where Δ
is a selected time value satisfying 0≦
Δ
<
Δ
τ
_{chip}, to an autocorrelation difference function Δ
AC#(τ
;
q) for the incoming digital signal s(t), and adding this contribution to A_{n1} to form A_{n} ;(7) when b_{n1} =b_{n}, computing the contribution of the second signal product difference over the time interval I_{n} to an autocorrelation difference function Δ
AC#(τ
;
q) for the incoming digital signal s(t), and adding this integral or sum to A_{n1} to form A_{n} ;(8) replacing the integer n by the integer n+1 and repeating steps (5), (6) and (7) at least once; (9) interpreting the accumulation A_{N} for a selected positive integer N as an autocorrelation difference function Δ
AC#(τ
;
q) for the incoming digital signal s(t);(10) determining at least one value t_{0} of the time shift variable τ
for which the autocorrelation difference function Δ
AC#(τ
;
q) changes sign; and(11) interpreting the time value t=t_{0} as an estimate of the time at which a signal, that is substantially free of the presence of a multipath signal, was received.
 18. A method of formation of an autocorrelation difference function of an incoming signal that reduces effects of presence of a multipath signal in the incoming signal, the method comprising the steps of:
(1) receiving an incoming digital signal s(t) that can vary with the time t and that has a digital signal bit period with a selected length Δ
τ
_{chip} ;(2) forming a first signal product difference
space="preserve" listingtype="equation">Δ
s.sub.1 (t;
τ
;
t.sub.E ;
t.sub.L ;
qE;
qL)=s(t)S.sub.d (t+τ
t.sub.E)w1(t+τ
t.sub.E ;
qE)s(t)S.sub.d (t+τ
t.sub.L)w1(t+τ
t.sub.L ;
qL),where S_{d} (t) is a selected reference signal, w1(t;
q) is a first selected, nonconstant weighting signal that may depend upon one or more parameters q and that is periodic with period equal to Δ
τ
_{chip}, τ
is a selected time shift, t_{E} and t_{L} are first and second selected time values satisfying 0<
t_{L} t_{E} <
2Δ
τ
_{chip} ;(3) forming a second signal product difference
space="preserve" listingtype="equation">Δ
s.sub.2 (t;
τ
;
t.sub.E ;
t.sub.L ;
qE;
qL)=s(t)S.sub.d (t+τ
t.sub.E)w1 (t+τ
t.sub.E)s(t)S.sub.d (t+τ
t.sub.L)w1 (t+τ
t.sub.L ;
qL),where w1 (t;
q;
k) is a second selected weighting signal that may depend upon one or more parameters q and that is periodic with period kΔ
τ
hd chip, where k is a selected integer≧
2;(4) selecting an integer n≧
3 and setting an accumulation A_{n1} =0;(5) examining the bit value b_{n}, for the time interval defined by nΔ
τ
_{chip} ≦
t≦
(n+1)Δ
τ
_{chip}, and the two immediately preceding digital bit values b_{n1} and b_{n2} of the reference signal S_{d} (t);(6) when b_{n2} =b_{n1} and b_{n1} ≠
b_{n}, computing the contribution of the first signal product difference over a time interval I_{n} ={t'"'"'t_{n1} +Δ
<
t'"'"'≦
t_{n} +Δ
}, where Δ
is a selected time value satisfying 0≦
Δ
<
Δ
τ
_{chip}, to an autocorrelation difference function Δ
AC#(τ
;
q) for the incoming digital signal s(t), and adding this contribution to A_{n1} to form A_{n} ;(7) when b_{n2} ≠
b_{n1} or b_{n1} =b_{n}, computing the contribution of the second signal product difference over the time interval I_{n} to the autocorrelation difference function Δ
AC#(τ
;
q) for the incoming digital signal s(t), and adding this integral or sum to A_{n1} to form A_{n} ;(8) replacing the integer n by the integer n+1 and repeating steps (5), (6) and (7) at least once; (9) interpreting the accumulation A_{N} for a selected positive integer N as the autocorrelation difference function Δ
AC#(τ
;
q) for the incoming digital signal s(t); and(10) determining at least one value t_{0} of the time shift variable τ
for which the autocorrelation difference function Δ
AC#(τ
;
q) changes sign; and(11) interpreting the time value t=t_{0} as an estimate of the time at which a signal, that is substantially free of the presence of a multipath signal, was received.
1 Specification
This application is a divisional of 08/650,338 filed May 20, 1996.
This invention relates to a method for minimizing the deleterious effects of multipath on incoming digital spread spectrum signals that are encountered in signal receivers, using analysis of transitions of consecutive bit values in the incoming signal.
The effects of multipath are well known in communications systems. Multipath is the term used to define the secondary signals that are locally induced reflections of a primary signal that enter the receiver in question a fraction of a second later than the direct path signal, and because of the relatively short delay between the original signal and the secondary signal, induce a type of destructive interference that results in some type of impairment to the desired signal. In analog FM band automobile receivers, the effects of multipath create an annoying flutter that causes a loss of intelligibility. In television signals, the impairment is called a "ghost" image. A similar impairment occurs in other forms of analog communication. In digital systems, whether for speech or for data transmission for other purposes, multipath basically adds noise to the desired signal, resulting in either outright errors or at least, much noisier data. In spread spectrum receivers, the effects of multipath are generally found in the correlators used to achieve signal timing synchronization. In GPS or GLONASS receivers, which seek to determine location based on triangulation of range distances determined from time delay measurements made from an orbiting constellation of satellites, the effect of multipath is to induce comparatively large instantaneous errors in the time of arrival measurements which translate into large errors in the indicated positions. Removal of these errors is the subject of most of the work done by previous workers in this field. Previous researchers have sought to deal with the effects of multipath by attempting to estimate the magnitude of the error introduced, and to subtract this error or to otherwise compensate for its effects.
The methods employed to acquire and demodulate data from spread spectrum transmissions is well known in the art. See R. E. Ziemer and R. L. Peterson, Digital Communications and Spread Spectrum Systems, Macmillan Publ Co., New York, 1985, pp. 419447 for a discussion of acquisition and demodulation of spread spectrum signals. A spread spectrum GPS receiver must obtain both code and carrier synchronization in order to demodulate the desired data successfully. Issues associated with tracking and accurately demodulating a spread spectrum signal, once the signal is acquired, are discussed in many references on GPS, such as Alfred Leick, GPS Satellite Surveying, John Wiley & Sons, New York, Second Edition, 1995, and Ziemer and Peterson, op cit.
A GPS signal contains a 50 bit/second navigation message and a unique spreading code (C/A) of length 1.023 kilobits, which is transmitted at a frequency of about 1.023 Mbits/sec. Signal acquisition requires that phase lock first occur with the radio frequency carrier and that the reference or local replica signal be synchronized with the spreading code. In signal synchronization, a local replica of the particular satellite code is synchronized in time with the incoming satellite signal code.
Once the Doppler error in the downlink signal from the satellite is appropriately compensated for and signal synchronization is obtained, the navigation message in the 50 bit/second modulation that forms the composite GPS signal (direct plus multipath) can be demodulated. This navigation message contains data on the satellite ephemerides and time pulses that indicate when the transmission originated from the satellite. By measuring the difference between the local clock time and the indicated satellite time of transmission, the time delay, and thus the instantaneous distance from GPS receiver to satellite, can be obtained by multiplying this time delay by the speed of light in the ambient medium.
Signal synchronization is performed using a signal correlator. The correlator constantly compares the incoming signal with a local replica of the desired signal; a microprocessor adjusts a time shift τ of the local replica signal until satisfactory agreement is obtained. Because the incoming signal and the local replica signal are substantially identical, a measure of the degree of agreement of these two signals is often referred to as an autocorrelation function. A variety of autocorrelation functions AC(τ) are shown in various texts, and an example is shown in FIG. 1A. An autocorrelation function AC(τ) can be described according to one of the equations ##EQU1## depending upon whether integration or summation of sampled values over a suitable contribution time interval is used to compute the composite signal autocorrelation function. The length T of the contribution time interval used to compute the autocorrelation function in Eq. (1A) or (1B) is often chosen to be N times the chip length Δτ_{chip}, where N is a large positive number.
Tracking the composite satellite signal requires maintaining signal synchronization. The peak of the autocorrelation function is rounded, not pointed, due to finite bandwidth effects, so that locating a true peak is difficult. Receiver designers have, therefore, resorted to an "earlyminuslate" correlation tracking method, as discussed by W. M. Bowles in "Correlation Tracking," Charles Stark Draper Laboratory, May 1980, by Fenton et al in U.S. Pat. No. 5,101,416, and by Lennen in U.S. Pat. Nos. 5,402,450 and 5,493,588. In the earlyminuslate tracking method, a first correlator measures an equivalent autocorrelation function when the local replica signal is shifted to an "early" time t_{E} relative to the position (τ=t_{P}) of an ideal or punctual replica, and a second correlator measures a second equivalent autocorrelation function when the local replica signal is shifted to a "late" time t_{L}. Early and late replicas of the punctual autocorrelation function AC(τ;P) are illustrated in FIG. 1B. By subtracting the late autocorrelation function from the early autocorrelation function, a correlation tracking function or autocorrelation difference function ΔAC(τ) with a zero crossing, corresponding to the autocorrelation function peak can be developed, if the separations of the early and late time shifts from the punctual time shift are chosen to be equal. A representative earlyminuslate tracking function ΔAC(τ) is shown in FIG. 1C.
If the tracking or time shift variable τ for the autocorrelation difference function ΔAC(τ) lies to the left (to the right) of the zero crossing point, the system uses the presence of positive (negative) values of ΔAC(τ) to increase (decrease) the value of τ and drive the system toward the zero crossing point for ΔAC(τ). The zerocrossing point is thus easily measured and tracked, and the equivalent peak value and peak location for the autocorrelation function is easily determined. At the zerocrossing point on this doubletlike tracking function, maximum correlation occurs between the incoming signal and the local replica signal. The zerocrossing point represents the best estimate of time shift τ for signal synchronization. The internal clock time corresponding to the zero crossing point is a good estimate for time of arrival of an incoming signal at the receiver.
