Spread-spectrum signal demodulator

US006088383A

United States Patent Suzuki et al.

[19] Patent Number:

Date of Patent: 6,088,383

Jul. 11,2000 [111

[451

[54] SPREAD-SPECTRUM SIGNAL 7-273713 10/1995 Japan .

DEMODULATOR OTHER PUBLICATIONS

[75] Inventors: Toshinolji Su_Zuki> Tokyo; Yoshio “Near—Far Resistance of Multiuser Detectors in Asynchro Takeuchh Saltama, both of Japan nous Channels”, R. Lupas et al., IEEE Transactions on

_ _ _ _ _ Communications, vol. 38, No. 4, pp. 496—508, Apr. 1990. [73] Asslgnee' gokllllsalTDlinshgn Denwa Kabushlkl “Multiuser Detection for CDMA Systems”, A. Duel—Hallen

als a’ 0 yo’ apan et al., IEEE Personal Communications, pp. 46—58, Apr.

[21] App1.No.: 08/952,342 1995 _ “A spread—Spectrum Multi—Access System With a Cascade

[22] PCT Flled: Mar‘ 5’ 1997 of Co—Channel Interference Cancellers for Multipath Fading [86] PCT NO‘, PCT/JP97/00679 Channels”, Y. Yoon et al., IEEE Second International Sym

posium on Spread Spectrum Techniques and Applications, § 371 Date: Oct. 15, 1997 Nov. 29—Dec. 2, 1992, pp. 87—90.

§ 102(6) Date? Oct- 15, 1997 Primary Examiner—Stephen Chin _ Assistant Examiner—ShuWany Liu

[87] PCT Pub' NO" “709763401 Attorney, Agent, or Firm—Westman, Champlin & Kelly, PCT Pub. Date: Sep. 12, 1997 PA.

[30] Foreign Application Priority Data [57] ABSTRACT

Mar. 7, 1996 [JP] Japan .................................. .. 8-078171 A Spread Spectrum Signal demodulator receives a received

[51] Int. C17 ................................................... .. H04B 15/00 igrgglriiijuiiglgagtpggzaggnj SS§§§§gIISIP§§g$§S§Ii§;aI1§;if [52] US‘ Cl‘ """""""""""" " 375/148; 542;_337755//1;I76; Which reduces a correlation signal outputted from a ?rst

_ ’ correlation detecting means or replica signals until the Fleld of Search ................................... .. Correlation Signal has been traveled through a replica Signal

375/152’ 148’ 285’ 144’ 147’ 370/342’ generating means, a replica signal subtracting means and a 479 second correlation detecting means. According to the

, invention, it is possible to improve interference cancellation [56] References Clted capacity With a computational effort substantially compa

U_S_ PATENT DOCUMENTS rable to that in a prior art replica signal cancellation system _ and improve interference cancellation performance With

1;/ gatfillllfhl et ‘1}’ """"""""""" " increasing number of interference cancellers irrespective of 5’74O’2O8 41998 HcuItelrrtlge I6 if ' ' 374346 the magnitude of the correlation. With a suf?cient number of 5’812’546 9/1998 Zhou et a1 ' 370/342 interference cancellers provided, it is possible to obtain the

5:898:740 4/1999 Laakso et aliiiii: .................. .3 375/346 Same/Pied as obtainéble with Pri"r a“ decorr‘ilation in _a real time processing irrespective of the correlation magni

FOREIGN PATENT DOCUMENTS mde~

7-212274 8/1995 Japan . 7-264112 10/1995 Japan . 18 Claims, 18 Drawing Sheets

1 9 O I- - - n _ _ _ _ - - ' _ _ _ l - _ _ - _ _ _ _ ' ' _ _ _ _ ' _ ' _I :- _l r 'I

5 1 9 o I

ILI i *I_o_I' : I' i_> _>I l—) I 1 9 1 D : I : I i

‘I‘ ; r121 H131 PIFI'MII I: EcEIvED GGREEEiiION I / REPLICA siIigmiH : i) SECOND I l l l i R l SIGNAL — CORREL 0 I SIGNAL DETECTING I GENERATING SUPRESSING _’ Z " DETEc'ilNGN T‘): r" ‘>: ‘_>

MEANs I MEANS MEANs MEANS I I ' I '

I 74 -r I II I - I ' I I I

‘J2 I mg ,m E11 :;1 I: IORIIITIII I I8 SIIIIIII I ' I ' I . sIGN 00R l . . . I

"’ DETECTING I GENERATING "’ suPREssING '_> oEfgéiiIlIgN JF’I :_> ‘*I 1”’) MEANs I MEANs MEANS I : I : I

