US005945948A
`5,945,948
`(114) Patent Number:
`United States Patent 55
`Bufordet al.
`[45] Date of Patent:
`Aug. 31, 1999
`
`
`[75]
`
`[54] METHOD AND APPARATUS FOR LOCATION
`FINDING IN A COMMUNICATION SYSTEM
`Inventors: Kevin A. Buford, Franklin, Mass.;
`John D. Reed, Arlington, Walter J.
`Rozanski, Jr., Hurst, both of Tex.;
`Amitava Ghosh, Vernon Hills, IIL.
`
`5/1996 Roy, Il et al.
`5,515,378
`.
`5,583,517 12/1996 Yokev et ab.
`oe. eeeseseeeenees 342/457
`OTHER PUBLICATIONS
`Joseph Kennedy and Mark C.Sullivan, “Direction Finding
`and ‘Smart Antennas’ Using Software Radio Architectures”,
`May 1995, pp. 62-68.
`
`[73] Assignee: Motorola, inc., Schaumburg,II.
`
`
`
`
`
`Primary Examiner—Theodore M. Blum
`Attorney, Agent, or Firm—Jeffrey G. Toler; Bruce Terry;
`Mario J. Donato,Jr.
`[21] Appl. No.: 08/706,751
`ABSTRACT
`[57]
`[22]
`Filed:
`Sep. 3, 1996
`A method for determining a subscriber unit location in a
`[SU] Mint, C19 eccsssssssessssneestneesenstsenetsesee G01s 3/02
`communication system is provided. The method includes the
`[52] U.S. Cl
`342/457
`
`steps of receiving a signal from the subscriber unit at a first
`[58] Field °f Search
`_342/457. 357:
`base station, determining a first receive time of the signal
`es3457200: 455fh429
`based on a sequence of spreading symbols at the first base
`,
`[56] station, determiningafirst angle of arrival of the signalat theReferences Cited
`first base station, and determining the location of the sub-
`scriber unit from the first receive time, the first angle of
`arrival, and further predetermined information aboutthe first
`base station. The signal is formed via modulation by the
`Sequenceofspreading symbols.
`
`U.S. PATENT DOCUMENTS
`ean tiy1903 ore iran 342/378
`5,317,323
`5/1994 Kennedy et al. esssnssnenene 342/457
`
`4/1996 Schilling .sssesseseseseseenen 375/205
`5,506,864
`5,508,708
`4/1996 Ghoshet al. wees 342/457
`
`51 Claims, 14 Drawing Sheets
`
`1
`
`APPLE 1020
`
`APPLE 1020
`
`1
`
`

`

`U.S. Patent
`
`Aug. 31, 1999
`
`Sheet 1 of 14
`
`5,945,948
`
`
`
`2
`
`

`

`U.S. Patent
`
`Aug. 31, 1999
`
`Sheet 2 of 14
`
`5,945,948
`
`201
`~~ bo LLL,
`210
`CDMA_RECEIVER
`
`
`
`
`
`
`DEMODULATOR
`
`VOCODER
`
`
`
`INPUT
`SIGNAL
`
`
`
`
`
` LOCATION
`SEARCHER/
`OFFSET
`TO
`
`DETECTOR
`DISPLAY
`
`
`
`LOW PASS
`FILTER
`
`202
`
`
`
`TIME
`
` CDMA
`j
`MODULATOR/
`
`SPREADER
`RESP
`|
`790
`
`bee ee eee eee eel a
`200
`
`DATA
`
`3
`
`

`

`U.S. Patent
`
`Aug. 31, 1999
`
`Sheet 3 of 14
`
`5,945,948
`
`PNB1(0)
`
`PNB2(256) PNB1(4)—PNB2(11)
`
`Tg
`
`Ty
`Ty
`be—1/8 CHIP
`sl
`FIG.3
`
` FIG.4
`
`TS(0
`PNB1(1088)
`|
`
`°
`
`Ars
`
`15(28 7/16) TS(29 7/16)
`PNS(100)
`PNS(100)
`| Ats
`
`TS(78 7/16)
`PNS(150
`
`Ats
`
`AtBit<>e—APi—>| AP2—»lArB1=ArB2_sid|| | ArBi
`
`
`
`
`
`TR(O)—TR(24 1/16) Ap3———+! TR(74 1/16)
`
`PNB1(1088)—PNS(100) PNS(150)
`
`4
`
`