Superposition of an equivalent autocorrelation function for the multipath signal (reduced in magnitude and delayed in time) onto the autocorrelation function AC(τ) for the desired satellite code signal is a useful model for analyzing the effects of presence of multipath signals, as noted in the Fenton et al patent and in the Lennen patent, op. cit. Superposition of any additional signal onto the desired incoming signal, during the time period when signal correlation occurs, will distort the desired autocorrelation function AC(τ;direct) and produce an altered autocorrelation function AC(τ;composite) for the composite signal (direct plus multipath). An autocorrelation function for an uncorrupted or "pure" direct signal is shown along with a representative, attenuated and time delayed, multipath autocorrelation function with positive relative polarity, compared to the direct signal, in FIG. 2A. The autocorrelation for the composite, corrupted incoming signal is obtained by summing the two autocorrelation functions and is compared with the uncorrupted autocorrelation function in FIG. 2B. FIGS. 2C and 2D are similar graphs, showing the autocorrelation function for a multipath signal with negative relative polarity, compared to the direct signal. Any such distortion produces errors in the indicated zerocrossing point on the earlyminuslate correlation tracking function. These errors in indicated punctual time shift produce errors in the pseudorange measurements, and will in turn produce an error in the final computed estimate of location coordinates for the receiver.
Another useful and equivalent model for analyzing the effects of presence of a multipath signal computes the autocorrelation functions AC(τ;x;direct) and AC(τ;x;multipath) (x=E, L) for the pure direct signal and the pure multipath signal, forms the differences ΔAC(τ;direct) and ΔAC(τ;multipath) and adds these two difference functions to obtain the autocorrelation difference function ΔAC(τ;composite) for the composite signal.
Representative autocorrelation difference functions for a direct incoming signal and a composite incoming signal are shown in FIGS. 3B and 3D for positive relative multipath polarity and negative relative multipath polarity, respectively, compared to the direct signal. The tracking error due to presence of the multipath signal, obtained from the difference in zero crossing points for the direct signal and for the composite signal, is easily seen from these figures.
Previous work in the area of multipath amelioration has focussed on two approaches: 1) estimating the effects and compensating for multipathinduced errors, and 2) attempting to limit the effects of the estimated multipath errors. In the Lennen patents, op. cit., both approaches are described. The estimation methods seek to model the distortions to the instantaneous autocorrelation function and to create a correction term to subtract from the indicated punctual time. Estimation methods are worthwhile but can never obtain perfection, wherein all multipath effects are removed, because the multipath signals are constantly varying and corrections can only be done after the fact.
A multipath limitation method, such as described in the Lennen patent, op. cit., operates the earlyminuslate correlation tracking loop with a shorter delay between the early signal and late signal correlators than previous methods had employed. This limitation method reduces the effects of the presence of multipath substantially. In FIGS. 1B and 1C, the autocorrelation function AC(τ) and the corresponding tracking function ΔAC(τ) are shown for the case where the earlyminuslate time delay is approximately 0.15 times the width Δτ_{chip} of a digital signal bit or chip.
Several workers have analyzed correlation functions and/or have used pseudorandom signal sequences in attempting to estimate or suppress the effects of the presence of multipath signals. Examples of these are Winters in U.S. Pat. No. 4,007,330, Tomlinson in U.S. Pat. No. 4,168,529, Bowles et al in U.S. Pat. Nos. 4,203,070 and 4,203,071, Guignon et al in U.S. Pat. No. 4,550,414, Dickey et al in U.S. Pat. No. 4,608,569, Liebowitz in U.S. Pat. No. 4,660,164, Borth et al in U.S. Pat. No. 4,829,543, McIntosh in U.S. Pat. No. 4,862,478, Wales in U.S. Pat. No. 5,091,918, Fenton et al in U.S. Pat. Nos. 5,101,416, 5,390,207, 5,414,729 and 5,495,499, Cai et al in U.S. Pat. No. 5,164,959, Scott et al in U.S. Pat. No. 5,282,228, Meehan in U.S. Pat. No. 5,347,536, Lennen in U.S. Pat. Nos. 5,402,450 and 5,493,588, Johnson et al in U.S. Pat. No. 5,444,451, Kuhn et al in U.S. Pat. No. 5,481,503, and Fox et al in U.S. Pat. No. 5,488,662.
In previous methods for multipath amelioration, samples are taken of the incoming direct (desired) signal plus the incoming multipath signal(s) over the entire width of the chip, using a uniform sampling rate and without assigning any variable weighting to the samples. Further, little or no account is taken of the effect of bit value transitions for consecutive bits for the incoming digital signal. What is needed here is an approach that obtains correlation information from portions of a chip width where the effects of presence of multipath signals are suppressed, by examining bit value transitions for consecutive bits for the incoming digital signal and by use of this information to choose a (nonuniform) weighting function for formation of a modified autocorrelation function for the incoming digital signal. Preferably, the approach should be adaptable to allow suppression of a controllable amount of multipath signal contributions to the autocorrelation function and should continue to provide tracking indicators that indicate the direction of time shift required to achieve signal synchronization as the unit operates.
These needs are met by the invention, which provides a method that dynamically changes the shape of a nonuniform weighting function w(t;q), used to form an autocorrelation function for an incoming digital signal, based on the relative values, b_{n2}, b_{n1} and b_{n}, of two or three consecutive bits in a replica S_{d} (t) of a digital bit sequence representing the incoming digital direct signal, with multipath and noise absent. Here q represents one or more parameters used to define the weighting function w(t;q). A weighting function is chosen from among a population of two or more such weighting functions to suppress or deemphasize the contributions of a multipath signal to an earlyminuslate autocorrelation difference function ΔAC(τ;q)=AC(τ;E;q)AC(τ;L;q) in a selected region of the time shift variable τ. In a first embodiment, the weighting function chosen to form the autocorrelation functions AC(τ;x;q) (x=E, L) and the difference function ΔAC(τ;q) will vary with the relationship between two consecutive bit values b_{n1} and b_{n}. If a bit value transition occurs so that b_{n1} ≠b_{n}, a first weighting function is used to form the autocorrelation difference function. If no bit value transition occurs so that b_{n1} =b_{n}, a second weighting function is used that suppresses the corresponding contribution of this region of the time domain to the autocorrelation difference function ΔAC(τ;q). Using this approach, the noise contribution is suppressed relative to the signal contribution so that the signaltonoise ratio (SNR) is enhanced.
In a second embodiment, the relationship between three consecutive bit values b_{n2}, b_{n1} and b_{n} is examined. If b_{n2} =b_{n1} ≠b_{n}, a first weighting function is used that enhances the contribution of the direct signal and/or reduces the contribution of a multipath signal that may be present in the composite signal; and if b_{n2} ≠b_{n1}, or if b_{n1} =b_{n}, or if both of these conditions occur, a second weighting function is chosen that suppresses the contribution of this portion of the time domain to the autocorrelation difference function ΔAC(τ;q). With this approach, the contribution of a multipath signal to the autocorrelation difference function is suppressed. The choice of weighting function thus varies dynamically as the bit values of two consecutive bit values, or three consecutive bit values, change.
In a third embodiment, if a bit value transition occurs so that b_{n1} ≠b_{n}, a first weighting function is used to form the autocorrelation difference function; and if no bit value transition occurs so that b_{n1} =b_{n}, a second weighting function, related to the first weighting function but with changed periodicity, is used to form the autocorrelation difference function.
In a fourth embodiment, if b_{n2} =b_{n1} ≠b_{n}, a first weighting function is used to form the autocorrelation difference function; and if b_{n2} ≠b_{n1} and/or b_{n1} =b_{n}, a second weighting function, related to the first weighting function but with changed periodicity, is used to form the autocorrelation difference function.
The invention begins with the realization, utilized in a companion patent application, U.S. Ser. No. 08/650,331, by the title "Suppression of Multipath Signal Effects", filed May 20, 1996 and assigned to the same assignee, that the useful information to be obtained from the incoming digital composite signal is available primarily in digital signal bit value transition regions, where the bit value for the incoming digital direct signal S_{d} (t) changes state. Contributions from samples that contain no useful information, but that do contain multipath signal effects that produce erroneous distortions in the corresponding autocorrelation function and in the resulting correlation tracking function, are ignored or suppressed. By ignoring or suppressing the contribution of samples taken in regions where no transition in a signal bit value can occur, the receiver can suppress the effects of multipath signals. The portion of a chip width where the autocorrelation difference function ΔAC(τ;q) is substantially zero is much larger than in any of the previous methods for ameliorating the presence of multipath signals. Previous methods produced wider regions of nonzero ΔAC(τ;q) for the desired signal. The invention, which may be characterized as nonuniform weighting of the sampling function used to create AC(τ;x;q), provides a method for substantially eliminating the possibility for experiencing multipath effects in correlation tracking.
A tracking point or zerocrossing point τ=t_{P} associated with the difference ΔAC(τ) of the autocorrelation functions AC(τ;E) and AC(τ;L) is determinable as before, but with reduced contributions from the direct signal and from the multipath signal in a selected central region that does not include the tracking point τ=t_{P}. This central region does not include any time shift values t for which the instantaneous value of a reference or local replica digital direct signal S_{d} (t+τ) can make a signal bit value transition (from 0 to 1, or from 1 to 0) relative to the instantaneous value of an incoming digital composite signal s(t). Thus, little or no qualitative information is lost by deemphasizing the contribution of this central region to a corresponding autocorrelation function. However, much of the important information arising from presence of the multipath signal is also present in this central region, and this multipath information is suppressed or deemphasized by suppressing the contribution of this central region to the autocorrelation function. Location of the "true" tracking point for the direct signal (absent multipath) is unchanged or is changed only minimally by this approach; the contribution of the multipath signal in the selected central region is reduced to substantially zero.