- 1 ~ - 19K1L'i2i" I. -' I 0 l a o l | : I o a: I o

' I {12K [13K $1 :'I i‘ ‘I i‘ IOIIIITIIII I i" STIIIIII I I 1 i I s A c0 DETECTING I GENERATING SUPRESSING _> néfgéfiriém 1‘): Il_) +1 IL) MEANs I MEANS MEANS I 5 Kr MEANS I I I I I

l I l I

11K I ' ' ' ' ' ' ' ' ' ' ' ' ' ' ' ' "? ‘ “ ' ' ' ' ‘ ' ' ' ' ‘ "I '7‘ "f

2 I 2 2 2 M

U.S. Patent Jul. 11,2000 Sheet 2 0f18 6,088,383

F l G. 2

322 a / b

—lj—) LATCH

J?‘ E SPREADING f3 2 3

SYMBOL TIMING ------------- "a -------- --> CODE

GENERATOR

3 5 1

SUPPRESSING FACTOR (GREATER THAN 0R EQUAL "0" AND LESS THAN "1")

FIG. 4 d

/

382

381 H 383 C

INTEGRATOR

1,\

SPREADING : CODE r384:

GENERATOR i 4 : ' i

SYMBOL T mum; --------- "Q --------------- -

U.S. Patent Jul. 11,2000

F I G . 5

3 3 1 3 3 2

a / / { HARD

DECISION —-—> LATCH ELEMENT

4*

SYMBRLHIW'JNP ................... ..; ..... -.>

Sheet 3 0f 18 6,088,383

SPREADING CODE

GENERATOR

RECEPTION LEVEL

362

FIG. 7

THRESHOLD LEVEL

b l ;—-> LIMITER

\ 361

FIG. 8

OUTPUT SIGNAL STRENGTH

A

SUPPRESS l NG FACTOR

x

THRESHOLD LEVEL

I

I

|

I

l I

I

I

l

l

I

|

THRESHOLD LEVEL

SUPPRESS | NG FACTOR

> INPUT SIGNAL STRENGTH

U.S. Patent Jul. 11,2000 Sheet 4 0f 18 6,088,383

82m AA|

.55 22581 wqm/ m?émzww 28

minimum M ........... . ...2 ............................ .- 62;: 3%;

hqm " / mvm :1“ m w. W 55m

A! AA . 2920mm

n . 92:

www

N 0. m\

w .0 _ m


d

/

3 9 1

C /

g MATCHED ) FILTER

3 9 3 3 9 4

T 3 9 2 / / TIMING TIMING SYMBOL

. . . . n) .--..

SPREADING CODE DETECTOR GENERATOR > TIMING

401 402

H /

._____, MFAITLCTHEERD —> LATCH _>

T SPREAD l NG CODE SYMBOL TIMING


v L'TU'TH'U'U 5+“ ' ‘G: ' _ _ " T’ " ' " ' ' "

~11

I d _l I I I I I I I I i I I I I I I I l I I I I | I I I I ‘ I I I I l

n _ _

_ _ x NOSE _ v_ NOSE _ “ 62681113 x m N v. N N " wzawwala

_ _ \ I _

_

_ “E5518 AI. “Emma? “ NEENNNS A1

H _ x m N _ x N N J

_ _ a _ K F N

_ _

_ _ J _

. O

_ _ N NOSE " VEN “ _ N NOSE ”

n _ wzzwwansw . N m N N N N _ 258E120, _ _

- - r r _

_ NEEDS:

m o

n _ NEENNNSAJ- INN _ NESNN o 4

J

n u 10'. U N N

_ IL _

_ _ _

_ _ P NOSE NNN 25: _ _ NOSE “ _ NZGNNNNNQ B N 62:05:26 F m N EZGSENDN _ .2266 .

_ _ 53mm; , _

N858? NEENNNS

_ _ 1mm _

_ P N N \

_ u a U _. F N

_ _ P w N\-.||L F _

_ _ _ 10

Q

_ _

"A||_.U/o m N _ u

._ .|||..||||..|n....|||..||| | | | | |....|||||L omN

w _1 w .

_. _. G _ m


FIG. 13

403

I404 SPREADING CODE

GENERATOR

FIG. 14

1 2 > 5 > 5 Q . ................................ .u . I I I

I I I

I I I - m I u

I A I

- Ll - TlLIlll-VWATM . LWMA . . - PIE-LL - - - ESNM 1 - I - R E - - 1|- - G 5 I I

I I I

1 - 4 - I

I I I

I

4 u u .

f . . u

u I u .

I I I N . G W G . n

I

mm - 1 DIIN " -

TTIS " MGSM NMHM " u

m m m m . 4 H w m w A m L c A . n l. E E E u 4 .l R P E E E E E . .

