CA2116127C

CA2116127C - Method and apparatus for canceling spread-spectrum noise

Info

Publication number: CA2116127C
Application number: CA002116127A
Authority: CA
Inventors: Eugene J. Bruckert
Original assignee: Motorola Inc
Current assignee: Motorola Solutions Inc
Priority date: 1992-06-29
Filing date: 1993-06-14
Publication date: 1999-03-16
Anticipated expiration: 2013-06-14
Also published as: BR9305563A; FI940952A; CA2116127A1; US5325394A; FI940952A0; TW263632B; CN1082287A; MX9303883A; SE9400545L; JP3564129B2; US5224122A; WO1994000917A1; KR960012479B1; CN1052361C; DE4392999T1; JPH06510415A; SE9400545D0

Abstract

A spread-spectrum noise canceller (182) is provided. A
received phase and a received amplitude for a first (216) and a second (236) component of a received spread-spectrum signal (200) is determined. The second component (236) is structurally similar to the first component (216), but differs by being received at a different time, being transmitted along a different path, or having a different phase. In addition, the spread-spectrum signal (200) includes a first and a second known signal. A portion of a spread-spectrum noise signal in the received signal (200) is canceled by generating an estimated signal (270) by spreading (260) the second known signal at the second component received phase (224) with the first known signal at the first component received phase (204) and adjusting a gain (268) of an integrated form of the spread second known signal as a function of the received amplitudes of the first (216) and the second (236) components. Subsequently, the second known signal is processed out of the received spread-spectrum signal (200) by subtracting (166) the estimated signal (270) from a demodulated form (216, 236) of the received spread-spectrum signal (200).

Description

wo 94/00917 2 1 1 6 1 2 7 PCI'/US93/05622 METHOD AND APPARATUS FOR CANCELING
SPREAD-SPECTRUrA NOISE

Fleld of the Invent;on The pr~scrft inven~ion relates to co~--n-~mication syst~...s which employ spr~aJ-spectrum si~nals and, more particularly, to a ,..~tl-oJ
and apparatus for canceling s~n~aJ-spectnum noise in a commu"ioation 1 0 channel.

und ofthe ~n~o~ n In general, the purpose of a communi~tion s~sle,-- is to tndnsl-li 15 infor".ation-bearing signals trom a SOUrCe9 lo~led at ona point, to a user destinat;on, located at an~.tl.Er poin;t some di;etnnoe away. A
communication system generally consists of three basic cG--,pono.l)ls:
tr~ns...;lter, channel, and ~c~iver. The tI~ns~ ler has the function o p~ -g the ll-6ss~e signal into a form suitable ~r tr.~nsl";ssion over 20 the channel. This proc~C;ng of the ",ess~se signal is referred to as mo~ul~ion. The function of the c~.~nn~l is ~o pr~vids a physical connection between the l-~r-s",itler output and the receiver input. The function of the rsc~:ver is to pr~cess the received signal so as to WO g4/00917 PCI/US93/05622 2116127 -2.

produee an estimate of the original .,.~ssage signal. This proce~sing of lhe ~ eJ signal is refened to as ~e.,-~dulation.
Two types of channels exist, namely, point-to point ehannels and l channels. Examples of point-to-point channels include wirelines (e.~., local telephone transn.;ssion), microwave links, and optical fibers. In con~ ~, L.ur 'ee-t channels pr~iJe a eapability where many receivin~ s~ions may be reaehed simultaneously from a ~ single transmi~ter (e.~., local tebvision and radio stations).
Analog and di~ital transmission methods are used to transmit a rnessage signal over a commun~eation channel. The use of digital methods offers several operational advantages over analog methods, includin~ but not limited to: increased immunity to channel noise and interference, flexible operation of the system, common format tor the -~
trans...~ssiGn of different Wnds of message signals, and in-~e~l security of eommunication throu~h the use of eneryption.
These advantages are attained at the eost of inc~&sed transmission ~channel) bandwidth and ir~ot sed system complexity.
ThrouQh the use of ve~ scale inte~ratbn (VLSI) teehnology a cost~tl~ve way of buildin~ the hardware has been developed.
One d~ital transmission method that may be used for the ~im~sbn o~t mess~oe d~nals over a communication channel is - ~ puls~code modubtion (F'CM). In ~CM, the messa~e si~nal is sampled, q~r~d. and then encod~d. The samplin~ operation permits rep~ent~ion of the ssage si~nal by a sequence of samples taken at un~ pacsd instants of time. bw~ion trims the amplltude of - each ~npb to the nearest value seb~ed 11om a finite set of represent~on hvels. The combination of samp~ng and quantization ~; permits~the use of a code (e.g., binary code) forthe t ansmission of a m~ge ~bnal. Other brms of dgital transmission use similar methods to transmit message signais over a communication channel.
When message signals are digitally transmitted over a band-~-~ limited channel. a hrm of interference known as intersymbol interference may result. The effect of intersymbol intorference, if lef~
- uncont-olled, is to severe~ limit the rate at which digital data may be 3~ tlansmmed without error over the channel. The cure for controtling the effects of intor~..,bol interfor~no~ may be controlbd by carofully shaping the transmitted pulse re"r~s~nting a binary symbol 1 or 0.