The invention in the related patent application, 08/650,331, suppresses the contribution of a central region to the computed autocorrelation function AC(τ;x;qx) (x=E, P or L) for the composite signal by imposing a nonuniform weighting function w(t;q) in the integral or sum of the digital signal product s(t)S_{d} (t+τ) over a chosen time interval t1≦t≦t2, to produce an adjusted or modified autocorrelation function. In one embodiment, the weighting function w(t;q) is positive (or negative) in signal bit value transition regions (t≈t1 and t≈t2=t1+Δτ_{chip}) and tends monotonically toward zero as τ approaches an intermediate time t3 (t1<t3<t2) from either side; this type of nonuniform weighting function is referred to as a "notch" function. The chip halfwidth Δτ_{chip} /2=(t2t1)/2 might be chosen to be the inverse of twice the carrier frequency, (2 f_{carrier})^{1}, associated with the reference digital signal S_{d} (t). The contributions of different portions of the "contribution interval", t1<t<t2, are weighted nonuniformly to reduce the effects of a multipath signal in a central region surrounding a tracking point, as discussed above. In another embodiment, the weighting function w(t;q) is chosen to be an "antinotch" function, which is defined as a constant minus a notch function. In another embodiment, the weighting function w(t;q) is unrestricted and may have positive, zero and negative values anywhere in the interval t1<t<t2.
One embodiment of the invention disclosed and claimed in this patent application improves the analysis of the autocorrelation functions AC(τ;x;qx) (x=E, P or L) and of the autocorrelation difference function ΔAC(τ;q) by providing at least two distinct weighting functions W1(t;q) and W1#(t;q), at least one of which is nonuniform, and by selecting which weighting function will be used at a given time to compute AC(τ;x;qx) and ΔAC(τ;q) based upon characteristics of two, three or more consecutive incoming digital signal bit values, b_{n2}, b_{n1} and b_{n}. In a first embodiment, if no bit transition occurs between bit value b_{n1} and bit value b_{n}, the contribution to the autocorrelation function AC(τ;x;qx) of the time interval t≈t_{n} (corresponding to the bit region where b_{n1} changes to b_{n}) is ignored or drastically suppressed, by use of a different weighting function than the weighting function used if b_{n1} ≠b_{n}. In a second embodiment, if (1) no bit value transition occurred between bit value b_{n2} and bit value b_{n1} (b_{n2} =b_{n1}) and (2) a bit value transition occurs between bit value b_{n1} and bit value b_{n} (b_{n1} ≠b_{n}), a first weighting function w1(t;q) is used to compute the autocorrelation functions AC(τ;x;qx) (x=E, L) and the autocorrelation difference function ΔAC(τ). If either or both of these two conditions is not both met, a second weighting function w1 (t;q) (which may be identically zero) is used to compute the autocorrelation functions AC(τ;x;qx) and the autocorrelation difference function ΔAC(τ;q). One result of this approach is that the contribution of any multipath signal that may be present in a composite signal (direct plus multipath) is further suppressed.
Another embodiment of the invention provides a method for decoding a signal to reduce signal distortion, using the steps of: (1) receiving a signal having at least two consecutive bits, with each bit having a value that is represented in the received signal over a time interval; (2) generating a reference signal having at least two consecutive bits, for matching the received signal; (3) generating a weighting signal, having an amplitude that varies over time, that conforms to (a) a first sequence of weighting values when each bit of a consecutive bit sequence of the reference signal is part of a predetermined bit sequence and (b) a second sequence of weighting values when at least one bit of a consecutive bit sequence of the reference signal is not part of a predetermined bit sequence; and (4) mixing the received signal, the reference signal and the weighting signal to determine a timing relationship between the received signal and the reference signal.
In addition, the first and second sequences of weighting values can be represented in the weighting signal over a selected time interval having a length that is less than the length of the time interval (one chip) for representing one of the bit values, and the selected time interval can be begun a selected time increment before the beginning of a respective time interval for each reference signal bit.
In one version, a reference signal bit sequence having two consecutive bits is examined. The first sequence of weighting values is used when the two consecutive bits have different bit values, and the second sequence of weighting values is used when the two consecutive bits have the same bit value.
In another version, a reference signal bit sequence having three consecutive bits is examined. The first sequence of weighting values is used when (i) the first and second consecutive bits of a threebit sequence have the same bit value and (ii) the third consecutive bit in the threebit sequence has a different bit value. The second sequence of weighting values is used when at least one of the conditions (i) and (ii) for use of the first sequence of weighting values is not satisfied.
FIGS. 1A and 1B are graphical views of a representative autocorrelation function AC(τ;x) (x=P, E, L).
FIG. 1C is a graphical view of an earlyminuslate autocorrelation function difference signal ΔAC(τ)=AC(τ;E)AC(τ;L) formed from the autocorrelation functions shown in FIG. 1B. In all earlyminuslate autocorrelation difference functions shown herein, unless otherwise stated, the earlyminuslate spacing Δt_{LE} =t_{L} t_{E} is 15 percent of the chip width Δτ_{chip} for definiteness. Other values for Δt_{LE} can be used here.
FIGS. 2A and 2B compare an autocorrelation function AC(τ;direct), formed using only a direct signal, with an autocorrelation function AC(τ;multipath) and AC(τ;composite), respectively, where the multipath signal included in the composite signal has positive polarity relative to the direct signal.
FIGS. 2C and 2D compare an autocorrelation function AC(τ;direct), formed using only a direct signal, with an autocorrelation function AC(τ;multipath) and AC(τ;composite), respectively, where the multipath signal included in the composite signal has negative polarity relative to the direct signal.
FIGS. 3A and 3B compare an autocorrelation difference function ΔAC(τ;direct), formed using only a direct signal, with an autocorrelation difference function ΔAC(τ;multipath) and ΔAC(τ;composite), respectively, where the multipath signal in the composite signal has positive polarity relative to the direct signal.
FIGS. 3C and 3D compare an autocorrelation difference function ΔAC(τ;direct), formed using only a direct signal, with an autocorrelation difference function ΔAC(τ;multipath) and ΔAC(τ;composite), respectively, where the multipath signal in the composite signal has negative polarity relative to the direct signal.
FIGS. 4B and 4C are graphical views of punctual (P), early (E) and late (L) autocorrelation functions AC0(τ;x) and of an earlylate autocorrelation difference function ΔAC0(τ), computed conventionally. FIG. 4A illustrates a uniform weighting function w0=1.
FIGS. 5B and 5C are graphical views comparing an autocorrelation function AC1(τ;x;qx) (x=E, P or L) and an earlylate autocorrelation difference function ΔAC1(τ;q), computed using a nonuniform weighting function (twovalue step function) w1 shown in FIG. 5A, with AC0(τ;x) and ΔAC0(τ;x). FIG. 5D illustrates a multiinterval weighting function.
FIGS. 616 illustrate other nonuniform weighting functions that can be used with the invention to suppress or to emphasize multipath effects.
FIGS. 17A and 17B show representative autocorrelation difference functions, before and after application of the invention, respectively.
FIGS. 18A, 18B, 18C, 18D and 18E illustrate a weighting function for noise reduction according to the invention.
FIG. 19 is a flow chart of one embodiment of a suitable procedure according to the invention.
FIG. 20 is a flow chart of a second embodiment of a suitable procedure according to the invention.
FIGS. 21 and 23 are schematic views of approaches suitable for implementing the twoconsecutivebit and the threeconsecutivebit procedures, respectively.
FIGS. 22A22E and 24A24L are timing diagrams for signals produced by the apparati in FIGS. 21 and 23, respectively.
FIG. 25 is a schematic view of other apparatus for practising the invention.
FIG. 26 illustrates an apparatus suitable for implementing a twoconsecutivebit procedure according to the invention, using readily available digital signal processor chips.
FIG. 27A is a graphical view of early and late replicas of a direct digital signal.
FIG. 27B is a graphical view of early and late replicas of a direct digital signal, time delayed relative to the signals shown in FIG. 27A.
FIG. 27C is a graphical view of early and late replicas of a direct digital signal, time advanced relative to the signals shown in FIG. 27A.
FIGS. 27D and 27E are graphical views of early and late weighting functions obtained from combining uniformly weighted correlators.
FIG. 28 illustrates another apparatus suitable for implementing a twoconsecutivebit procedure according to the invention.
In a conventional approach, no variable weighting is used in computing the autocorrelation function. Equivalently, the weighting function applied to the integrand or summand is constant: w0(t)=g (constant)>0, with g=1 preferred, across the contribution interval, for example, t_{P} Δτ_{chip} /2≦t≦t_{P} +Δ96 _{chip} /2, as illustrated in FIG. 4A. The quantity Δτ_{chip} is the temporal length of one bit in the incoming digital signal of interest. An autocorrelation function ΔC0(τ), computed without a variable weighting function, has small or vanishing values outside the interval t_{P} Δτ_{chip} ≦t≦t_{P} +Δτ_{chip}, due to the properties of pseudorandom sequences that are used for the direct signal. As the autocorrelation displacement variable τ increases from τ=t_{P} Δτ_{chip} to τ=t_{P} +Δτ_{chip}, the autocorrelation function increases approximately linearly to a maximum value at a tracking point τ≈t_{P}, then decreases approximately linearly beyond the tracking point (FIG. 1A). This conventional approach is equivalent to using the uniform weighting function w0(t), shown in FIG. 4A, in computing a conventional autocorrelation function AC0(τ;x) (x=E, P or L), shown in FIG. 4B, and in computing a conventional earlyminuslate autocorrelation difference function ΔAC0(τ), illustrated in FIG. 4C. The time shift values τ=t_{E} and τ =t_{L} chosen for the approximate peak values of the replica autocorrelation functions AC0(τ;x) (x=E or L) satisfy the constraints t_{P} Δτ_{chip} <t_{E} <t_{P} <t_{L} <t_{P} +Δτ_{chip}, where τ=t_{P} is the time shift value for the actual peak in the measured autocorrelation function. The time shift values τ=t_{E} and τ=t_{L} may be, but need not be, chosen symmetrically about time shift value τ=t_{P}.