`

`U.S. Patent
`
`Aug. 31, 1999
`
`Sheet 4 of 14
`
`5,945,948
`
`10
`VOCODER
`
`| | | | |
`
`|!”
`
`305
`
`DEMODULATOR
`
`
`
`Tag 7
`BS1310
`
`| |
`
`10/FROM
`RAKE
`ANTEWNAS
`Me
`CDMA
`ANTENNA =
`"aoRFF—]| a !
`
`FRONT END
`Noo 2 UN
`||
`=
`||
`cles PR”
`Te
`|
`No. 3
`No. 4
`;
`[SEARCHER(2)
`|
`|
`
`||PN/TIME || |
`
`PROCESSOR
`|
`DETECTOR
`|
`
`355
`l
`
`
`[CDMA MODULATOR/
`|
`|
`
`||SPREADER ENCODER |
`
`
`|
`|
`
`
`
`
`)OUTPUT SYSTEM(GPS)IL353FETE TIME |
`ADJUST
`CLOCK
`bee eee ee eee ee a ee eee eee eee I
`
`
`|yioenLOCATION
`
`
`
`et
`
`OTHER BS’s
`
`360
`
` DATABASE
`
`5
`
`

`

`U.S. Patent
`
`Aug. 31, 1999
`
`Sheet 5 of 14
`
`5,945,948
`
`INITIATE LOCATION PROCESSING
`
`F405
`
`STATUS OF SUBSCRIBER
`
`#10
`
`FIG.7
`
`CHECKED BY BASE
`415
`IN 3-WAY SOFT HANDOFF
`
`I$
`SUBSCRIBER
`
`Y8
`
`400
`
`
`
`
`
`ARE THERE 3
`BASES IN THE CANDIDATE SET
`
`?
`
`
`
`
`
`
`
`
`
`
`?
`
`
`
`
`CHOOSE THREE LARGEST PILOTS
`FROM UNIQUE BASES AND ASSIGN
`TO ACTIVE SET
`
`440
`
`445
`
`COLLECT AND PROCESS
`MEASUREMENTS
`
`RETURN TO NOMINAL
`SETTINGS, DONE
`
`460
`
`OUTPUT LOCATION COORDINATES
`AND CONFIDENCE LEVEL OF
`MEASUREMENT
`
`IS
`SUBSCRIBER IN A 3-WAY
`SOFT HANDOFF
`?
`
`
`CHOOSE THREE LARGEST PILOTS
`
`FROM AT LEAST TWO UNIQUE
`BASES AND ASSIGN TO
`ACTIVE SET
`
`
`NO
`
`465
`
`6
`
`

`

`U.S. Patent
`
`Aug. 31, 1999
`
`Sheet 6 of 14
`
`5,945,948
`
`INITIATE LOCATION PROCESSINGJ-505
`
`STATUS OF SUBSCRIBER
`CHECKED_BY_BASE
`
`sia
`
`YES
`
`515
`
`IS
`SUBSCRIBER
`IN 3-WAY SOFT HANDOFF
`?
`
` ARE THERE 3
`
`BASES IN THE CANDIDATE SET
`
`FIG.&A
`
`|--940
`
`
`CHOOSE THREE LARGEST PILOTS
`FROM UNIQUE BASES AND ASSIGN
`T0_ ACTIVE SET
`
`
`
`
`IS
`Tapp THRESHOLD AT MINIMUM
`?
`
`200
`
`? —
`
`
`
`
`
`ADJUST Tapp THRESHOLD
`
`DOWN T0 OBTAIN 3 BASES
`
`IN CANDIDATE SET
`
`
`CAN EACH
`BASE RECEIVE THE SUBSCRIBER
`?
`
`NO
`
`546
`
`NO
`
`ARE
`AUXILIARY BASES AVAILABLE
`2
`
`
`
`
`CONTINUE WITH TWO
`BASE MEASUREMENTS
`
`547
`
`
`YES
`
`ACTIVATE AUXILIARY BASE
`RECEIVERS
`
` IS AN
`
`EMERGENCY INDICATED
`?
`
`
`
`YES
`
`7
`
`

`

`U.S. Patent
`
`Aug. 31, 1999
`
`Sheet 7 of 14
`
`5,945,948
`
`200
`
`565
`
`
` I$
`
`SUBSCRIBER POWER AT
`MAXIMUM
`?
`
`NO
`
`570
`
`YES
`
`INSTRUCT SUBSCRIBER
`TO_INCREASE POWER
`
`
`
`CAN
`
`EACH BASE RECEIVE
`
`THE SUBSCRIBER
`?
`
`
`
`
`
`
`NO
`
`REDUCE CELL LOADING BY
`PRE-DEFINED STEP SIZE
`
`HAS
`LOAD SHEDDING LIMIT BEEN
`REACHED
`
`
`
`
`
`
`SEND LOC_S SIGNAL IF IN
`ACTIVE MODE; COLLECT AND
`PROCESS AVAILABLE DATA
`MEASUREMENTS
`
`
`
`
`RETURN TO NOWINAL
`SETTINGS, DONE
`
`OUTPUT LOCATION CO-ORDINATES
`AND CONFIDENCE LEVEL OF
`MEASUREMENTS
`
`FIG.&B
`
`550
`
`999
`
`590
`
`8
`
`