WO 94/00g17 2 1 1 6 1 2 7 Pcr/US93/05622 Further, to transmit a message signal (either analog or digital) over a ba..J~.ass communication channet, the message signal must be rnanipulated into a form suitable tor efficient transmission over the channel. Modification ot the messa~e signal is achieved by means of a I,.ucesstermedmodulatiûn. Thispr~cessinvûlvesvaryingsome parameter of a canier wave in accordance with the ~.-~ssage signal in such a way that the message infomlation is pr~se~ved and that the spectrum of the modulated wave contained in the assigned channel bandwidth. CG.-~ndingly, lhe receiver is required to re~reate the original message signal trom a degraded version of the ~ s.--;tl.
signal after propagation through the channel. The re~f~latio.- is a~,.-plished by using a p~cess known as demodulation, which is the inverse of the modulation proo~ss used in the transminar.
In addition to providing efficient transmission, there are other reasons for performing modulation. In particular, the use of modulation permits mul1iplexing, that is, the simultaneous transmission of signals ff~m several message sour~es over a common channel. Also, modulation may be used to conveft the message signal into a torm bss susceptible to noise and interterence.
Typically, in propagating thfwgh a channel, the transmilted signal is distQfted because ot nonlinearlties and imperfections in the ~requency response of the channel. Other sources of degradation are noise and interference added to the f~ceived signal during the course of tfansmission thf~u9h the ~channel. Ndse and disloftion cor~stitute two basic !imi~atbns in the desi~n of oommunication systems.
There are various sour~es of noise, intemal as well as extemal to the system. Although noise is fandom in nature, it may be desaibed in tefms of its sta~cal prope~ties such as the avera~e power or the spectral distribution of the average power.
In any communicahon system, there are two primary communic~ion resources to be employed, namely, average transmitted power and channel band~Adth. The avera~e transmitted power is the average power of the transmitted si~nal. The channel bandwidth defines the range ot trequencies that the channel uses for the transmission of signals with satisfactory fidelity. A general system design objective is to use these two resources as efficiently as possible.
In most channels, one resource may be considered more i.--pG,lant than WO 94/00917 PCI'/US93/05622 the other. Henee, we may also c assi~y communieation ehannels as power-limited or band-limited. For example, the telopl-ore eircuit is a typical band-limited channel, whereas a deep-space communication link or a satellite ehannel is typieally power-limited.
The tran~.~ teJ power is i~ o.lant bee~use, for a r~c 3iver of preseribed noise figure, it determines the allowable separation between the transmitter and r~ivcr. In other words, fora r~c~ivor of ~ ,sc,;~.J
noise figure and a preseribed distanee between it and the transmi~ter, the availabb transmitted power determines the signal-to-noise ratio at the receiver input. This, slJ~s~uent~, determines the noise performanee of the reeeiver. Unless this pc.fo.---al)ee exeeed~ a eertain design level, the trans---iss-on of message signals over the ehannel is not eonsidered to be sa~isf~ory.
Additionally, ehannel bandwidth is i.--pG.lanl, be~se, for a pr~&.ibaJ band of trequeneies eharaeterizing a message signal, the ehannel bandwidth determines the number of such message signals that can be multiplexed over the ehannel. In other words, for a preseribed number of independent message signals that have to share a common ehannel, the channel bandwidth determines the band of frequencies that may be albtted to the transmission of eaeh message si~nal without discemible distorlion.
For spread-spectrum communication systems, these areas of concem have been optimized in one par~icular manner. In spread-spectrum systems, a modulation technique is utilized in wbich a t~ansnitbd si~nal is spread over a wWe frequen~y band. The frequency band is wider than the minimum bandwidth required to transmit the information being sent. A voice sbnal, hr exampb, can be sent with amplitude modulation (AM) in a bandwidth onb twice that of the inhrmation itself. Other hrms of modubUon, such as ~ow deviation frequency modulation (FM) or single sideband AM, also permit information to be transmitted in a bandwidth comparable to the bandwid~h of the inhrmation itself. A spread-spectNm system, on the other hand, often takes a baseband signal (e.g., a voice channel) with a bandwidth of only a few Idlohertz, and distributes it over a band that may 3~ be many megahertz wide. This is a~.-~plished by modulating with the inhrmation to be sent and with a wideband oncGding signal. Through the use ot spre~s~e~.1.um mo~l~J'~ion, a ..,&ssage signal may be 2 1 1 6 1 2 7 PCI'~US93/05622 tr~ns..,illed in a channel in which the noise power is higher than the signal power. The mo~ on and de.--uJulation ot the n.essage signal provides a signal-to-noise gain which enables the r~cwe.y of the ...essage signal trom a noisy channel. The gr~ter the signal-to-noise 5 ratio for a g~ven system equates to: (1 ) the smaller the bandwidth required to transmit a message s4nal with a bw rate of error or (2) the lower the average transmitted power required to transmit a ...~s~g~
si~nal with a low rate ot error over a given bandwidth.
Three general types ot spr~&~spectrum communication 10 techniques exist, including:

The modulation of a carrier by a digital code sequence whose bit rate is much higher than the info, ---dtion signal bandwidth. Such systems are reterred to as ~direct sequence~ modulated systems.

Carrier frequency shffling in discrete increments in a pattem dictated by a code sequence. These systems are called ~frequency hoppers~. The transmitter jumps from frequency to frequency wtthin some predetermined set; the order of frequency usage is determined by a code sequence. Simibrly ~me hopping~ and ~time-frequency hopping~ have times of transmission which are regulated by a code sequence.
;-- Pulse-FM or ~chirp~ modulation in which a carner is swept over a wide band during a given pulse inten~al.

Information (i.e., the message signal) can be ~ ~Wed in the 30 spectrum signal by several methods. One method is to add the information to the spreading code before ~t is used 10r spreading modulation. This technique can be used in direct sequence and frequency lloppi.~ systems. It will be noted that the infonnation being sent must be in a digital form priorto adding it to the spreading code, 35 because the combination of the spreading code, typically a binary code, imolves modulo-2 ~-d~l;tion. Altematively, the infG.--,ation or ,-,ess~
signal may be used to mo~ te a carrier before sp.e~i-,!a it.

WO 94/00917 PCI'/US93/05622 2~1 61 2 7 -6- ~

Thus a spread-spectrum system must have two properties: (1 ) the trar,s."illecJ bandwidth should be much greater than the bandwidth or rate of the ir,lGr"lalion being sent and (2) some function otherthan the information being sent is employed to deter",ine the resulting 5 modulated channel bandwidth.
The essence of the spread-spectrum communication involves the art of expanding the bandwidth of a signal l~dna-,-itling the expanded signal and recovering the desilu.l signal by r~",apping the received spr~J-spectrum into the original infG""dtion bandwidth. Furthermore 10 in the ,c rucess of carrying out this series of bandwidth trades the purpose of spread-spectrum techniques is to allow the system to deliver info-",dtion with low error rates in a noisy signal envirul"nent.
The present invention enhances the ability of s~re~-spectrum syste."s and in particular code division muttiple ~ss (CDMA) 15 cellular radio-telephone systems to r~ /er sp.~-spectrum signals from a noisy radio communication cl,annel. In CDMA cellul~ radio-tGlop~,Gne systems the users~ are on the same frequency and separdtQ~I only by unique user codss. The noise interference level in the communication char,nel is directly related to the inle, fbr~nce level 20 cr~ale.J by the users plus additive R ~"5s;~n noise and not solely by additive ~i~uss~ noise like in other communication systems. Thus the number of users that can simultaneously use the same frequency band in a given cel~ ar region with a low relative of additive ~ ~ Is~i~n noise is limited ~c ,i,llalily by the code noise of all active users~. The presenl 25 invention reduces the effects of urlJesireJ user code noise and thus si~nifican~ incleases the number of users which can simultaneously be serviced by a given cellular region.