Assume that a multipath signal S_{m} (t;Δt_{m}) is present, with a time delay Δt_{m} and reduced in magnitude relative to the direct signal S_{d} (t), in the incoming digital composite signal s(t)=S_{d} (t)+S_{m} (t;Δt_{m}). The timedelayed and reducedamplitude multipath signal S_{m} (t;Δt_{m})≈χS_{d} (tt_{P} Δt_{m}) (χ<1) is also summed across the contribution time interval in formation of the autocorrelation signal AC(τ;x) (x=E, P or L). Because of linear superposition, the multipath signal S_{m} (t;Δt_{m}) contributes a proportional amount (reduced in amplitude and delayed in time as indicated) to the early and late autocorrelation functions AC(τ;E) and AC(τ;L) and to the difference function ΔAC(τ)=AC(τ;E)AC(τ;L) for the incoming digital composite signal s(t). Typical multipath signal contributions to AC(τ;x) (x=E, L) are illustrated in FIGS. 2A, 2B, 2C and 2D.
The polarity of a multipath signal S_{m} (tΔt_{m}) that is part of an incoming digital composite signal may be positive or negative relative to the desired digital direct signal S_{d} (t). A multipath signal with positive relative polarity will add to the direct signal contribution to the autocorrelation function for a composite signal. FIGS. 2A and 2C compare the autocorrelation function AC(τ;x;direct) with the multipathonly functions AC(τ;x;multipath;positive) and AC(τ;x;multipath;negative) for a "bare" multipath signal having positive and negative multipath relative polarity, respectively. A multipath signal with positive (negative) multipath relative polarity will add to (subtract from) the direct signal contribution in the autocorrelation function AC(τ;x;composite). FIGS. 2B and 2D compare the autocorrelation function ΔAC(τ;x;direct) with the respective composite signal functions AC(τ;x;composite;positive) and ΔAC(τ;x;composite;negative).
Because the effects of presence of a multipath signal are additive, one can analyze the autocorrelation difference functions by using the sum ΔAC(τ;direct)+ΔAC(τ;multipath) or by forming the sum composite signal=direct signal+multipath signal before the autocorrelation difference function ΔAC(τ;composite) is computed.
FIGS. 3A and 3C graphically illustrate the autocorrelation difference functions ΔAC(τ;direct) (solid curve) and ΔAC(τ;multipath) (broken line curve) for two representative situations, using the constant weighting function w0(t)=1 in FIG. 4A, where the multipath signal has positive and negative relative polarity, respectively. FIGS. 3B and 3D graphically compare ΔAC(τ;direct) and ΔAC(τ;composite) for positive and negative relative multipath polarity, respectively. Note that the apparent tracking or zero crossing point, where ΔAC(τ;composite) for the composite signal is shifted to the right (t_{P+}) or to the left (t_{P}) relative to the desired tracking point τ=t_{P} for the direct signal, depending in part on the relative multipath polarity.
In one embodiment of the invention, illustrated in FIGS. 5A, 5B and 5C, a nonuniform weighting function w1(t;g1;t1'"'"';t2'"'"') is used that may be continuous or discontinuous and is not necessarily symmetric. The endpoints t1 and t2 of the time interval used for definition of the weighting function, such as w1(t;g1;t1'"'"';t2'"'"'), satisfy the constraints
t1≈t.sub.P, (2A)
t2=t1+Δτ.sub.chip, (2B)
For notational convenience herein, a set of weighting function parameters, such as g1, t1'"'"' and t2'"'"' for a nonuniform weighting function, such as w1, will often be denoted by a symbol "q". With this notation adopted, this weighting function becomes w1(t;q). In FIG. 5A, the weighting function w1(t;q) is a twovalue step function that is nonzero only in regions near t=t_{P}, and near t =t_{P} +Δτ_{chip}. In a center region, away from the tracking points τ=t_{P} and τ=t_{P} +Δτ_{chip}, where no signal bit transition can occur, little or no new qualitative information is presented, and use of a nonuniform weighting function w1(t;q) suppresses or deemphasizes this largely redundant information, as discussed above.
A weighting function, such as w1(t;q), is preferably extended periodically, using a prescription such as
w1(t;q)=w1(t+n Δτ.sub.chip ;q) (n=0,±1,±2, . . . ).(3)
Where nonuniform weighting is used, an autocorrelation function is formed according to one of the prescriptions ##EQU2## depending upon whether integration or summation of sampled values over a suitable contribution time interval is used to compute the composite autocorrelation function. The length T of the time interval used to compute the weighted autocorrelation function in Eq. (4A) or (4B) is often chosen to be N times the chip length Δτ_{chip}, where N is a large positive integer.
FIG. 5B compares the approximate forms of the conventional punctual autocorrelation function AC0(τ;P), which uses a constant weighting function (w0(t)=1), with an adjusted punctual autocorrelation function AC1(τ;x;q) (x=P), which uses the nonuniform weighting function w1(t;q). The early and late autocorrelation functions AC1(τ;E;q) and AC1(τ;L;q) are formed in a similar manner and have the same shape as the punctual autocorrelation function AC1(τ;P;q).
FIG. 5C illustrates the approximate form of the autocorrelation difference function ΔAC1(τ;q)=AC1(τ;E;q)AC1(τ;L;q), which should be compared with the corresponding autocorrelation function difference ΔAC0(τ), shown in FIG. 4C and repeated as a dashed line curve in FIG. 5C, that uses a uniform weighting function w0(t)=1. The spacing intervals Δt_{PE} =t_{P} t_{E} and Δt_{LP} =t_{L} t_{P} are selected to be equal in each of FIGS. 4C and 5C. If the time shift variable τ for the autocorrelation difference function ΔAC1(τ;q) lies a small amount to the left (to the right) of the peak correlation location τ=t_{P} or zero crossing value, the tracking system uses the presence of positive (negative) values of ΔAC1(τ;q) near this peak correlation location to increase (decrease) the value of the shift variable τ and drive the autocorrelation function AC1(τ;x;q) toward its peak correlation value, where ΔAC1(τ;q)=0.
More generally, a nonuniform weighting function w1'"'"'(t;g1;t1'"'"';t2'"'"';t2"; . . . ; tn'"'"';tn";t(n+1)'"'"') that has a sequence of steps of equal amplitude (=g1) in the time intervals t1<t<t1'"'"', tk'"'"'<t<tk" (k=3, . . . , Q; Q≧3) and t2'"'"'<t<t2, illustrated in FIG. 5D, can be used here.
Where the nonzero amplitudes of the steps are equal, as in FIG. 5A or 5D, use of a nonuniform weighting function, such as w1(t;q) or w1'"'"'(t;q), is equivalent to use of nonuniform sampling density in forming the autocorrelation function AC(τ;q) in Eq. (4A) or (4B). In this nonuniform sampling approach, samples of incoming digital signal values in the regions t1 <t<t1'"'"', tk'"'"'<t<tk" and t2'"'"'<t<t2 are given a uniform, nonzero weight g, and samples in all other regions are assigned a weight of zero. Alternatively, nonuniform weighting and nonuniform sampling may be combined to compute an autocorrelation function.
Other suitable nonuniform weighting functions w(t;q), each having one or more parameters q to define the function, are illustrated in FIGS. 616. Each of the weighting functions w2(t;q) and w3(t;q) in FIGS. 6 and 7 is a threestep weighting function, with amplitudes g1, g2 and g3, with one of these amplitudes being zero in FIG. 6. FIG. 8 illustrates a fivestep weighting function w4(t;q). FIGS. 9 and 10 illustrate a triangular weighting function w5(t;q) and a power law weighting function w6(t;q), respectively. FIGS. 11 and 12 illustrate a first situation in which the weighting functions w7(t;E;q) and w7(t;L;q), used to form the autocorrelation functions AC(τ;E) and AC(τ;L), are chosen independently. FIGS. 512 illustrate nonuniform weighting functions wm(t;q) (m=1, 2, . . . , 7) that are "notch" functions. A "notch" function has positive values w(t1;q) and w(t2;q) at the two ends of the defining interval t1≦t≦t2 and decreases monotonically toward zero as the time variable t approaches some intermediate value t3 (t1≦t3≦t2) from either direction within the defining time interval.
Each of FIGS. 13 and 14 illustrates a nonuniform weighting function, w8(t;q) and w9(t;q), that is an "antinotch" function, having the form w(t;q)=g0 w(t;q;notch), where g0 is a constant and w(t;q;notch) is a notch function. An antinotch function can be used to emphasize, rather than suppress, multipath effects in some situations.
FIG. 15 illustrates another suitable nonuniform weighting function w10(t;q) that is neither a notch function nor an antinotch function. The nonuniform weighting function w(t;q) need not be nonnegative everywhere on the defining interval t1≦t≦t2. FIG. 16 illustrates a suitable nonuniform weighting function w11(t;q) that is positive in some regions and negative in some other regions of the defining interval.
In the preceding analysis, the correlator spacings Δt_{PE} =t_{P} t_{E} and Δt_{LP} =t_{L} t_{P} have been chosen to be equal, for definiteness, but this is not required. It is generally required here that 0<Δt_{LE} =Δt_{PE} +Δt_{LP} <2Δτ_{chip}.
A multipath signal S_{m} (t;Δt_{m}), if it is present in an incoming digital composite signal s(t), usually arrives "late," that is, after the direct signal S_{d} (t) has begun to arrive. FIG. 17A shows an autocorrelation difference function ΔAC(τ;q), using the weighting function w1(t;q) shown in FIG. 5A, before application of the invention. Ideally, after application of the invention ΔAC(τ;q) appears as in FIG. 17B, where the "early spike" ES has been removed, and only the "late spike" LS (which is irrelevant and thus ignored) and the "central spike" CS remain. This is accomplished in the invention by dynamically changing the weighting function used to compute the autocorrelation functions AC(τ;x;qx), depending upon the characteristics of a portion of the incoming digital signal that is presently arriving.