FRTM . STPM SRTM . . RE . SU RE " .

OD I S I I C - C - -

I I I

I x I I

I

" 1| " -

I 1 I

2 _ .

n < 6 n u

. 4 G 4 . .

I

n A N u -

I I

n u .

. D V LMAA . V“ - + PIRE - I

. ESTM . . - R B - .

u w u u

I I I

I I I I I I

r

RECEIVED SIGNAL

V


I

I I

I

I

I

I I l

I I

I

I I

I I

I.

T'IL'L'TT"

I-.'.._..'..-..".-.I-i

A]. . 1|“ m _ m m _ m u _ n n _ u u _ n n _ u n _ n u _ n n _ u u _ n HAIL" mwjmuzé II“ m m wozmm?m?i m m

.1]. TI“


FIG. 19 PRIOR ART 1 001 1 O 1

RECEIVED / ............... SIGNAL : :

—o—-> CORRELATOR --->5 §__->

*--> CORRELATOR i-—>

/ E DECORRELATOR E 1 002 5 5 .

1/0 0 K E E .

‘——> CORRELATOR I~——>

F I G. 2 O PRIOR ART

RECEIVED 1 1 9 1 f

8% CORRELATOR >

Q2 7 9 1

RE—MODULATOR E

H I 1 2 9 1 E

CORRELATOR l > r.- ----------------------------------------------------- --L,-2 7 9 2 i I—>


F l G . 2 3 PRIOR ART

FORMULA (1) : r = aldlq + 4211,52 +---+a,{dKE,, + n K

= xatdkb‘k +n k=l

FORMULA (2) : dike,‘ = 1d,‘

FORMULA (3) : 2, = re,k K

= zakd,‘ + (Z aidiEi + r08,‘ is):

FORMULA (4) : a =( “I 32 3‘) = "A

: '(El ‘2 Ex) = Ad - é+n?

A _ (a1 a2 ax)

d _ (d1 d2 dx) 2 E162 F‘15x

a = 51.5‘ z __ 5A ' cK-lcK

5K5‘ EKG: z

FORMULA (5) : £1 = Ad - c+na

1 cm C11:

c2.I 1 C — z : Cir-1.x

CK] 6x3 1

FORMULA (6) ;. (“1(-->=&.c-'

FORMULA (7) ; 3"‘) = Ad + n6 - c-I

U.S. Patent

PRIOR ART

FORMULA (a);

FORMULA (9):

FORMULA (10):

FORMULA (11):

FORMULA (12):

FORMULA (13):

FORMULA (14):

FORMULA (15):

FORMULA (16):

FORMULA (17):

Jul. 11,2000 Sheet 17 0f 18 6,088,383

FIG. 24

81"” = zm'Ad — Ad - (zl — (3),” + izmmnqzl — (3),"

lim(I-C)M -> 0 ArI-NI

U.S. Patent

FORMULA (18) :

FORMULA (19) :

FORMULA (20) I

FORMULA (21) :

FORMULA (22) :

FORMULA (23) :

FORMULA (24) :

FORMULA (25) :

FORMULA (26) :

Jul. 11,2000 6,088,383 Sheet 18 0f 18

FIG. 25

s m=0

Li}? (I —sC)M -+ 0

a:l 0 0 a I

c=x. 2. ' “ -. 0 X

0 0 ax

(1-s0q)“ 0 O M .

= X 0 (l-saz) I X_] 0

0 O (I-sQ'K)M

miaxll — sail <1

M

81"" = zm'Ad — Ad - (zI — séym + sZz”""m“:(zI — sé)"l

6,088,383 1

SPREAD-SPECTRUM SIGNAL DEMODULATOR

TECHNICAL FIELD

This invention relates to spread spectrum signal demodu lators and, more particularly, to spread spectrum signal demodulators used for cellular communication systems, indoor Wireless communication systems, Wireless LAN (local are network) systems, etc.

BACKGROUND ART

Systems dealing With spread spectrum signals for code division multiple access (CDMA), in Which a plurality of signals are transmitted in the same band, are subject to inter-signal interference that is generated in dependence on the correlation among codes assigned to individual signals. The characteristics or quality of the signals are deteriorated increasingly With increasing number of signals involved. In addition, signal level ?uctuations result in increased inter ference of higher level signals on loWer level signals, thus greatly reducing the characteristics of the loWer level sig nals.

Some methods for improving signal characteristics by reducing such inter-signal interference have been proposed. One such method is realiZed by a system called decorrela tion system. In this system, decorrelation is made by using knoWn code correlations. FIG. 19 shoWs an example of the circuit construction Which realiZes this system. Referring to the ?gure, designated at 1001 to 100K are correlators, and at 101 is a decorrelator.