Summary of the Invention A spread-spectrum noise canceller is provided. A received phase and a received amplitude for a first and a second component of a received spread-spectnum signal are ~lele""ined. The second cG",ponent is structurally similar to the first component but differs by 3~ being received at a different time being trans",i~leJ along a different path or having a différent phase. In ~J~Iition the spread-spectrum signal inclu~les a first and a second known signal. A portion of a WO 94/00917 2 1 1 ~ 1 2 7 P~r/us93/os622 spread-spectnum noise signal in the received signal is canceled by generating an estimated si~nal by s~r~aJiny the secel,d known signal at the seoGn~ co-..po.~ant received phase with the first known signal at the first c;c ,..pGnent r~ceivGd phase and adjusting a gain of an integrated form of the secon~ spr~.a~l known signal as a tunction of the l.,cGiveJ amplitudes of the tirst and the sew,.J co---poncnt~.
Subsequently, the second known signal is ~,r. c ~s~ out of the received ~"~a~spec~rum signal by subtracting the esti.nal~J signal trom a demodulated torm ot the r~iveJ spr~aJ ~r~um signal.
Briet D3~c-i~)tion of the Drawings - FIG. t is a diagram showing a prior art s~r~aJ-spectrum communication system.
FIG. 2 is a diagram showing a preferred emL~imo.~t intemal stn~cture of a receiver having a spread ~pe~um noise canceller for use in the prior art spread-spsctrum communication system shown in FIG. 1.
FIG. 3 is a tbwchart sumrnarizing the operation of the preferred embodiment noise cancelbr shown in ~IG. 2.

Detailed Description Reterring now to FIG. 1, a prior art spread-spectrum communication system, as substantially described in U.S. Ratent No.
5,103,459 for Gtlhousen et al. tiled June 25, 1990, and ~On the .. ystem Design Aspects ot Code Division Multiple Access (CDMA) Applied to Digital Cellular and Personal Communication Networks,~ Albn Salmasi and Klein S. Gilhousen, presented at the 4t~t IFFF Vehi~
Teohnol~y ~ n~ n~ on May 19-22,1991 in St. Louis, MO, pages 30 ~7-62, is shown.
In the prior art spr~a~J ~Fwtrum communication system, traffic channel data bits 100 are input to an encoder 102 at a particular bit rate (e.~., 9.6 kilobits/second). The traffic channel data bits can include either voice converted to data by a vocGJ~r, pure data, or a combination 35 of the two types of data. ~coJer 102 convolutionally Gnw~J~s the input data bits 100 into data symbols at a fixed encoding rate. For example, encoder 102 encoJes rec~ive~ data bits 100 at a fixed enco~ing rate of .. .... . .. . . . .... . ... . .. .

WO 94~nO917 - PCI'/U$93/0~622 2116127 -8- :