Comparison of consecutive incoming digital signal bit values is useful in reducing the noise associated with computation of the autocorrelation function and the autocorrelation difference function. Where two consecutive incoming digital signal bit values, such as b_{n1} and b_{n}, differ from each other, a bit value transition has occurred, and useful information is present that is needed to maintain tracking of the incoming digital signal: the signal bit values near t=t_{P} for both bits should be sampled. Where two consecutive signal bit values coincide, no transition has occurred and no additional useful information is present. Sampling the bit signal values near t=t_{P} for both bits in this situation will merely introduce additional noise into the computations. A digital direct signal S_{d} (t), such as the one shown in FIG. 18A, represents the baseband incoming digital direct signal. In FIG. 18A, a bit value transition for S_{d} (t) occurs near t=t1 but does not occur near t=t2. The nonzero step, of amplitude g1, for the nonuniform weighting function w1(t;q) is two clock pulses wide and overlaps the time interval endpoints t=t1 and t=t2=t1+Δt_{chip}, as shown in FIG. 18B. A sequence of clock pulses with pulsetopulse separation Δt(clock) (usually<Δτ_{chip}) is shown in FIG. 18C. The weighting function w1(t;q) is superimposed on each of an early replica and a late replica of the digital direct signal S_{d} (t+τ) in FIGS. 18D and 18E, where the correlator spacings satisfy Δt_{PE=}Δt_{LP} ≦Δt(clock)/2. The weighting function w1(t;q) overlaps the early replica signal domain in the time intervals A1 and B1 and overlaps the late replica signal domain in the time intervals B1 and C1, where a bit value transition occurs in the bit sequence for the incoming digital direct signal S_{d} (t). The periodic extension of the weighting function w1(t;q) overlaps the early replica signal domain in the time intervals A2 and B2 and overlaps the late replica signal domain in the time intervals B2 and C2, where no bit value transition occurs in the bit sequence for the incoming digital direct signal S_{d} (t). Let EA and EB (A=A1, A2; B=B1, B2) denote the sampled value of the early replica signal S_{d} (t+τt_{E}) in the time intervals A and B, and let LB and LC (B=B1, B2; C=C1, C2) denote the sampled value of the late replica signal S_{d} (t+τt_{L}) in the time intervals B and C. Let IA, IB and IC denote the bit values of the incoming digital direct signal S_{d} (t) in the time intervals A, B and C, respectively.
The contributions to the autocorrelation difference function ΔAC1(τ;q) of the early and late replica signals in the regions A1, B1 and C1 are
Δs=g1 EA1 IA1+g1 EB1 IB1g1 LB1 IB1g1 LC1 IC1. (5)
If noise is absent, the value of Δs is zero if the system is tracking perfectly. However, if the system is not tracking perfectly, the value of Δs will be positive or negative. The net result from the difference function formed in Eq. (5) is a positive or negative contribution to ΔAC1(τ;q) from the time intervals A1, B1 and C1 and will indicate whether the value of the tracking variable τ should be increased or decreased to drive the value Δs=Δs(τ) toward zero.
The contributions to the autocorrelation difference function ΔAC1(τ;q) of the early and late replica signals in the regions A2, B2 and C2 are
Δs=g1 EA2 IA2+g1 EB2 IB2g1 LB2 IB2g1 LC2 IC2=0, (6)
because no bit value transition occurs in the incoming digital direct signal S_{d} (t), and thus in the early and late replica signals, in any of the time intervals A2, B2 and C2. The net result from the difference Δs formed in Eq. (6) is zero contribution to ΔAC1(τ;q) from the time intervals A2, B2 and C2. However, noise will contribute to each of the differences formed in Eqs. (5) and (6). For these reasons, if tracking is not perfect, the nonzero contribution of the difference in Eq. (5) to ΔAC1(τ;q), where a bit value transition occurs in the incoming digital direct signal S_{d} (t), should be included in the computation, but the net zero contribution of the difference in Eq. (6) should not be included. Thus, for narrow correlator spacings, if the contribution of noise to a modified autocorrelation difference function ΔAC1#(τ;q) is to be minimized, the signal product difference
Δs.sub.EL (t)=w1(t+τt.sub.E ;q)s(t)S.sub.d (t+τt.sub.E)w1(t+τt.sub.L ;q)s(t)S.sub.d (t+τt.sub.L),(7)
which contributes to the integral or sum used to compute the autocorrelation difference function ΔAC1#(τ;q), should be included near a possible signal bit value transition point t≈t1 or t≈t2 only if the bit value transition for the incoming digital direct signal S_{d} (t) satisfies
Δb.sub.n =1, (8)
for the incoming digital signal at or near that time point, where the bit value change function Δb_{n} is defined by ##EQU3## One method of insuring that the contribution of the signal product difference Δs_{EL} to the autocorrelation difference function ΔAC1#(τ;q) is included only from regions where an incoming digital direct signal bit value transition occurs, is to use the weighting function
w(t.sub.n ;q)=w1(t.sub.n ;q) if Δb.sub.n =1, (10A)
w(t.sub.n ;q)=w1 (t.sub.n ;q) if Δb.sub.n =0, (10B)
where w1 (t;q) is a selected weighting function that is different from w1(t;q) and that suppresses the contribution of the signal product difference Δs_{EL} in the time interval t_{n} Δτ_{chip} /2≦t<t_{n} +Δτ_{chip} /2. One suitable choice is w1 (t;q)=0. This approach will minimize the noise contribution from any region that does not affect location of the zero crossing or tracking point for the autocorrelation difference function ΔAC1(τ;q).
FIG. 19 is a flow chart illustrating suitable procedural steps that can be taken for analysis of an incoming digital signal according to one embodiment of the invention. In step 140, the incoming digital composite signal s(t) is received. In step 141, a replica of the expected bit sequence for the incoming digital direct signal S_{d} (t) is generated. In step 142, two consecutive signal bits, with signal bit values b_{n1} and b_{n}, are received and examined for the digital direct signal S_{d} (t). In step 143, the bit value change function Δb_{n} is computed. In step 145, the system determines whether Δb_{n} =1. If Δb_{n} =1, a first nonuniform weighting function W1(t;q) is used to compute a contribution to the autocorrelation function AC1#(τ;x;qx) (x=E, P or L) and to an autocorrelation difference function ΔAC1#(τ;q), in step 147, by integration or summation of the quantities W1(t+τt_{E} ;q)s(t)S_{d} (t+τt_{E}) and W1(t+τt_{L} ;q)s(t)S_{d} (t+τt_{L}) over a time interval I_{n}, of length Δτ_{chip} and defined by
I.sub.n ={tt.sub.n1 +Δ<t≦t.sub.n +Δ},(11)
where Δ is a selected time value satisfying 0≦Δ<Δτ_{chip}. If Δb_{n} =0, a second weighting function W1 (t;q) (not necessarily nonuniform) is used to form a contribution to the autocorrelation function AC1#(τ;x;qx) (x=E, P or L) and the autocorrelation difference function ΔAC1#(τ;q), in step 149, by integration or summation of the quantities W1 (t+τt_{E} ;q)s(t)S_{d} (t+τt_{E}) and W1 (t+τt_{L} ;q)s(t)S_{d} (t+τt_{L}) over the time interval I_{n}.
Comparison of consecutive incoming digital direct signal bit values is also useful in suppressing the multipath signal contribution in computation of the autocorrelation function and the autocorrelation difference function. Where three consecutive incoming digital direct signal bit values, b_{n2}, b_{n1} and b_{n}, satisfy the conditions b_{n2} =b_{n1} ≠b_{n}, a bit value transition (b_{n1} →b_{n}) has occurred, and useful information is present that is needed to maintain tracking of the incoming digital signal: signal bit values for both bits should be sampled. Where b_{n1} =b_{n}, no transition has occurred and no additional useful information is present. Sampling the bit signal values for both bits in this situation will merely introduce additional noise into the computations. Where b_{n2} ≠b_{n1}, a bit value transition in the multipath signal, delayed in time by approximately Δτ_{chip} relative to a bit value transition in the corresponding digital direct signal, will appear near the bit value transition point t=t_{n}. Sampling the bit signal values for both bits in this situation will introduce additional multipath noise into the computations. The analysis is similar to the analysis associated with FIGS. 18A18E. Table 1 indicates the eight possibilities for three consecutive bit values, b_{n2}, b_{n1} and b_{n}, for an incoming digital direct signal, where b_{n} is the present bit value and the two consecutive preceding bits (having values b_{n2} and b_{n1}) have already arrived at the signal receiver.
TABLE 1______________________________________Bit Value Transitions (n2)to(n1) bit (n1)ton bit Weightingb.sub.n2b.sub.n1 b.sub.n transition occurs transition occurs used______________________________________0 0 0 No No w10 0 1 No Yes w10 1 0 Yes Yes w10 1 1 Yes No w11 0 0 Yes No w11 0 1 Yes Yes w11 1 0 No Yes w11 1 1 No No w1______________________________________
Here, w1(t;g1;t1'"'"';t2'"'"') is the weighting function shown in FIG. 5A, or some other suitable nonuniform weighting function, and w1 A(t;q) is an alternative weighting function (e.g., w1 (t;q)=0). Other choices of the weighting function w1 (t;q) can be used.
One motivation for this approach is as follows. If a multipath signal S_{m} (t;Δt_{m}), with corresponding time delay Δt_{m}, is present in the incoming digital composite signal s(t), this multipath signal will contribute significantly to computation of an autocorrelation difference function ΔAC#(τ;q) only if the multipath signal delay time satisfies Δt_{m} ≦Δτ_{chip} +Δt_{LE} /2, where Δt_{LE} =Δt_{PE} +Δt_{LP}. If Δt_{m} >Δτ_{chip} +Δt_{LE} /2, this multipath signal contributes nothing to computation of an autocorrelation difference function ΔAC#(τ;q), because of the nature of pseudorandom sequences.