`

`U.S. Patent
`
`Aug. 31, 1999
`
`Sheet 8 of 14
`
`5,945,948
`
`
`
`9
`
`

`

`Sheet 9 of 14
`
`5,945,948
`
`U.S. Patent
`
`Aug. 31, 1999
`
`FIG. 12
`
`10
`
`

`

`U.S. Patent
`
`Aug. 31, 1999
`
`Sheet 10 of 14
`
`5,945,948
`
`1420
`
`TIME
`
`1530
`
`1520
`
`1540
`
`1400
`
`1410
`
`1500
`
`1510
`
`TIME
`
`LIG. 15
`
`11
`
`ud
`
`a——e
`
`I
`—_!
`S
`<<
`
`La
`a
`F—_/
`
`Qa
`
`=<
`
`<
`
`11
`
`

`

`U.S. Patent
`
`Aug. 31, 1999
`
`Sheet 11 of 14
`
`5,945,948
`
`1660
`
`162]
`
`LITIG. 16
`
`12
`
`12
`
`

`

`U.S. Patent
`
`Aug.31, 1999
`
`Sheet 12 of 14
`
`5,945,948
`
`YIGOOOA
`
`OL
`
`r-ocore
`
`S%
`
`L
`
`eee we ee ae ee wee eee 4
`
`VNNILNV
`ose|I!
` YOSSI00UdYO193130
`
`
`/YOLVINGOWYNG9
`
`GOE~=SWNNILNV
`ada1LN@|SLNSN313-W {|||||!|{Sitbec|SWV38-N
`
`
`SNVI8—-NYdILNG|SLNAW31S-W
`
`eeaeeeeeeead
` SWVadaNCATHORIVNJdaXI4J90LI
`
`[y°4OLOALFQ
`
`ZOLTAVYYVJNO
`e__JTONY.
`
`d3qdOONI
`
`YI0VIudS
`
`||
`
`(Sd)WALSAS
`
`SWILANTS
`
`LsArdv
`
`INdLAo
`
`l
`
`OOLT
`
`YOLOATAS
`
`JuNV
`
`
`
`QN3LNOY4
`
`
`
`(Z)Y3HOYNV3S
`
`AWIL/Nd
`
`Ove
`
`Wous/0L
`
`I|
`
`XIULVN
`
`||||
`
`JOVAwad
`
`Y4MOLNO
`
`13
`
`13
`
`
`
`
`
`
`
`

`

`U.S. Patent
`
`Aug.31, 1999
`
`Sheet 13 of 14
`
`5,945,948
`
`
`
`
`LNOU|~~BIC|_|yoxwingonad<-v081yo19313S
`
`roe|YIGOONT"|7yQyINGOWWNGO|,|¥oSS300ud 49070Lsnray
`40103130IIWIL/Ndl]lI9YOL9aS
`!LndLno||waqvaudS|
`
`
`
`
`eerFrerTereeeOE——_—_—_—OO—O_eES———_—_—_——_—_—_—__——_—oe7_—_—_—_—_—_—_—_—_—_
`,ON3,
`
`
`
`(Sd)WALSASIWIL3NT4
`
`
`
`
`ZOSTGOE—SYNNALNY~Wous/OLVYOLOAS
`
`!“~~908I
`
`WNNGINY
`
`@4O103S
`
`OO8T
`
`14
`
`14
`
`
`

`

`U.S. Patent
`
`Aug.31, 1999
`
`Sheet 14 of 14
`
`
`
`|
`
`5,945,948
`
`vacoo0n!fzON]SROMLINSININSTS87coerNomOLHoLy
`
`INGOH3datONINSINING138l
`
`roe|BBQ00N3wna}|"syivyt™L/uouwinGon
`
`\Q+geeyGGEjOm
`
`
`YOSS300ud|098fee>ONaeYAHOUWASWW)|TOYINOD[aver|EONvy|Nous
`
`LNOY4]9067:r_-____O46FI0éJuNVyaNyo4|AWv3d‘|2/61SH3LLT1dSIGNa
`
`
`
`
`aFALLAVOVSINTeo}ZY08|SVNNJINV4.v06lhaeave
`W3LSASANIL3NI4i!LndLno|loger
`
`IuadvaudsWaaTg03y
`po—Pta1I49070
`isnrayI\(Sd)
` yoss300ud|P|tneSWIL/Ndolyindonad
` 4cL6l”“S5WILWW3dYOIOMLNODONYyOVadaa4
`
`
`61OLlAol|O06Tager961
`
`SOEWOud/0l“E06T
`
`Oleeg;ZO6I
`.ave=H)
`
`waATIO3e
`
`NMOG®GNFNOUS4u
`O¥I40G/V+YSLYAANOD
`
`STYNOIS
`
`E061
`
`
`
`15
`
`15
`
`
`
`
`
`