one data bit to two data Sy.-~ls such that the anco~ler 102 e~nr~S data ~y.-,~ols 104 at a 19.2 ki'o~ bols/second rate. The anc~uer 102 ~,"",~ ~1es the input of dala bits 100 at variable rates by encoding repetition. That is when the data bit rate is slower than the panicular bit 5 rate at which the encoder 102 is designed to operate, then the encoJer 102 repeats the input data bits 100 such that the input data bits 100 are provided to the encoding elements within the encoder 102 at the equivalent of the input data bit rate at which the e~ elements are designed to operate. Thus, the encoder 102 outputs data sy.nbols ~04 10 at the same fixed rate regardless of the rate at which data bits 100 are input to the encoder 102.
The data symbols 104 are then input into an ir,terled~er 106.
Interieaver 106 interleaves the input data s~ bols 10~. This interleaving of related data a~ bols 104 c~ses bursts of errors in a 15 communication channel 1~8 to be spr~aJ out in time and thus to be handled by a decoder 178 as if they were independent rdn.lo.-. errors.
Since, communication channel 138 memoly deuer ses with time separation, the idea behind interleavin~ is to separate (i.e., make independent) the related data symbols 104 of an encoded data bit 100 20 in time. The inten~enin~ space in a transmission bbck is filled with other data s~ bels 104 related to other encoded bils 100. Separating the da~a ~.-~bols 104 sufficiently in time ettectively transforrns a communication channel 138 with memo~y into a memorybss one, and thereby enables the use of the random-error co~cting codes (e.g., 25 ' convolutional codes and block codes). Subsequently, a maximum likelihood convolutional deooder 178 can make a decision based on a sequence ot data sampbes 176 of a received signal in which each data sampb 176 is assumed to be independent trom the other data sampbs 176. Such an assumption of independence of data samples 176 or 30 memo~ylessness of the communicatbn channel 138 can in-pr~/e the performance ot a maximum likeUhood decoder 178 over a d~er which does not make such assumptions. The interleaved data s~ ols 108 are output by the interleaver 106 at the same data symbol rate that they wers input (e.g., 19.2 kilosy..,bols/second) to one input of an 35 FY~ sive-OR/mu~tiplier 112.
A long psQudo-noise (PN) generator 110 is oporatively coupl~d to the other input ot the FYcb~sive-OR/multiplier 112 to enhance the WO 94/00917 2 1 ~ ~ 1 2 7 Pcr/us93/05622 g securi~ o~ the communication in the communication channel by scrambling the data sy,.~bols 108. The lon~ PN ~enerator 110 uses a lon~ PN sequence to ~enerate a user .spe~f~c sequence of sy",l,ols or unique user spreading code at a fixed rate equal to the data symbol rate 5 of the data sy..lbols 108 which are input to the other input of the FYclusive-OR gate 112 (e.~., 19.2 kilosy..,bols/seco.,J). The scn~,.lbled data symbols 114 are output trom the FYc'~sive-OR/multiplier 112 at a fixed rate equal to the rate that the data sy--~Lols 108 are input to the FYc~sive-OR gate 112 (e.~., 19.2 kilosy.-llJols/seco.,d) to one input of an Exclusive-OR/multiplier 118.
A code ~i1/ision channel selec~ion generator 116 provides a particular predetermined len~th Walsh code to the other input of the Exclusive~Rlmultiplier 118. The code d;.i~on channel selsction generator 116 can provide one of 64 G,lho~onal codes cG,-~s~nding to 64 Walsh codes from a 64 by 64 Hadamard matrix wherein a Walsh code is a single row or column of the matrix. The Exclusive-OR/multiplier 118 uses the particubr Walsh code input by the code division channel ~enerator 116 to spre~ the input scrambled data symbols 114 into Walsh code spread d~ta symbols 120. It will be appreciated by those sldlled in the art that ~spnading~ is a term used to ~-describe the operation of increasin~ the number of symbols which represent input data symbols. For example, the combiner 118 may receive a sequence of scrarnbled data s~r.-~ols 114 at a rate of 19.2 kilosymbols~ nd. Each scrambbd data symbol 114 is,combined with a Walsh spreading code 116 such that each scrambled data symbol 114 is represented by or spread into a single 64 bit bngth Walsh spreading code 120. As a resutt, the Walsh code ~ _~ data symbols 120 are output from the FY~lt~sive~Rlmultiplier 118 at a fixed chip rate (e.g., 1.2288 Megachip~seco..d). The term ~chip~ is used in 30 the art interchangeable with the term ~bits~ when describing segments o~
a spread digital signal.
The Walsh code spread data symbols 120 are provided to an input of two FYclusive-OR/multipliers 122 and 128, respectively. A pair of short PN sequences (i.e., short when compared to the bng PN
35 sequence used by the long PN generator 110) are generated by 1-channel PN generator 124 and ~channel PN generator 130. These PN genGr~tors 124 and 130 may gener~te the same or different short WO 94/00917 PCI'/US93/05622 PN sequences. The FY~ sive-OR/mu~tipliers 122 and 128 fur~her spread the input Walsh code spread data 120 with the short PN
sequences generated by the PN l-channel generator 124 and PN Q-channel generdter 130, l~spe.;ti~ely. The resuHing l-channel code ~re~.J s~uence 126 and ~channel code :",r~J sequence 132 are used to quadrature-phase shift key (QPSK) modulate a quadrature pair of sinusoids 134 by ~;J;n~ the power level controls of a pair of sinusoids. The sinusoids' output si~nals are summed, l,a,-.l~ass filtered, tr~..slalh~ to a radio frequency (RF), arnplified, filtered and radiated by an antenna 136 to complete tra--s,--;ssion of the traffic channel data bits 100 in a communication channel 138.
Antenna 140 r~ce VGS a spread-spectrum signal such that the received signal can be ,ur.t~ sed in a wbstantialb complementary set of operations as compared to the set of operations p~.",eJ on the traffic channel data bits 100 prior to their transmission over communication channel 138 by antenna 136. The received spreaJ-spectrum si~nal is translated to a baseband frequency, fittered, and QPSK demodulated 142 into a demodulated spread 6~Y1-um signal 144,146. Subsequently, the demodulated spread-spectrum signal 144,146 is quadrature despread. A pair of short PN sequences are ~enerated by l-channel PN ~enerator lU and Q-channel PN generator 154. These PN ~enerators 1U and 154 must ~enerate the same short PN sequences as the PN ~enera~ors 124 and 130, respectively. The Exclusive-OR~/multipliers 150 and 152 despread the input demodulated sprea~spectrum signals 144 and 146, l~pe~ eb. The resulting 1-chanriel code despread sequence lS6 and ~channel code despread sequence 158 are combined into quadrature despread data samples 160.
A code di~shn channel selection generator 164 ~roJ;d~s a particular predetennined length Walsh code to an input o~ the FY~ sive-OR/multiplier 162. The code division channel selection generator 164, like generator 116, can provide one of 64 orthogonal codes corresponding to 6~ Walsh codes from a 64 by 64 Hadamard matrix wherein a Walsh code is a single row or column of the matrix, bu~ to prope~y despread a particular code transmission the same Walsh code as the tra.,s",iller generator 116 generated must be gen~rdlecJ. The htc~sive-OR/multiplier 162 uses the particular Walsh code input by the WO 94/00917 2 1 1 6 1 2 7 PCl'/US93/05622 code di~[~bn channel generator 164 to despread the input quadrature despread data samples 160 into Walsh code despread data samples 166. It will be appreciated by those skilled in the a~t that ~d6s~.r~ading~
is a term used to descr~be the operation o~ decreasing the number of 5 sampbs which represent input. For exampb, the combiner 162 may receive a sequence of despread data sampbs 160 at a rate of 1.2~88 Megasamp~e~ nd. A group of 64 despread data samples 160 is combined with a selected Walsh despreading code 164 such that the group of 64 despread data samples 160 is represented by or despread 10 into a single Walsh despread data sampb 166. As a result, the Walsh code despread data samples 166 are output from the FYc~u~ive-OR/multiplier 162 at a fixed rate (e.~., 19.2 kibsampbs/second).
A bn~ PN generator 170 is operativety coupl~d to the input of the FYclu~ve-OR/muttip~er 168 to desctamble the dGs~r ,z~ data samples 166. The bng PN generator 170 uses a bn~ PN sequence to generate a user spedfic sequence o1 sampbs or unique user spreading code at a fixed rate equal to the data samples rate of the despread data samples 166 which are input to the other input of the FYclu~iv~OR gate 168 (e.g., 19.2 Wlosampb~second). This operation uses the sa ne bng PN sequence as ~ener~ed by bn~ PN generator 110 and is the ical compbment of the sc~nblin~ operation per~ormed by the Exclusive-OR gate 112. The descrambbd data sampbs 172 are out~wt trom the Exclusive OR/multiplier 168 at a fixed rate equal to the rate that the despread data samples 166 are input to the Exdusive4R
gate 168 (e.g., 192 kibsamphs/second).
-The d~mbbd data sampbs 172 are then input into a deinterleaver 174. Delnterleaver 174 deinterleaves the input descrambbd data samples 172 in a manner whi'ch is the bgical compbment ot the interteaver 106. The deinterbaved data sarnples 176 are output by the deinterleaver 174 at the same data sample rate that they were input (e.g., ~9.2 Idlosamphs/second). Subsequently, a maximum likelihood convolutional decoder 178 makes decisions based on the input sequence of deinterleaved data samples 176. The maximum likelihood decoder 178 preferably generates estimated data bits 180 by utilizing maximum likelihood decoding techniques which are substantialq similar to the Viterbi doc~in!a aborithm.