As a spectrum spreading scheme, a direct sequence (DS) system is assumed. A received signal r is expressed by formula (1) in FIG. 23.

In the formula (1), K represents the number of simulta neously transmitted signals, dk a k-th (k=1, 2, . . . , K) transmitted symbol, Ck the spreading code of a k-th signal, 6, a despreading operation caused by spreading code ck, n background noise introduced into signal on the transmission line, and ak a reception level.

The rule of formula (2) in the same ?gure is met When the despreading operation 6, and the correlating operation ck are synchroniZed to each other.

The correlating operation ck is completed after one sym bol has been transmitted, and this is expressed by the provision of a delay element Z for one symbol.

With the received signal r, the k-th correlator 100K provides an output dk, Which is given by formula (3) in the same ?gure.

Formula (4) in the same ?gure is a vector expression of all the signals to be demodulated in a single form. In the

formula (4), C is a matrix type operator corresponding to a correlation matrix.

For the sake of brevity, it is assumed that the symbol timings of the signals are synchroniZed, and denoting the correlation betWeen signals i (i=1, 2, . . . , K) and signals j (i=1, 2,. . . , K) by ci,j (ci, i=1, |ci, i 1) We can use formula (5) in the same ?gure, Where the matrix C of the elements ci, j represents the correlation of signals.

The decorrelator 101 obtains an inverse matrix (C_1) to the correlation matrix C, and executes matrix multipli?ca tion on the outputs of the correlators 1001 to 100K as in formula (6) in the same ?gure.

Since the spreading codes of the individual signals are knoWn, the elements ci,j of the correlation matrix C can be

10

15

25

35

45

55

65

2 calculated in advance, and the inverse matrix (C_1) can be obtained in advance.

It Will be seen that by substituting the formula (5) into the formula (6) the signal d(°°) obtained by the above decorre lation can be expressed by formula (7) in the same ?gure.

This means that the decorrelated signal d(°°) comprises the product of the reception level Aof the original signal and the transmitted symbol d and noise component n that is introduced, and is not affected by the other signals that are simultaneously received. In other Words, it is meant that inter-signal interference is cancelled, and that interference cancelled detected signals (or detected signals With can celled interference) can be obtained. The interference can celled detected signals can be demodulated through phase synchroniZation, for instance, and bit determination.

While the operation of the decorrelation system in the case of presence of symbol synchroniZation has been shoWn, in the case of absence of the symbol synchroniZation (asynchronous case) the decorrelation is obtainable as Well, as shoWn in “Near-Far Resistance of Multiuser Detectors in Asynchronous Channels” Lupas S. Verdu, IEEE Trans. Com. Vol. 38, No. 4, April 1990). Denoting the number of symbols transmitted in a sufficiently long time by L, it can be considered that LK synchroniZed signals are transmitted during this time. This means that the decorrelation in the asynchronous case can be attained by dealing With the correlation matrix C having a magnitude of LK><LK.

In the decorrelation system, a change in the number of signals causes a change in the siZe of the correlation matrix C, thus making it necessary to compute again the inverse matrix (C_1) used for the decorrelation process. In other Words, in such case as When signals are frequently turned on and off by voice activation or the like or When delay times are changed quickly in mobile communication, the inverse matrix computation should be made in a very short period of time. The inverse matrix computation generally requires a

computational effort Which is proportional to the cube of the matrix siZe. Speci?cally, the computational effort is propor tional to about the cube of K in the case of presence of the symbol synchroniZation and to about the cube of LK in the asynchronous case. This means that it is difficult to make the inverse matrix computation on the real time basis.

Aside from the decorrelation system, a replica signal cancellation system has been proposed as another system for improving signal characteristics by cancelling inter-signal interference. In this system, replica signals of individual signals are produced and subtracted from the originally received signal for the interference reduction. The replica signals of the individual signals may be subtracted either one by one (serial system) or collectively (parallel system) from the original signal in FIGS. 20 and 21 shoW examples of the constructions of the serial system (or also called successive system) and the parallel system, respectively. The illustrated systems are for the case Where the number

of signals is K. Re-modulators 1291 to 129K generate replica signals by re-spreading symbols detected in correla tors 1191 to 119K. Replica signal cancellers 1491 to 149K subtract the replica signals of the other stations either one by one or collectively from the original signal (i.e., received signal), and correlators 1591 to 159K again detect symbols. The interference cancellers 2791, 2792 and also the inter ference cancellers 2891, 2892 are identical or alike in construction. Aplurality of such interference cancellers are connected such as to iteratedly carry out signal processing With the re-modulators, replica signal cancellers and corr

Spread-spectrum signal demodulator

Documents