If a bit value transition has not occurred at an immediately preceding bit value transition time (i.e., b_{n2} =b_{n1}, or Δb_{n1} =0) and a bit value transition occurs at the present bit value transition time (i.e., b_{n1} ≠b_{n}, b_{n} , or Δb_{n} =1), the contribution of this time interval I_{n} is emphasized by using the weighting function w1(t;q) to compute the autocorrelation function AC#(τ;x;qx). If a bit value transition has occurred at the immediately preceding bit value transition time (t=t_{n1}, where Δb_{n1} =1), a multipath signal S_{m} (t;Δt_{m}) with an associated time delay Δt_{m} ≈Δτ_{chip} will arrive near the current bit value transition interval (t≈t_{n}). In this situation, the contribution of this current bit value transition interval I_{n} to the autocorrelation function AC#(τ;x;qx) should be deemphasized to deemphasize the signal information arising from the multipath signal bit value transition (with the associated time delay Δt_{m} ≈Δt_{chip}). Here, deemphasis is achieved by using the weighting function w1 (t;q)=0. However, any other weighting function w1 (t;q) that substantially reduces the contribution of the associated time interval I_{n} to AC#(τ;x;qx) (x=E, L) and/or to ΔAC#(τ;q) can be used in place of the weighting function w1 (t;q)=0. Representative autocorrelation difference functions, ΔAC(τ;q) and ΔAC#(τ;q), which would be obtained before and after application of this process, are shown in FIGS. 17A and 17B, respectively.
A formally equivalent approach for identifying the bit value transitions of interest in Table 1 uses the bit value change function Δb_{n} defined in Eq. (9), for the incoming digital signal. The weighting function used for computation of the autocorrelation function AC#(τ;x;qx) is selected to be w1(t;q) if
B.sub.n =(Δb.sub.n1)*·Δb.sub.n =1, (12A)
and is selected to be w1 (t) if
B.sub.n =(Δb.sub.n1)*·Δb.sub.n =0, (12B)
where b* is the Boolean complement for the binary value b (=0 or 1).
The weighting function w1(t;g1;t1'"'"',t2'"'"') shown in FIG. 5A can be replaced by any other suitable nonuniform weighting function W1(t;q), such as w2(t;q), w3(t;q), w4(t;q), w5(t;q), w6(t;q), w7(t;q), w8(t;q), w9(t;q), w10(t;q), and w11(t;q) shown in FIGS. 6, 7, 8, 9, 10, 11/12, 13, 14, 15 and 16, respectively. The weighting function w1 (t;q) can be replaced by any other weighting function W1 (t;q) that deemphasizes the corresponding contribution to the autocorrelation function.
FIG. 20 is a flow chart illustrating suitable procedural steps that can be taken for analysis of an incoming digital signal according to one embodiment of the invention. In step 120, the incoming digital composite signal s(t) is received. In step 121, a replica of the incoming digital direct signal S_{d} (t) is generated. In step 122, three consecutive signal bits, with signal bit values b_{n2}, b_{n1} and b_{n}, are received for the digital direct signal S_{d} (t). In step 123, the value of the Boolean function B_{n} =(Δb_{n1})*·Δb_{n} is determined. In step 125, the system determines whether B_{n} =1. If B_{n} =1, a first nonuniform weighting function W1(t;q) is used to form a contribution to the autocorrelation function AC1#(τ;x;q) (x=E, P or L) and to the autocorrelation difference function ΔAC1#(τ;q), in step 127, by integration or summation of the quantities W1(t+τt_{E} ;q)s(t)S_{d} (t+τt_{E}) and W1(t+τt_{L} ;q)s(t)S_{d} (t+τt_{L}) over the time interval I_{n}. If B_{n} =0, a second weighting function W1 (t;q) (uniform or nonuniform) is used to form a contribution to the autocorrelation function AC1#(τ;x;q) (x=E, P or L) and the autocorrelation difference function ΔAC1#(τ;q), in step 129, by integration or summation of the quantities W1 (t+τt_{E} ;q)s(t)S_{d} (t+τt_{E}) and W1 (t+τt_{E} ;q)s(t)S_{d} (t+τt_{L}) over the time interval I_{n}.
As a third embodiment, an alternative or adjunct to dynamically changing the weighting function w(t;q) according to the relative bit values b_{k2}, b_{k1} and b_{k} (or b_{k1} and b_{k}) of the incoming digital signal, the periodicity M_{k} of the weighting function W1(t;q) can be changed dynamically. This alternative is implemented as follows. For any particular time t≈t_{k}, the periodicity of the weighting function W1(t;q) is defined by
W1(t;q)=W1(t+n M.sub.k Δτ.sub.chip ;q) (n=0,±1,±2, . . . ),(13)
where M_{k} is a positive or negative integer that may vary with the time interval (t≈t_{k}) within which the time variable t is found. Where Δb_{k} =1 (first embodiment) or B_{k} =1 (second embodiment), the choice M_{k} =1 is appropriate in Eq. (13); when Δb_{k} =0 (first embodiment) or B_{k} =0 (second embodiment), the choice M_{k} >1 is appropriate. In both situations, the same weighting function, such as w1(t;q) shown in FIG. 5A, is used, but with two or more different periodicities.
Implementation
FIG. 21 illustrates apparatus for implementation of a first embodiment, wherein presence of a bit value transition (b_{n1} ≠b_{n}) determines whether or not a time interval I_{n}, of length Δτ_{chip} and including the bit value transition time t=t_{n}, is included in the computation for the autocorrelation function AC1#(τ;x;qx). A periodic clock pulse signal is provided on a first signal line 161, and a code signal, Code(t;τ)=S_{d} (t+τ), is provided on a second signal line 163. Optionally, the clock pulse signal can be passed through a dividebyN module 162 that issues a clock pulse after the module 162 has received N consecutive clock pulses on the first signal line 161.
The clock signal is received at the clock input terminal, denoted CLK, of each of a plurality of time delay or "D" flipflops 165, 167, 169, 171, 173 and 175, with the code signal Code(t;τ) being received at the D input terminal of the first flipflop 165. An output signal issues from the Q terminal of the D fliptlops 165, 167 and 169 and is received at the D input terminals of the D flipflops 167, 169 and 171, respectively. The output signals from the Q output terminals of the D flipflops 167, 169 and 171 are denoted "E", "P" and "L", respectively. The code signal and the output signal from the Q output terminal of the flip flop 171 are received at two input terminals of an EXclusive OR ("XOR") gate 177. The output signal Code(t;τ)⊕L(t) from the XOR gate 177 is received at the D input terminal of the flipflop 173 and also serves as a gating signal, denoted "EG." The output signal from the Q output terminal of the flip flop 173 is received at the D input terminal of the flipflop 175, and the output signal from the flipflop 175 serves as a second gating signal, denoted "LG." The gating signals EG(t) and LG(t) serve as enable signals for the early and late correlation signals, respectively. That is, if the enable signal is low or "false," the correlator does not accumulate the present contribution to the autocorrelation function AC1#(τ;q); and if the enable signal is high or "true," the correlator does accumulate the present contribution to the autocorrelation function AC1#(τ;q). The various output signals from the flipflops and the XOR gate 177 can be expressed mathematically by the relations
E(t)=Code(t2Δt;τ), (14B)
L(t)=Code(t4Δt;τ), (14C)
EG(t)=XOR(t)=Code(t;τ)⊕L(t), (14D)
LG(t)=XOR(t2Δt), (14E)
where Code(t;τ) is the instantaneous signal that appears on the code line 163 and Δt=Δt(clock) is the minimum period of the clock pulse appearing on the clock line 161. Table 2 sets forth the output signal values of the flipflops and the XOR gate for some sequences of code signals.
TABLE 2______________________________________Output Signals For Twoconsecutivebit SchemeCode(t;τ) E(t) P(t) L(t) XOR(t) = EG(t) LG(t)______________________________________0 0 0 0 0 01 0 0 0 1 01 0 0 0 1 01 1 0 0 1 11 1 1 0 1 11 1 1 1 0 11 1 1 1 0 11 1 1 1 0 0. . .0 1 1 1 1 00 1 1 1 1 00 0 1 1 1 10 0 0 1 1 10 0 0 0 0 10 0 0 0 0 10 0 0 0 0 0. . .1 0 0 0 1 0______________________________________
The notation ". . ." indicates that the present line in Table 2 will be identical to the preceding line, unless and until the code signal Code(t;τ) changes.
Returning to FIG. 21, a digital computer C1 receives the clock pulse signals on the signal line 161 and receives the signals E(t)=S_{d} (t+τt_{E}), L(t) =S_{d} (t+τt_{L}), EG(t) and LG(t). A memory for the computer C1 contains a weighting functions w(t;E;qE) and w(t;L;qL), which may be the same or may differ from each other. The computer C1 provides a sequence of two or more sampling times t=t_{k} (k=1, 2, 3, . . . ) for an incoming signal that has been frequency converted to baseband, s(t). For a given sampling time t=t_{k}, the computer C1 examines the incoming signals EG(t_{k}) and LG(t_{k}). If the value of EG(t_{k}) is high, the computer computes a first signal product w(t_{k} +τ;E;qE)S_{d} (t_{k} +τ)s(t_{k}) and adds this to a first accumulated integral or sum that will become the early autocorrelation function AC(τ;E;qE), when fully accumulated. If the value of EG(t_{k}) is low, the value 0 is added to the first accumulated integral or sum for the sampling time t_{k}. If the value of LG(t_{k}) is high, the computer computes a second signal product w(t_{k} +τ;L;qL)S_{d} (t_{k} +τ)s(t_{k}) and adds this to a second accumulated integral or sum that will become the late autocorrelation function AC(τ;L;qL), when fully accumulated. If the value of LG(t_{k}) is low, the value 0 is added to the second accumulated integral or sum for the sampling time t_{k}. The computer C1 computes the difference ΔAC(τ;qE;qL)=AC(τ;L;qL)AC(τ;E;qE) and determines a value τ=t0 for which ΔAC(t0;qE;qL) changes sign or passes through the value 0. The computer C1 interprets the value t0 as a time at which the desired signal S_{d} (t), relatively free of the presence of any accompanying multipath signals, arrived at the signal antenna or receiver.