`

`5,945,948
`
`1
`METHOD AND APPARATUS FOR LOCATION
`FINDING IN A COMMUNICATION SYSTEM
`
`FIELD OF THE INVENTION
`
`to wireless
`in general,
`invention relates,
`The present
`communication systems and, moreparticularly, to a method
`and apparatus for locating a subscriber unit
`in a Code
`Division Multiple Access (CDMA)wireless communication
`system.
`
`10
`
`Background of the Invention
`
`2
`mentis determined by receiving a signal from the subscriber
`unit at a first base station, determining a first receive time of
`the signal based on the sequenceof spreading symbols at the
`first base station, and determininga first angle of arrival of
`the signal at the first base station. The signal is formed via
`modulation by the sequence of spreading symbols.
`In accordance with another aspect, a communication
`system includes a controller and a location processorthatis
`responsive to the controller. The controller is responsive to
`a first and a second basestation, each ofthe first and second
`base stations comprising a receiver operable for receiving a
`signal from the communication unit, the signal being formed
`via modulation by a sequence of spreading symbols, and a
`detector operable for determining a receive time of the
`signal based on the sequence. The location processor is
`responsive to the controller and is operable for requesting
`the first and second base stations to determine first and
`
`second receive times of the signal based on the sequence,
`and for determining a location of the communication unit
`from the first and second receive times and further infor-
`mation about the first and second basestations.
`
`second signals are formed based on the first sequence of
`symbols and the second sequence of symbols, respectively.
`BRIEF DESCRIPTION OF THE DRAWINGS
`
`In a wireless communication system it is often desirable
`to locate users who are making calls. Applications for such
`a technology would include 911-emergencyservices, so that
`police/fire/ambulance services could be dispatched to a user
`making a call. Other applications would include fraud
`detection, police investigations, and the like.
`Previously installed cellular systems had little capability
`in this regard. For example, in AMPS (Advanced Mobile
`Phone System) Cellular Radio, a user could be located
`within a cell by determining which base station antenna was
`In accordance with yet another aspect, a method for
`used to serve the user. However a cell could be as large as
`determining location of a subscriber unit includes receiving
`3-5 miles in radius, making this information practically
`a first signal from a first base station of the plural base
`useless. Since many of the dense urban cell sites are now
`stations and a second signal from a secondbasestation of the
`much smaller, and many of the urban/suburbancell sites are
`plural basestations, determiningafirst receive time based on
`nowsectored, using sectored antennas to limit a channel’s
`a first sequence and a secondreceive time based on a second
`service area to just one sector of a cell, the coverage areas
`sequence, and determining the location of the subscriber unit
`from the first and second receive times and further infor-
`of a cell are now smaller. However, the area even in these
`mation about the first and secondbasestations. The first and
`smaller cells canstill be more than one square mile. Thisstill
`makes locating a user impractical for most purposes. Other
`radio systems, like US Digital Cellular (USDC) and Group
`Speciale Mobile (GSM) use the same methodofidentifying
`the cell or sector, and thus could do no better than the AMPS
`system.
`While there are other location alternatives, such as the use
`of Global Positioning System (GPS) units at the subscriber
`unit, or triangulation onto a transmitting subscriber unit,
`these and similar approaches are too costly to be used by
`most subscribers, or in the case of triangulation, require
`other costly and time-consuming resources to be dedicated.
`There remains therefore a need for an improved, cost-
`efficient approach for locating subscribers in a wireless
`communication system.
`
`15
`
`20
`
`25
`
`30
`
`35
`
`40
`
`45
`
`The inventionitself, together with its intended advantages
`may best be understood by reference to the following
`detailed description, taken in conjunction with the accom-
`panying drawings, of which:
`FIG. 1 is a simplified diagram illustrating a cellular
`system which may employ the present invention;
`FIG. 2 is a block diagram of a CDMAreceiver at a
`subscriber unit according to a first embodiment of the
`invention;
`FIG. 3 is a diagram illustrating location finding of a
`CDMAsubscriber unit according to an embodimentof the
`invention;
`These problems and others are solved by an improved
`FIG. 4 is a diagram illustrating a timing sequence used in
`method and apparatus according to the invention. In accor-
`determining propagation delay for location of a CDMA
`dance with a first aspect, a method for determining a
`subscriber unit according to an embodiment of the inven-
`subscriber unit location in a communication system includes
`tion;
`the steps of receiving a signal from the subscriber unit at a
`FIG. 5 is a block diagram of a CDMAreceiver at a base
`first base station, determining a first receive time of the
`station according to an embodiment of the invention;
`signal based on a sequence of spreading symbolsat thefirst
`FIG. 6 is a timeline diagram illustrating propagation and
`55
`base station, determiningafirst angle of arrival of the signal
`delay times used in calculating a subscriber according to an
`at the first base station, and determining the location of the
`embodiment of the invention;
`subscriber umit from the first receive time, the first angle of
`FIG. 7 is a flow chartillustrating the process by which a
`arrival, and further predetermined information aboutthefirst
`subscriber measures base station signals according to an
`and second base stations. The signal is formed via modula-
`embodiment of the invention;
`tion by the sequence of spreading symbols.
`FIG. 8 is a flow chartillustrating the process by which a
`In accordance with a further aspect, a method for esti-
`base station measures subscriber signals according to an
`mating a subscriber unit
`location includes the steps of
`embodiment of the invention;
`performinga first location measurement havinga first con-
`FIG. 9-13 are diagramsillustrating location finding of a
`fidence level, performing a second location measurement
`subscriber unit in accordance with a second embodiment.
`having a second confidence level, and determining an esti-
`mated location of the subscriber unit based on the first and
`second location measurements. The first location measure-
`
`SUMMARYOF THE INVENTION
`
`50
`
`60
`
`65
`
`16
`
`FIG. 14-15 are general diagramsillustrating reception of
`a signal from a subscriber unit by a base station;
`
`16
`
`