WO 94/00917 ' PCI'/US93/05622 Referring now to FIG. 2, a Jic.y.~llHs shown of a preferred embodiment intemal structure of a "GniGn 182 of a ~~c~i~for having a spr~-~spectnum noise canceller for use in the prior art spread spe~rum communication system shown in FIG. 1.
S The receiver pGftion ~82 as d6sc-il,cJ hereinafter preferably is implemented in a mobile communication unit of a cellular radio communication system also having a plurality of base st~tions or central communication sites. It will be appreciated by those skilled in the art that the parlicular receiver portion structure 182 having a noise canceller d6so-;~ herein could readity be adapted tor use in the central communication sites or in any other communication system having similar knowledge of the muttipath characteristics of the signal received on a communication channel.
It will be appreciated by those skilled in the art that spreading S codes other than Walsh s~r~&~in,~ codes 116, 164 can be used to separate data signals from one ano!her in a CDMA communication system. For instance, PN spreading codes can be used to separate a plurality ot data signals. A particular data signal can be separated trom the other data signals by using a palticular PN s~reading code which is offset by a paracular phase to spread the particular data signal. For example, in a CDMA sp~ad-spectrum communication system, a particular PN spreading code can be used to generate a pluratity of channels by using a different offset phase for the PN spreading code for ea~h channel of the communication system. Furthermore, the modu!ation scheme of the signals is assumed to be quadra~ure phase shm-keyin~ (QPSK). However, it will be appreciated by those skilled in the an that other modulat on tect n~ues can be used without departing from the teachings of the present invernion. Finally, in the preferred embodiment, the communication channel 138 ~or the cellular communication system is in the 900 MHz region of the electromagnetic spectrum. However, other r~ions of the ebctromagnetic spc t~um may be used without departin~ from the teachin~s of the present invention.
The por~ion 182 of the rGc~iver shown implements ~Rake~
receiving techniques to reduce the e~tect ot multipath fading in the communication channel. It will be appreciated by those skilled in the art that ~Rake~ r~cei~iry techniques are well known in the art of radio communication. For example, ~A Communication Technique ~or .