FIGS. 22A, 22B, 22C, 22D and 22E are contemporaneous timing diagrams illustrating how the apparatus shown in FIG. 21 operates where a bit value transition (0→1) occurs at a time t=t_{n} (b_{n1} ≠b_{n}), where a second bit value transition (1→0) occurs at a second time t=t_{n+1} (b_{n} ≠b_{n+1}), and where a bit value transition does not occur at a third time t=t_{n+2} (b_{n+1} =b_{n+2}). In FIG. 22A, a bit value transition in the code signal Code(t;τ) occurs at time t=t_{n}. An early replica signal E(t) and a late replica signal L(t), corresponding to the bit value transition at t=t_{n}, change from 0 (or low) to 1 (or high) at times t≈t_{n} +2Δt and t≈t_{n} +4Δt, respectively, as shown in FIGS. 22B and 22C. The signal EG(t)=XOR(t) shown in FIG. 22D changes from 0 to 1 at a time t≈t_{n} and changes from 1 to 0 at a time t≈t_{n} +4Δt, when the signal L(t) changes from 0 to 1, assuming that the code signal Code(t;τ) has not changed in the time interval t_{n} ≦t<t_{n} +4Δt. The early gating signal EG(t) thus acts as an enable signal for the early signal correlation and is high (EG(t)=1) only during the time interval t_{n} ≦t<t_{n} +4Δt. In a similar manner, the late gating signal LG(t), shown in FIG. 22E, acts as an enable signal for the late signal correlation and is high only during the time interval t_{n} +2Δt≦t<t_{n} +6Δt. For ease of illustration, it is assumed that t_{n+1} t_{n} >4Δt here.
At a second time t=t_{n+1}, the code signal Code(t;τ) changes from 1 to 0, as shown in FIG. 22A. The early replica signal E(t) and the late replica signal L(t) change from 1 to 0 at the times t=t_{n+1} +2Δt and t=t_{n+1} +4Δt, as shown in FIGS. 22B and 22C, respectively. The early gating signal EG(t) and the late gating signal LG(t), shown in FIGS. 22D and 22E, respectively, become high only during the time intervals t_{n+1} <t≦t_{n+1} +4Δt and t_{n+1} +2Δt<t≦t_{n+1} +6Δt, respectively. The early gating signal EG(t) and the late gating signal LG(t) thus serve as enable signals for the early signal correlation and the late signal correlation, whenever the signal Code(t;τ) changes from 0 to 1 or from 1 to 0.
At a third time t=t_{n+2} the code signal Code(t;τ) does not change, as shown in FIG. 22A, over a time interval of length, say, greater than 4Δt. During this last time interval, Code(t;τ), E(t) and L(t) have the same value (0 or 1) so that XOR(t)=Code(t;τ)⊕L(t) is 0 over this time interval, as are the gating signals EG(t) and LG(t). The early and late enable signals remain low (0) in this situation, as should occur where no bit value transition occurs.
FIG. 23 illustrates apparatus for implementation of the second (threebit transition) embodiment, wherein the value of B_{n} =(Δb_{n1})*·Δb_{n} determines whether or not the interval I_{n} is included in the computation for the autocorrelation function AC1#(τ;x;qx). A periodic clock pulse signal is provided on a first signal line 191, and a code signal, Code(t;τ)=S_{d} (t+τ), is provided on a second signal line 193. The clock signal is received at the clock input terminal, denoted CLK, of each of a plurality of time delay or "D" flipflops 195, 197, 199, 201, 203, 205 and 213 and at a JK flipflop 211. The code signal is received at the D input terminal of the first flipflop 195. An output signal issues from the Q terminal of the D flipflops 195, 197 and 199 and is received at the D input terminals of the D fliptlops 197, 199 and 201, respectively. The output signals from the Q output terminals of the D fliptlops 197, 199 and 201 are denoted "E", "P" and "L", respectively. Optionally, a dividebyN module 192 is included for the clock pulse signals.
The code signal and the output signal from the Q output terminal of the flip flop 201 are received at two input terminals of an EXclusive OR ("XOR") gate 207. The output signal Code(t;τ)⊕L(t) from the XOR gate 207 is received at the D input terminal of the flipflop 203. The output signal from the Q output terminal of the flip flop 203 is received at the D input terminal of the flipflop 205. The output signal from the flipflop 203 is also received and inverted at a first input terminal of an AND gate 209. The output signal from the flipflop 205 is received at a second input terminal of the AND gate 209. The output signal from the AND gate 209 is received at the "J" input terminal of a JK flipflop 211, and the "K" input terminal receives a CCOTO signal, which is discussed below. The output signal from the JK flipflop 211 is received at the D input terminal of a D flipflop with enable 213 (referred to here as a D/EN flipflop).
The output signal from the XOR gate 207 and the inverted output signal from the flipflop 213 are received at two input terminals of an AND gate 215, whose output signal is a first gating signal, denoted "EG." The output signal from the flipflop 205 and the inverted output signal from the flipflop 213 are received at two input terminals of an AND gate 217, whose output signal is a second gating signal, denoted "LG." A table of value assignments similar to that shown in Table 2 is easily set down for the apparatus in FIG. 23.
FIGS. 24A24L are contemporaneous timing diagrams illustrating how the apparatus shown in FIG. 23 operates where a bit value transition (0 →1 or 1→0) occurs in the code signal Code(t;τ) at times t=t_{n} (b_{n1} ≠b_{n}), t=t_{n+1} (b_{n} ≠b_{n+1}), t=t_{n+3} (b_{n+2} ≠b_{n+3}) and t=t_{n+4} (b_{n+3} ≠b_{n+4}), and does not occur at times t=t_{n+2} (b_{n+1} =b_{n+2}) and t=t_{n+5} (b_{n+4} =b_{n+5}), as shown in FIG. 24A. For ease of illustration, it is assumed that t_{n+1} t_{n} =Δt_{chip} >6Δt. FIG. 24B illustrates the signal CCOTO(t), which is defined by ##EQU4## The CCOTO signal goes high one clock pulse interval before each chip transition time occurs (at t=t_{r} Δt), for one clock pulse interval, and is low at all other times. This particular choice is used for illustrative purposes in FIGS. 23 and 24B. Other choices of the CCOTO signal can also be used, with appropriate changes in the apparatus shown in FIG. 23. The signals ##EQU5## are shown in FIGS. 24C24L, respectively. The functions F_{J},K (f(t), g(t)) and F_{D},EN (f(t), g(t)) are determined by the output signals for the JK flipflop 211 and for the D/EN flipflop 213, viz. ##EQU6## The timing diagrams shown in FIGS. 24A24L can be verified by analysis similar to that employed for FIGS. 22A22E. The gating signal EG(t) is high, indicating enablement, only for times in the interval t_{n+3} ≦t≦t_{n+3} +4Δt, which includes the only code signal transition point (t=t_{r} with r=n+3) for which B_{r} =(Δb_{r1})*·Δb_{r} =1 in FIG. 24A; that is, B_{n} =B_{n+1} =B_{n+2} =B_{n+4} =B_{n+5} =0 in FIG. 24A. Similarly, the gating signal LG(t) is high, indicating enablement, only for times in the interval t_{n+3} +2Δt≦t≦t_{n+3} +6Δt.
Returning to FIG. 23, a computer C2 receives the clock pulse signals on the signal line 191 and receives the signals E(t)=S_{d} (t+τt_{E}), L(t)=S_{d} (t+τt_{L}), EG(t) and LG(t). The computer C2 provides the weighting functions w(t;E;qE) and w(t;L;qL). forms the first and second signal products w(t_{k} +τ;E;qE)S_{d} (t_{k} +τ)s(t_{k}) and w(t_{k} +τ;L;qL)S_{d} (t_{k} +τ)s(t_{k}), computes the first and second accumulations AC(τ;E;qE) and AC(τ;L;qL), forms the accumulation difference ΔAC(τ;qE;qL), determines a time shift value τ=t0 at which the accumulation difference changes sign, and interprets the value t0 exactly as does the computer C1 in FIG. 21.
FIG. 25 illustrates more general apparatus for implementation of the various embodiments of the invention. An incoming composite signal is downconverted from an intermediate frequency to a baseband signal and converted to an incoming, digital composite signal s(t), using carrier signal mixing in a baseband conversion module 261, in a well known manner. The incoming digital composite signal s(t) is despread, using code phase mixing to form an early signal product s(t)S_{d} (tτt_{E}) in an early signal despread module 263, and using code phase mixing to form a late signal product s(t)S_{d} (t+τt_{L}) in a late signal despread module 265. An early weighting function w(t;E;qE) is combined multiplicatively with the early signal product to form a weighted early signal product w(t+τt_{E} ;E;qE)s(t)S_{d} (t+τt_{E}) in an early weighting module 267, and an early signal correlation function is formed in an early correlator 269. A late weighting function w(t;L;qL) is combined multiplicatively with the late signal product to form a weighted late signal product w(t+τt_{L} ;L;qL)s(t)S_{d} (t+τt_{L}) in a late weighting module 271, and a late signal correlation function is formed in a late correlator 273. The early and late correlation signals are received by a computer 275 that determines the direction the time shift variable τ should move to drive the system toward the zero crossing or tracking point. The time shift adjustment information is fed back to the early and late signal despread modules 263 and 265 and to the early weighting and late weighting modules 267 and 271, for use in forming an adjusted early signal product and an adjusted late signal product, and is fed back to the baseband conversion module 261.
FIG. 26 illustrates another apparatus 280 for implementing the invention, where the choice of weighting function is based upon the values of two consecutive bits, b_{n1} and b_{n}, in the reference signal, using readily available digital signal processor chips. The apparatus 280 includes a first channel 281A, a second channel 281B and a computer 283 and uses weighting functions W1(t;q) such as are shown in FIGS. 5A, 6, 27D and 27E, with W1 (t;q)=0. The apparatus 280 uses early and late weighting functions W(t;E;qE) and W(t;L;qL), which may be the same or may be different, having nonzero values only near the two ends of a time interval width t1≦t <t2, with periodic extensions as discussed above. The weighting functions W(t;E;qE) and W(t;L;qL) may be chosen, for example, to be step functions as shown in FIGS. 27D and 27E, with the same or different step function widths and the same or different step function gains or amplitudes.