`

`5,945,948
`
`3
`FIG. 16 is a diagram illustrating location finding of a
`subscriber unit when there is an obstruction between the
`subscriber unit and a base station.
`
`FIG. 17 is a block diagram ofa first receiver implemen-
`tation in a base station for use in location finding according
`to the second embodiment;
`FIG. 18 is a block diagram of a second receiver imple-
`mentation in a base station for use in location finding
`according to the second embodiment.
`FIG. 19 is a block diagram ofa third receiver implemen-
`tation in a base station for use in location finding according
`to the second embodiment.
`
`DETAILED DESCRIPTION OF THE DRAWINGS
`
`A first embodiment of the invention is a system for
`determining the location of a user in a Code Division
`Multiple Access (CDMA)cellular system. Using the CDMA
`modulation information, an estimate of the time of flight or
`propagation is madeof the first arriving ray at a subscriber
`unit. The first ray received typically represents the shortest
`path between the base and subscriber, and the time of flight
`estimate allows the calculation of the distance between the
`subscriber and the base station. By calculating the distance
`to multiple, e.g., three, sites, a specific subscriber location
`can be calculated limited by the accuracy’s of the measure-
`ment timing and other processing delays.
`In the first embodiment the time of flight of the signal
`between each base and subscriber is calculated automati-
`cally within a correlation receiver. The processing steps
`involve the transmission of a Pseudo Noise (PN) sequence
`coded signal time-aligned to under a chip accuracy (e.g.,
`“oth of a chip), and correlating on this signal at the receiver
`using a correlation algorithm. Because the modulation
`sequence (e.g., a PN sequence) is known and used in
`synchronization/despreading, a precise time of reception of
`a given chip can be determined. By determining reception
`time for multiple related signals, a time delay can be
`calculated and used to determine a position estimate.
`In one implementation, the subscriber uses known PN
`sequence and offset information to determine which related
`PN chips from different bases (standard and/or auxiliary
`bases) that were transmitted at
`the same time, and also
`determinesthe time of reception of these related chips. From
`the difference between the reception times, a time differen-
`tial and thus distance differential is determined. Using the
`distance differentials and known positions of the bases, a
`position estimate is determined. Where a subscriber is only
`in communication with one or two bases, additional bases
`maybe forced into an active set (including auxiliarysites, if
`needed) so that time measurements can be made by the
`subscriber.
`
`In another implementation, receiving base sites are con-
`trolled to make time measurements of selected chips, and the
`difference in receive time is used to similarly calculate the
`subscriber position. Where additional receive sites are
`needed because of interference and the like, auxiliary sites
`are controlled so as to receive the signals transmitted from
`the subscriber unit. If necessary, in case of an emergency, the
`subscriber unit is powered up to a maximum powerlevel
`such that at least three base stations can receive and make a
`
`time estimate of the signal. Further, where more precise
`measurements are needed, a special location message can be
`transmitted to the subscriber. Upon receipt, the subscriber
`determines a chip/time offset for a response signal, encodes
`the offset and transmits the response signal. Upon decoding
`the offset and comparing the receive times of a same chip
`
`10
`
`15
`
`20
`
`25
`
`30
`
`35
`
`40
`
`45
`
`50
`
`55
`
`60
`
`65
`
`4
`(e.g., the first chip of a frame) used in determiningthe offset,
`a delay compensated time value is determined for the
`various propagation paths, and the position determined
`therefrom. Finally, since it might be difficult
`to get a
`received signal at bases further away, an emergency load
`shedding can be performed at the nearby bases to provide
`extra range, since capacity can be traded off for range in a
`CDMAradio system. Thus coverage is improved, and
`location finding is made morereliable.
`Turning now to FIG. 1, a cellular system is generally
`depicted as 100 having a hexagonal cell pattern with base
`stations 110, 120, 130, and a subscriber 140. Auxiliary base
`units 121 are also located between bases 110, 120 and 130.
`The distance between bases 110, 121 and 130 and the
`subscriber umit 140 is estimated by determining the time of
`flight or propagation of the first arriving ray which is
`measured from a predefined reference time to the point in
`time that the receiver performs a correlation on the trans-
`mitted signal. This is made moredifficult, in that the distance
`estimate may be overestimated, or underestimated since the
`measurementis made to an arbitrary time reference point in
`the receiver (a precise measurement would only be available
`if a more accurate (and costly) timing system such as one
`derived from a GPS signal or atomic clock is used in the
`subscriber 140). Thus,
`the distances 150, 160 and 170,
`respectively, may be longer or shorter than the actual dis-
`tance between each base 110, 121, 130 and the subscriber
`140 based on correlation to a chip rate (at an approximately
`814 nanosecond (ns) chip rate (i.e., the rate of the fully
`spread signal, which is determined in TIA
`(Telecommunications Industry Association) Interim Stan-
`dard IS-95A by the PN sequencerate), or approximately 250
`meters (m) per chip; so it
`is desirable to achieve time
`measurements at faster than the chip rate). In FIG. 1, the
`distance 150 is shownto be overestimated indicating a point
`125 beyond the subscriber unit’s actual location. Likewise
`points 115 and 135 are also overestimated. These points will
`be corrected by the distance processing described below,
`yielding an estimate much closer to the subscriber’s true
`location.
`
`FIG. 2 is a block diagram illustrating a CDMAsubscriber
`unit 200 having a CDMA receiver 201, locator unit 202, and
`transmitter 203. The receiver 201 has a common RF(radio
`frequency) front end 205 which feeds three independent rake
`inputs, 210, 220, 230. These rake units 210, 220 and 230 can
`lock onto three different received rays that are approxi-
`mately one PN chip time or more apart, which is typical of
`a direct sequence spread spectrum (DSSS) receiver. The
`searcher 240 scans for new correlation peaks at faster than
`the chip rate (in the preferred case allowing for resolutions
`as fast as the 50 ns clock rate), and can reassign the rake
`inputs based on its best estimate of current channel condi-
`tions. Normally, the correlators for rakes 210, 220 and 230
`lock onto the three strongest rays that are available, and
`when a second or third base station can supply a signal
`sufficiently strong, they are reserved for locking onto these
`other base stations signals which are also delayed in time
`more than one PN chip time respectively, as described by the
`IS-95A Standard. If only two base stations are sufficiently
`strong,
`then two rays are dedicated, one for each base
`station, and the third ray to the strongest remaining ray for
`either base station.
`
`When a location finding function is desired by the sub-
`scriber 200, it is preferable to attempt to find three different
`base stations, one for each ray so that sufficient information
`is available to accurately estimate the location. Thus,
`to
`connect to three base sites the rakes 210, 220 and 230 are
`
`17
`
`17
`
`