WO 94/00917 2 1 1 6 1 2 7 Pcr/US93/05622 Multipath Channels,~ R. Price and P.E. Green, Jr., r~ d~Q nf tlle LBE. March 1958, pa~es 5~570 describes the basic operation ot a ~Rake~ receiver. Bnefly, a ~Rake~ receiver performs a continuous, detailed measurement ot the multipath characteristic of a l~ceived 5 si~nal. This knowbdge is then expbited to combat the selective fading by detecting the echo signals individually, usin~ a correlation method, and abebraically combining those echo signals into a single detected signal. The intersymbol interference is attenuated by varying the time deby or phase between the various detected echo si~nals prior to their 10 algebraic combination.
Similar to the prior art communication system shown in FIG. 1, the antenna 140, shown in FIG. 2 receives a spread-spectrum signal such that the received signal can be pr~95e 1 in a substantial~
complementary set of operations as compared to the sat ot opa~tions ~-16 penommed on the traffic channel data bits 100 prior to thair tra--s-,-;ssion over communication channel 138 by antenna 136. The received sprea~spectrum signal is a composite signal including several signals in different spread-spectrum channels. At least one of these sprea~
spectrum si~nals i~8 a h~own pilot data si~nal. Each of spread-spectrum 20 signals in the con~osite received spread-spectrum dgnal may be received by ~iver 182 trom one or more base stations and alon~ one or more communication paths. As a result, each ot the signals In a pa~icubr spread-spectrum channel may have several components which vary in ~mp~ and/or phase hom the other signals in the 25 channel. In the preten~d embodiment, similar pilot data si~nals are t~nsmitted ~rom each~base slabon in the communication system.
However, when a mobib communication unit is attempting to retdeve (i.e., demodubte and decode) a puticular signal from a spread- ~ ' spe~rum channel, these pibt data sbnals contribute to the non-30 deterministic noise in the communication channel 138. These undesired signals can be cancebd when the receiver has obtained particubr infomnation oonceming 1he communication channel and the received ~.."osite spread-spectrum si~nal.
The spread-spsctrum si~nal r~csivsd on antQnna 140 is 35 translated to a baseband frequency, filtered, and QPSK demodulated 142 into a demodulated spread-spectrum signal 200. During this demodulation pro~ss 142, a received phase and a receiveJ amplitude WO 94/00917 PCI/USg3/0~622 ~or each c~ onent of the received spread-spectrum signal is determined. The phase represents the moment in time that a particular component is received relative to the other components. The amplitude represents the relative received signal strength or .3~iveJ accuracy of the component relative to the other components. During the tolbwing discussion, the re~i~ spread-spectrum si~nal is assumed to have a signal in one particular spread spe~rum channel and further that the signal has three components. These signal components have followed dmerent communication channel paths on their way to receiver 182.
For this example, the first component was transmilted by a base station in a primary seming cell and was received at a phase ~1 and an amplitude A1. Simibrly, the second component was transmitted by 1he base st~ion in the primaly sen~ing cell, but traveled along a different communication path than the flfst component, and was received at a phase ~2 and an amplitude A2. Fmally, the third c~r "~onent was transmilted b~y a base station in a secondary serving cell (e.g., during a so~t hand off situation) and was received at a phase ~3 and an amp~de ~.
In the preterred embodiment ~Rake~ receiver 182, demodulated spread-spectnJm signal 200 is input to individual receiver po~tions which manipubte each of the three sbnal components. The first signal component is quadrature despread by inpu~n~ the demodulated spleKl~ctnum si~nal 200 into Exolusive~OR com~ner 202. A pair of short PN seq~Jences are ~enerated by l channeî PN ~ene ator 148 and O-channel PN ~enerator 154 (shown in FIG. 1). The pair ot short PN
sequences is input 204 to the E~lusive-OR combiner 202 at the first componentphase~1. Exclusive4Rcombiner202despreadsthe input demodubted spread-spectrum signal200. Inadditbn, FY~ v~OR
~ combiner 202 combines the resulting I-channel code despread sequence and a-channel code despread sequence into quadrature despread data sampbs 206. It will be appreciated by those skil~d ~n the art that althou~h a single Exclusive OR combiner 202 is described above, Uke in the prior an recsiver, shown in FIG. 1, t Ho E dusive-OR/multipliers (e.g., multipliers 1S0 and lS2) could be used.
The quadrature despread d~a samples 206 for the first signal component are input to FYr~sive-OR/multipîier 208. A code ~ii.ision channel sebction generator 164 (shown in FIG. 1) proJi~es a particular predetermined length Walstl code (Wi) at the tirst signal cG.,-~)onent phase ~1 210 to the other input of the Exclusiv~ORUmultiplier 2~8. The Exclusive-ORtmultiplier 208 uses the particular Walsh code ~ ) 210 input by the code diJ~sion channel generator 164 to JespreaJ the input 5 quadrature despread data samples 206 into Walsh code despread data samples 212.
These Walsh code despread data samples 212 are then input to integrator 214 which integrates the data samples 212 over a predetermined time period (T) and adjusts the gain of the input data 10 samples 212 signal. The predetermined time period (T) prG~r~ly C0.-~5pO..~ to a desired output rate of data sampbs from the ~Rake"
receiver 182 (e.~., 19.2 kilosamplests~co,-d output rate which co..~nds to T,1/19,200 of a s~oo.~). The gain of the input data sampbs 212 si~nal is adjusted by a ~ain factor 91 which is a function of 15 the amplitude of the first signal cG.-.ponent A1 (g1 = f(A1)). This gain factor 91 is also determined such that it enables ...a~ -um ratio cG.-~b:ning of the three signal components. In addition, the input data sample 212 gain is divided/adjusted by the predetermined time period (T) so that the output signal 216 gain better re~ s the gain assoc~iated ~Ath each input data sample 2t2. The output of integrator 214 is a Walsh code despread data sample si~nal 216 forthe first signal component. This first si~nal component Walsh code despread data sample signal 216 may optionally be switched into an input 218 of a signal pr~ssor 220. R will be appreciated by those skilbd in the art that the integrator 214 function may be implemented with a data sample summing circuit and multiplier.
The B~-l,l signal component can be derived from the demodulated spread ~e~l~um signal 200 in a manner similar to that previous~ des~.;bcJ tor the ~irst si~nal component. The seosn~i si~nal component is quadrature despread by inpu~ting the demodulated sprea~spectnum signal 200 into FY.~sive-OR combiner 222. A pair of short PN sequences are ~enerated by l-channel PN ~enerator 148 and Q-channel PN generator 154 (shown in FIG. 1). The pair of short PN
sequences is input 224 to the ~tcl~sive-OR combiner 222 at the seco"J co-l-pGne.lt phase q~2 F~ sive-OR combiner 222 da-~pre~Js the input .le."G-hllated spleaJ-specl,-lm signal 200 and combines the 2116127 -~6-resulting l~hannel code d6~rda~ sequence and ~channel code despread sequence into quadrature despread data samples 226.
The quadrature .I~ ,r~,ad data samples 226 for the seco..J
signal component are input to FYch~sive-OR/multiplier 228. A code S di:ision channel selection generator 164 (shown in FIG. 1) provides a particular predetermined bngth Walsh code (Wi) at the second signal component phase ~210 lhe o~her input of lhe FYc~usive~R/multiplier 228. The Exclusive-OR/multiplier 228 uses the particular Walsh code (Wl) at lhe s~oor,J signal cG.--ponent phase ~2 230 input by the code 10 division channel generator 164 to despread the input quadrature despread d~a samples 226 into Walsh code despread data samples 232.
ll~ese Walsh code d~,reaJ data sampbs 232 are 1hen input to integrator 234 which integrates the data samples 232 over a 15 predetermined time period (T) and adjusts the gain of the input data samples 232 si~nal. The gain ot the input data samples 232 signal is adjusted by a ~ain tactor 92 which is a tunction of the amplitude of the seoond si~nal component A2 (92 ~1(A2)). This ~ain tactor 92 is also determined such that it enables ma~amum ratio combining of the three 20 signal components. In addition, the input data sample 232 ~ain is divided/adjusted by the predeterrnined ffme period (T) so that the output si~nal 236 ~ain be~ter reflects the ~ain associ~ed with each input data sample 232. The output of inte~rator 234 is a Walsh despread data sample si~nal ,236 for the second si~nal component. This second 25 si~nal oomponent Walsh oode despread data sampb si~nal 236 may opffonally be switched into an input 238 of the si~nal "r~sor 220.
The third si~nal oomponent can be derived from the demodulated spread-spectrum si~nal 200 in a manner similar to that previously desuibed forthe first and second si~nal components. The third signal 30 component is quad~ature despread by inpufflng the ~...G~Iated spread-spectrum si~nal 200 into FYc'~sive-OR combiner 242. A pair of short PN sequences are generated by l-channel PN ~enerator 148 and ~channel PN ~enerator 154 (shown in FIG. l ). The pair ~t short PN
sequences is input 244 to the FYol~sive-OR combiner 242 at the third 35 component phase ~3. FYo~usive~OR combiner 242 desprea~ls the input demodulated spr~-spedrum signal 200 and co.,ll,i.)es the resulting 1-WO 94/00917 2 1 1 6 1 2 7 Pcr/US93~0s622 channel code despread sequence and Q-channel code despread sequence into quadr~ture despread data samples 246.
The .quadrature despread data samples 246 for the third signal component are input to F~clllsive4R/multiplier 248. A code ~ sion S channel sebction ~enerator 164 (shown in FIG. 1) provides a particular predetermined length Walsh code (WI) at the third signal component pha~se ~3 to the other input of the Exclusive-OR/mu~tipHer 248. The FYc'~sive-OR/mu~tiplier 248 uses the particubr Walsh code (Wl) at the third si~nal component phase ~3 250 input by the c~de di~tsion channel ~ 10 ~enerator 164 lo despread the input quadrature despread data samples 246 into Walsh code despread data samples 252.
These Walsh code despread data sampbs 252 are then inpln to integrator 254 which inte~rates me data samples 252 over a predeterrnined time period (T) and adjusts the gain of the input data sarnpbs ~52 signal. The ~ain of the input dsta samples 252 si~nal is adjusted by a ~ain tactor 93 which is a ~unction of the amplitude of the third si~nal component A3 (93 ~ t(A3)). This ~ain tactor ~3 is also detemlined such that it enabbs maximum ratio combining of the three si~nal oomponents. In addition. lhe input data sarnpb 252 ~ain is d'iv~adpJsted ioy the predetennined time period (T) so that the output signai 256 gain be~ter ~cls lhe gain ~d vnth each input d~ta sa~nple 2S2. The output of integ~ator 2U is a Walsh dbspread data sampk si~nal 256 for the third ~ignal component. This third signal component Walsh code despread data sample signal 256~may bptionally be s~,ilched into an input 2S8 of the signal pn~sor 220.
The demodulated spread-spsctrum signal 200 turther includes non deterministic noise consistin~ ot two components. The two components to the non deterministic noise are: :
- - All ot the CDMA spread-spectnJm si~nals which are not bein~
demodulatsd by the receiver. These consist ot a br~e number of bw-bvel interfefin~ users usin~ the sarne communication channel as the receiver whioh ue in nearby cells ot the oommunioation system.
Receiver ~ront end noise. By design, additive noise preferably is ~ 35 below the demodulated spread-spectnJm signal 200 when the :: communication channel is operating at tull capacity.