Each channel 281z (z=A, B) has a code phase NCO and generator 285z whose output signal is received by the first of a sequence of three D flipflops 287z, 289z and 291z, each with an associated time delay Δt(clock). An incoming signal I(t) is received by a first input terminal of a first signal multiplier 293z. A second input terminal of the signal multiplier 293z receives a carrier signal from a carrier phase NCO 295z that is controlled by signals received from the computer 283. The carrier signal mixes with the incoming signal to produce a baseband signal. The output signal from the first signal multiplier 293z is received by first input terminals of second and third signal multipliers 297z and 299z. Second input terminals of the signal multipliers 297z and 299z receive the output signals from the first and third D flipflops 287z and 291z, respectively. Output signals from the signal multipliers 297z and 299z are received by an early signal correlator module 301z and a late signal correlator module 303z, respectively. Output signals from the early signal correlator module 301z and late signal correlator module 303 are received by the computer 283 to vary the time shift variable τ that is part of the reference signal S_{d} (t+τ) received from the code phase NCO and generator 285z. The code phase NCO 285z, the D flipflops 287z, 289z and 291z, the carrier phase NCO 295z, the early signal correlator module 301z and the late signal correlator module 303z are all driven by a master clock 305. The D flipflop clock signals are optionally passed through a dividebyNz module 307z as shown, where NA and NB are independently selectable positive integers The carrier phase NCOs 295A and 295B are run identically. The clock pulse width is Δt(clock).
In one preferred embodiment, the code phase NCO 285B in the second channel 281B has a selected positive time delay Δt(BA)=τ_{B} τ_{A} >0 relative to the code phase NCO 285A in the first channel 281A, as indicated by a comparison of FIGS. 27A and 27B. In another preferred embodiment, the code phase NCO 285B in the second channel 281B has a selected negative time delay Δt(BA)=τ_{B} τ_{A} <0 (that is, an advancement) relative to the code phase NCO 285A in the first channel 281A, as indicated by a comparison of FIGS. 27A and 27C.
Four time regions are identified from FIGS. 27A, 27B and 27C, corresponding to the four regions for the weighting functions shown in FIGS. 27D and 27E. A first region X, of temporal width Δt(X), spans the separation between the early and late signals for the channel 281A. A second region Z, of temporal width Δt(Z), spans the separation between the early and late signals for the channel 281B. A third region YA, of temporal width Δt(YA), spans the separation between the late signal in the channel 281A and the early signal in the channel 281B. A fourth region YB, of temporal width Δt(YB)=Δt(YA), spans the separation between the early signal in the channel 281B and the late signal in the channel 281A. The temporal widths are
Δt(X)=Δt.sub.LE,A, (18)
Δt(Z)=Δt.sub.LE,B, (19)
Δt(YA)=Δt(YB)=(t.sub.E,B t.sub.L,A). (20)
The weighting functions w12(t;x;qx) (x=E,L) used in the timedelayed version or timeadvanced version illustrated in FIGS. 27A, 27B and 27C are constructed as follows. Define a square wave function by the relations ##EQU7## where t1≦ta<tb≦t2. The weighting functions w12(t;x;qx) (x=E,L) used in the timedelayed embodiment in FIGS. 27A and 27B are defined by
w12(t;E;qE)=g1·SQ(t1;t1+Δt(YA))+g,2·SQ(t2Δt(X);t2), (22)
w12(t;L;qL)=g3·SQ(t1;t1+Δt(Z))+g4·SQ(t2Δt(YB);t2). (23)
The autocorrelation difference function ΔAC(τ;qE;qL) is then computed, for τ=τ_{A} or τ_{B}, as the integral, or sum of sampled values, of the difference
Δs.sub.EL (t)=w12(t+τt.sub.E ;E;qE)S.sub.d (t+τt.sub.E)s(t) w12(t+τt.sub.L ;L;qL)S.sub.d (t+τt.sub.L)s(t), (24)
over a suitable time interval 0≦t≦T that contains the time interval t1≦t≦t2, with t2t1=Δτ_{chip}.
The square wave functions SQ(t;ta;tb) used in Eqs. (22) and (23) are obtained from uniform weighting correlations, such as shown in FIGS. 27A and 27B, by the following linear combinations of correlation values.
SQ(τ;t1;t1+Δt(YA))={AC0(τ.sub.A ;L)AC0(τ.sub.B ;E)}/2,(25)
SQ(τ;t2;t2Δt(X))={AC0(τ.sub.A ;L)AC0(τ.sub.A ;E)}/2,(26)
SQ(τ;t1;t1+Δt(Z))={AC0(τ.sub.B ;E)AC0(τ.sub.B ;L)}/2,(27)
SQ(τ;t2;t2Δt(YB))={AC0(τ.sub.A ;L)AC0(τ.sub.B ;E)}/2(28)
Proceeding in a similar manner, the weighting functions w12(t;x;qx) (x =E,L) used in the timeadvanced embodiment in FIGS. 27A and 27C are computed as set forth in Eqs. (22) and (23) for the timedelayed embodiment. The square wave functions SQ(t;ta;tb) are obtained for the timeadvanced embodiment from the following linear combinations of correlation values.
SQ(τ;t1;t1+Δt(YA))={AC0(τ.sub.A ;L)+AC0(τ.sub.B ;E)}/2,(29)
SQ(τ;t2;t2Δt(X))={AC0(τ.sub.A ;L)AC0(τ.sub.A ;E)}/2,(30)
SQ(τ;t1;t1+Δt(Z))={AC0(τ.sub.B ;E)AC0(τ.sub.B ;L)}/2,(31)
SQ(τ;t2;t2Δt(YB))={AC0(τ.sub.A ;L)+AC0(τ.sub.B ;E)}/2.(32)
The digital reference function S_{d} (t+T_{y} t_{x}) (x=E,L; y=A,B) used to compute the autocorrelation functions AC0(τ_{y} ;x) has a fixed offset t_{x} and a variable offset τ_{y}, where the difference Δt(BA)=τ_{B} τ_{A} is selected initially. For the timedelayed embodiment, or the timeadvanced embodiment, the offset τ_{A} and/or the offset τ_{B}, subject to the constraint Δt(BA)=τ_{B} τ_{A} =constant, is varied to determine the value of τ_{A} (or τ_{B}) for which the difference Δs_{EL} (t), integrated over the time interval 0≦t≦T, is zero. The value of τ_{A} (or τ_{B}) thus found becomes the tracking point for the incoming signal.
In the apparatus shown in FIG. 26, the dynamic, or time varying selection of W1(t;q) and of W1 (t;q) (=0 here) is performed automatically. If two consecutive digital signal bit values satisfy b_{n1} =b_{n} (or Δb_{n} =0), Eqs. (25)(28) and Eqs. (29)(32) become identically zero for the time interval I_{n}, which effectively insures that W1 (t;q)=0. If b_{n1} ≠b_{n} (or Δb_{n} =1), Eqs. (25)(28) and Eqs. (29)(32) are not identically zero and do provide a nonzero weighting function W1(t;q).
FIG. 28 illustrates a single channel apparatus 310 that provide similar weighting functions w12(t;x;qx) by combining uniformly weighted autocorrelation function values. This apparatus includes a code phase NCO and generator 311 whose output signal is received by the first of a sequence of D flipflops 3131, . . . , 313N1, . . . , 313N2, . . . , 313N3, with associated time delay Δt(clock), where the numbers N1, N2 and N3 are chosen to satisfy
(N11) Δt(clock)=t.sub.L,A t.sub.E,A, (33)
(N21) Δt(clock)=t.sub.E,B t.sub.E,A, (34)
(N31) Δt(clock)=t.sub.L,B t.sub.E,A, (35)
with N1>1 and N3>N2, where the subscripts A and B refer to the desired different desired times for an equivalent "A" channel and an equivalent "B" channel. Here, N2 may be less than, equal to or greater than N1. An incoming signal I(t) is received by a first input terminal of a first signal multiplier 315. A second input terminal of the first signal multiplier 315 receives a carrier signal from a carrier phase NCO 317 that is controlled by signals received from a computer 319. The carrier signal mixes with the incoming signal to produce a baseband signal. The output signal from the first signal multiplier 315 is received by first input terminals of second, third, fourth and fifth signal multipliers 321, 323, 325 and 327. Second input terminals of the signal multipliers 321, 323, 325 and 327 receive the output signals from the D flipflops 3131, 313N1, 313N2 and 313N3, respectively. Output signals from the signal multipliers 321, 323, 325 and 327 are received by a first early signal correlator module 329, a first late signal correlator module 331, a second early signal correlator module 333 and a second late signal correlator module 335, respectively. Output signals from the four early signal and late signal correlator modules 329, 331, 333 and 335 are received by the computer 319 to vary the time shift variable τ that is part of the reference signal S_{d} (t+τ) that is produced by the code phase NCO and generator 311. The code phase NCO 311, the D tlipflops 331y (y=1, . . . , N1, . . . , N2, . . . , N3), the carrier phase NCO 317, and the early signal and late signal correlator modules 329, 331, 333 and 335 are all driven by a master clock 337, with the D flipflop clock signals being optionally passed through a dividebyN module 339 as shown. The clock pulse width is Δt(clock).
Computation of the selected reference signals S_{d} (t+τ_{y} t_{x}) (x=E,L; y=A,B), of the autocorrelation functions AC0(τ_{y} ;x), and of the tracking point for τ_{A} (or τ_{B}) proceeds as in the earlier development for FIG. 26.
The approach developed here is applicable to analysis of any incoming digital signal that has been transmitted using a code division multiple access (CDMA) format, in which the underlying incoming digital signal is a known pseudorandom code and the time of arrival of the direct signal (with multipath signals absent) is unknown. The source of the incoming digital signal may be satellite based or be ground based, with fixed location or variable location relative to the observer. By variably suppressing the contribution of a particular time interval I_{n} to the autocorrelation difference function ΔAC#(τ,q), depending upon the transitions in bit values at t=t_{n}, at t =t_{n1}, and/or at t=t_{n2}, the contribution of noise and/or multipath signals to a modified autocorrelation difference function ΔAC#(τ;q) can be suppressed.
The phrase "autocorrelation function" is used herein to refer to a sum or integral over a collection of time values of a digital signal product w(t+τ;q)s(t)S_{d} (t+τ), where the digital composite signal s(t) includes the digital direct signal S_{d} (t) and, possibly, one or more multipath signals S_{m} (t;Δt_{m}).