`

`5,945,948
`
`10
`
`15
`
`20
`
`25
`
`30
`
`6
`Returning to FIG. 2, DLLs 215, 225, and 235 are fed back
`to each rake 210, 220 and 230, respectively, for adjusting the
`signals to output fine time aligned signals. As noted above,
`the DLL outputs can also serve as fine phase offset infor-
`mation for adjusting the receive times of the PN chips,
`preferably after filtering in Low Pass Filters (LPFs) 250,
`260, 270 for each channel, respectively, which effectively
`averages the outputs of each DLL 215, 225, 235. This
`averaged fine phase offset information, together with the
`chip number/times/base identification or offset (i.e. B1-B3
`information) from searcher 240 (which is also adapted for
`PN chip/time detection), are fed to location searcher 280.
`Location searcher 280 takesthe fine phase offset information
`from each branch and corrects the time of reception from
`searcher 240 for each chip, to give a corrected relative time
`of reception for each branch. From the earliest time, say B1
`(ie.,
`the time the signal from base 1 is received),
`the
`difference tB21 and tB31 in reception time for the other
`signals B2 and B3 is determined, and the corresponding
`distances dB21 and dB31 determined. One thus knowsthat
`the distance from bases 1 (110), 2 (120) and 3 (130) are dB1,
`(dB1+dB21) and (dB1+dB31), respectively. Further, from
`the PN offsets, the identity of the bases are known andtheir
`geographic position can be retrieved from memory 281. It is
`then a simple matter of performing a search routine to
`determine, one suchasillustrated in FIG. 4, to determine the
`geographic coordinates of the mobile. In the example of
`FIG. 4, the known base locations are used to define three
`lines L12 (151), L23 (152) and L13 (153). The distances
`dB21 and dB31 are subtracted from lines L12 (151), L23
`(152) and L23 (152), L13 (153)respectively, and the remain-
`ing segments bisected by normallines N12 (154), N23 (156)
`and N13 (155). The intersection of these lines N12 (154),
`N23 (156) and N13 (155)is the position of the subscriber
`140. This information could then be sentto the serving base
`station for forwarding to a requesting party of serving
`location register, or could be forwarded for use by the
`subscriber(e.g., on a map grid or other location device, not
`shown).
`if base site location information is not
`Alternatively,
`available to the subscriber, the phase offset, chip, timing and
`base offset information can be sent in a location request
`signal to a serving base station. There, a location searcher
`can access its own database and determine the subscriber
`location. This location information is then transmitted back
`in a location response message to the subscriber or other
`requesting entity.
`A preferred approach, however, for location using infra-
`structure equipment can be seen with reference now to FIG.
`5, which generally depicts a block diagram of a CDMA
`infrastructure system 300 having a first CDMAbasestation
`301. Base 301 has a common RFfront end 305 which feeds
`
`5
`adjusted so that at least three base unit signals are decoded.
`If available, emergency pilot generators (such as auxiliary
`base unit 121 of FIG. 1) physically located between the base
`sites could be activated in response to a beacon request in
`order to blanket the area with additional reference signals,
`allowing the subscriber to make location estimates based on
`these pilot generators as well as the standard basesites.
`These auxiliary units would have a different PN offset than
`the surrounding base stations, and would typically be
`equipped with a GPSreceiver for proper synchronization/
`timing. They would be coupled to the base stations or other
`controller in the infrastructure by any convenient means,e.g.
`wireless or twisted pair cable. Their activation is preferably
`accomplished by a request to the controller, or command
`from the serving basestation to a local auxiliary unit under
`its control, upon indication by the subscriber that less than