WO 94/00917 PCI/US93/0~;622 A portion of this first spr~ad-spectrum noise component can be canceled from the demodulated spread-spectrum si~nal 200 provided sufficient infonnation is known to the fsceiver. This inforrnation includes several pieces of data already known to a typical ~Rake~ receiver like the 5 preterred embodiment receiver portion 182 described above. This known data ir~bJc~es the amplitude (i~e. A1, A2, and A3) and phase (i.e., 2. and ~3) of each si~nal component, the shon PN spread code sequences 148 and 154 used by the communication system, and the Walsh code (Wl) ~or the particular channel bein~ rec~iv,ed With 1his 10 known d~a, the receiver poltion 182 may be configured to cancel the noise related to other si~nal componerns such as a pilot channel carrier signal which may be interfefin~ with the desired si~nal components.
Typically each Walsh code channel does not cofltnbute any noise to the other Walsh o.ode channels because oftho~onality is maintained.
15 However, this is not l~ue when there is signfflcant delay spread (2 one chip cblay) and~or when the receiving unit is in a communication channel hand-off state between two or more t~nsmilters. A possibb situ~tion in which these other channels may cont-ibute noise or cause interlerence in the desirsd communication channel is when either a 20 debyed replica of the transmitted canier ortransmitted carriers of ofi~inating in other cells is received in the desired communic~tion ch:annel by the receiver poltion 182 and the receiver portion 182 does not distlnguish beh~n the desired signal and the interferin~ signals.
AS more ot these interfering si~nals contribute to the demodulated 25 ~sp~d~rum si~nai 200 received by the receiver, the signal to noise ratio may deteriorate to near or bebw a prefened threshold.
In the prefe~ed embodiment communic~tion system, the debyed pibt signal rep~cas of the prima~y sen~ing cell and pilot signal energy trom other nearby cells cause approx~mately 1 dB ot the tota~ noise in 30 the desired communbation channel. Through the tollowing cancellation prccss~ most of that 1 dB ot noise can be canceled which results in a greater signal to noise ratio tor the desired si~nal. Some of the advantages of ~his cancellation technique include: removing or reducin~ undesired pilot channel signal interference from the receive~
35 si~nal and albwing an inc.~~~e in the number ot users on a particubr CDMA communication channel due to the inc- ~ a~e~ capabili~ ot the receivers to handle interi~rGnce in the communication channel.

WO g4/OOgl7 2 1 1 6 1 2 7 PCI/US93/05622 A first estimated interference signal can be derived from the known data The previous~ generated pair of short PN sequenoes havin~ a second component phase ~2 are input 224 to an FY~sive-OR
combinér 260. Similarb~, pr~iously ~enerated pair o1 shorl PN
sequences havin~ a first component phase ~IPl are input 204 to an FY~ ve-OR combiner 260.
~clusn~e-OR combiner 260 spreads the second component phase ~2 sequences 224 ~nth the first component phase ~1 sequences 204 and oombines the resultin~ l-channel code spread sequence and ~channel code spread sequence into quadrature spread data samples 262.
The quadrature spread data sampbs 262 for the ffrst estimated interferenoe signal are input to Exclusive-OR/multiplbr 264. The previously generated particular predetermined bngth Walsh code (Wl) at the first si~nal component phase ~ 210 is provided to the other input of the FYc'~she4R/mufflplier 264. The E clusive-OR/multiplier 264 uses the pa~icubr Walsh code (Wl) at the first si~nal component phase ~1 210 to spread the input quadrature spread data samples 262 into Walsh code spread data samples 266.
These Walsh code spread data sunpbs 266 are then input to integrator 268 which inte~rates the data sampbs 266 over a predetermined time pefiod (T) and adjusts ~he gain ot the input data sampbs 266 signal. The ~ain ot the input d~ samples 266 si~nal is adjusted by a ne~ative of a prodwt of 9ain tactor 91 and ~2 (- :' 91~2) which as previousb~ noted are tunctions ot the ampli~des of the flr t and second signal components A1 and ~2. respective~. This ~ain 91~2 is dso detemlined swh that n enables a subtr~ction trom the maximum ratio oombination of the three signal components (i.e., a n~bve factor). In addi00n, the input data sarnple 266 gain is adjusted by the predetermined time period (T) so that the output signal 270 gain better re~bcts the gain associated ~Ath each input data sampb 266.
The output of integrator 268 is a flrst estimated Walsh despread data sa ~mpbd interference signal Z70. This first estimated inter~erence signal 270 may optionally be switched into or input to 272 a signal processor 220.
A second estimated interference signal can also be derived from the known data. The previously generated pair of short PN se~ences having a third co...pon~nt phase ~3 are input 244 to an F~c~us~ve~R