`three bases are available. Alternatively, the auxiliary units
`could be equipped with scanning receivers that, in response
`to a request signal by a subscriber, would begin transmitting
`for a limited period (e.g., 5 seconds, in order to minimize
`system interference). By appropriate placement, such aux-
`iliary units can be used to reduce uncertainties at certain
`locations or generally increase the accuracy of position
`finding in strategic areas, such as major highways, malls, or
`central business districts. Because of the interference-
`limiting nature of a CDMA system, in some cases only one
`base station will be able to receive the subscriber’s signal,
`and vice-versa, so the auxiliary units are needed to obtain the
`necessary multiple readings.
`Therelative time of reception of each signal is determined
`by using information about the leading edge(or alternately,
`the peaks) of related correlation peaks in the searcher, and
`adjusting this by an offset determined in a fine time align-
`mentcircuit (e.g., delay lock loops (DLLs) 215, 225 or 235
`for each branch, coupled with filters 250-270). Preferably
`related correlation peaks are those received on different
`branches but within one chip of each other. In this approach,
`the precise time of the leading edge is determined, along
`with the PN sequence number(i.e., the chip position (e.g.,
`number 245) of the repeating PN sequence (e.g., approxi-
`mately 16,000 chips in length)). Using the already deter-
`mined PN sequenceoffset, and the system design where the
`base PN sequence is the same for each base station, and
`transmitted at the same system time plus or minus a unique
`PN sequenceoffset, the difference in relative times yields a
`difference in propagation path delay. This is illustrated in
`FIG. 3. At time TO two bases B1 and B2 are transmitting, but
`base B1 transmits PN chip 0, while base B2 transmits PN
`chip 256 since it has a PN sequence offset of 256 chips. At
`some time T1, after location finding is activated, the sub-
`scriber determines that the leading edge of PN chip 4 from
`B1 has been received. The next leading edge of a PN chip
`from base B2 is received sth of a chip later at time T2, and
`the chip is determined to be the 280th in the PN sequence.
`From these receive times and PN numbers, the propagation
`delay difference is calculated to be (PNB2-offset) +(receive
`difference, T2-T1))-(PNB1-offset)=((261-256)+(4))-(4-
`lation peaks, and can reassign the rakes based on its best
`O)=1%% chips*814 ns/chip=916 ns. At approximately %
`estimate of current channel conditions. Normally, the four
`meter (m) per ns propagation speed for a radio signal, this
`correlators of rakes 310, 320, 330 lock onto the four stron-
`translates into about 300 m difference in propagation path
`gest rays that are available.
`distances. The precision in location is only limited by the
`
`system clock rate being used and degree of synchronization. Whenalocation finding function is desired, two general
`approachesare available—either passive (i.e., no subscriber
`Where all base stations are using GPS timing information,
`synchronized transmissions (i.e., of the leading edges of
`unit response) oractive. In either case it is preferable to find
`chips) to within 50 ns (or approximately “6th of the chip
`at least three different base stations capable of receiving a
`rate) are currently possible. With a local clock generating at
`subscriber signal, so that sufficient information is available
`least the same 20 MHz clockrate, location to within 100 ns
`to estimate the location. In a first embodimentpassive mode,
`four rake branches 310, 320 .
`.
`. 330 of base 301 are used to
`or 30 m is possible.
`
`35
`
`40
`
`45
`
`50
`
`55
`
`60
`
`65
`
`18
`
`. 330.
`.
`four independent rake inputs, shown as 310, 320, .
`These rakes can lock onto four different received rays that
`are at least one PN chip time apart, which is typical of a
`DSSSreceiver. The two searchers 340 scan for new corre-
`
`18
`
`

`

`5,945,948
`
`7
`detect an uplink signal. From each rake, a Delay Lock Loop
`(DLL) is used to generate an estimate of the timing (ie.,
`adjustment) of the correlated ray. This more accurately
`estimates the time of the correlation, similar to the process
`used by the subscriber unit above. Searcher and Chip/Time
`Detector 340 peak correlates the signal on each branch, and
`also determines the best branch to use (preferably based on
`the earliest received peak for the same chip, but other
`selection techniques may be used to determine a current best
`branch); this best branch signal is used in determining PN
`chip and receive time information, similar to that in sub-
`scriber searcher 240.
`
`To initiate a location process, in a preferred embodiment
`a commandis initiated within the system 300, most likely at
`a regional entity such as a mobile switching center (MSC)
`365, operations center, or perhaps within a connected net-
`work such as PSTN (public switched telephone network)
`375. A location request is then processed via home location
`register (HLR) 366 to determine the currently serving base
`station(s). Upon receipt of a location command, processor
`350 of base 301 (and similar processors of other serving
`bases) uses detector 340 to determine a chip receive time.
`Preferably this is accomplished byall bases determining the
`leading edge rise time of a specified group of PN chips, for
`example by determining the rise time for each 64th chip (.e.,
`PN sequence number 0, 64, 128, etc.) for a predetermined
`numberofchips, e.g., 10. This informationis then forwarded
`by each basereceiver, along with its ID (identification), to a
`designated entity, e.g. location searcher 361 of BSC (base
`site controller) 360, or location searcher 367 of HLR 366,
`etc.. Thus, the difference in receive time for the same chips,
`each chip being derived from the same single chip
`transmission, may be used to determine propagation delay
`differences. In other words, for each chip number the dif-
`ferential between receive times at the different bases yields
`a propagation difference, and location may be determined
`from this information in conjunction with the known loca-
`tion of the receiving bases, in a similar manneras described
`above with FIG. 4. By taking plural sets of information in a
`relatively short time frame (e.g., 10 times, every 64 chips,
`across about 500 microseconds), and averaging or otherwise
`best-fit calculating using the determined positions, position
`errors can be minimized. A skilled artisan will appreciate
`that other approaches can be used in the actual calculation.
`For example, a detection at the same system time(s) for
`leading edges within one chip of the designated time(s),
`along with time differences from the designated system time
`and chip number, could be used in determining the propa-
`gation delay differences (albeit, an additional error may arise
`because the transmit time for the different chips is limited by
`the accuracy of the subscriber’s clock rate; even if a 50 ns
`clock cycle were present, this is still more error than present
`from a transmission of the same chip (which has no timing
`error). What is important is that the chip ID (e.g., number/
`position in the PN sequence) and precise time of reception
`(e.g., leading edge, or peak, at the oversampled clock rate)
`at different bases be used in determining the subscriber
`location.
`
`In a preferred embodimentfor active location, a two-way
`ranging system is implemented using both chip receive time
`information and certain response information from the sub-
`scriber. In this embodiment, the process is again initiated
`with a l