WO 94/00917 - PCI'/US93/05622 211~127 -20-combiner 280. Similar y, previous~ ~enerated pair of short PN
sequences havin~ a first component phase ~1 are input 204 to an E_clusive-OR combiner 280.
FY~Isive-OR combiner 280 spreads the third co...~,onent phase ~3 sequences 244 ~nth the first componern phase ~1 sequences 204 and combines the resulting l-channel code spread sequence and O~hannel code spread sequence into quadrature spread data samples 282.
The quadrature spread data samples 282 for the second estimated interference si~nal are input to F-Y~ ve-OR1multiplier 284.
The previous~ ~enerated panicular predetermined len~th Walsh code ~WI) at the first *nal component phase ~1 210 is provided to the other inpùt of the Exclusive~R~multipaer 284. The Exclusive-OR/mul~plier 264 uses lhe panicuhr Walsh code ~WI) at the first signal component phase ~1 210 to spread the input quadrature spread data sampbs 282 into Walsh code spread data sa nples 286.
These Walsh code spread data sampbs 286 are then input to inte~rator 288 which inte~rates the dsta samples 286 over a predetermined ffme pe~iod (T) and adJusts 1he ~ain ot the input data sarnpbs 286 si~nal. The ~ain of the input data sa nples 286 signal is adjust~d by a ne~tive of aprodu~ of ~ain fa~ lor~l and ~3 (-~3) which as p~lou~ noted are funclions of the ampli~cbs of the fi~st and third sbnal oo~onents Al and A3, respec~vely. This ~ain fa~or ~1~3 is al~o deter nined such that it enabbs a subtrac~ion from the maximum latio combin-lion of the three si~nal components (i.e., à
n~be faclor). In addition, the input data sarnple 2B6 ~ain is adpJsted by the predetennined time perbd (T) so that the output si~nal 290 gain be1ter reflects the gain~associated wTth each input data sarnpb 286.
The output of int~ralor 21~8 is a second estimated Walsh despread data sampled interference si~nal 290. This seoond estimated interference signal 290 may optionally be s~dtched into or input to 292 a signal pr~s60r 220.
The ~ene~tion ot the first and second estimated ~nterference - signal was matb by way ot exampb only. It wtll be appreciated by those skilbd in the art that this estimated interfer~n~a si~nal p.oc~ss 36 may be continued tor any other interfering si~nal tor which sufficient inforrnation is known.

WO g4/00917 2 1 1 6 1 2 7 PCl'/US93/05622 Flnally, signal pr a6sor 220 preferably maximum r~tio combines several signal components (e.g., si~nal components 216, 236, 256, 270, and/or 2~0) into a dngle Walsh code despread data sampb 166 signal. This single Walsh code despread data sample 166 S signal is output ~rom the signal pr~s~or 220 at a fixed rate ~e.~., 19.2 kilosampbs/second). Subsequently, the Walsh code despread data sampb 166 signal is preblably fulther prooessed in a manner similar to the prior art receiver, shown in FIG. 1, to generate estimated data bits 180.
It will be appreciated by 1hose skilled in the art that it may not be dedrabb to cancel all of the interfedng signals from the desired signal.
Thus, the signal stren~ths of the interfering signals may be compared to the desired signal. Fu~ther, only the undesired interfering signa~
having a signal strength greater than the desired signal should be removed fr~m the composite demodu~ted spread s!-eYt~um signal 200.
If the weaker undesired interlering signals are removed, then the desired data signal may be partially conupted. In addition, it will be appreciated by those skilbd in the a t that a spread-speclrum signal ~e.g., the desired signal) typically can ~e detected and retrieved trom a composrw si~nal when ~s si~nal strength is ~reater than the si~nal strengths o~ inter~erin~ signals. Thus, the removal of interfering signals trom composlte signal which have a 8i~ strength bss than the desired si~nal is unnecessary and may unduly increase ~he detection and retrieval time of the desired si~nal.
; For exampb, in the caSe of the desired si~nal having the three si~nal component Walsh code despr~ad data sample signals 216, 236, and 256, an interfe~r is removed from the combined single Walsh code desp~ data sampb 166 signal, if it has a stronger signal strength ltlan the desired sbnal components. For in~tance, the tirst estimated sbnal 270 may have a signal stren~th greater than the third si~nal component Walsh code despread data sampb si~nal 256 and as such shouid be removed t om the combined single Walsh code despread d~a sample 166 signal. Thus, the tirst estimated signal 270 is swltched input 272 of the signal processor 220. In contrast, the secG.-d 36 estimated signal 290 may have a si~nal stren~th less than the third signal co.-.ponent Walsh code despread data sample signal 256 and as such should not be r~ o~eJ trom the combined single Walsh code WO g4/00917 PCr/USg3/05622 2~1~i12~ ' -22-despread data sample 166 si~nal. Thus, the second estimated siç~nal 290 is not ~ntched input 292 of the si~nal processor 220. It will be appreciated by those skilbd in the art that another si~nal quality or communication system metfic may be used to determine which 5 interfering si~nals should be canceled from the compo~te spread-spectrum si~nal without depalting from the teachings of the present invention. for exampb, the canceUation of particular interlering signak may be determined by a comparison of a predetermined threshold to a function of the adjusted gains (91. ~2. and 5~3) of the inte~rators 214, 234, and 254, respectiveq~.
The operation ot the preferred embodiment noise canceller is summarized as the flov~ shown in FIG. 3. A spread-spe~rum signal havinSI a ffrst and~ second component is received 300 from over a communic~tion channel. The ffrst component being received at a 15 dfflerent time from the second component. In addition, the received - spread-spect~um si~nal includes a known si~nal (e.g., a cellular communication system pibt channel si~nal). A recen~ed phase (~
and a received amplitude (A1, ~) for the fi~t and the second components of a reoeived spread-spectrum si~nal is determined 302, 20 respecth~eb~ qu~, the receh~ed spreadrspectrum si~nal flrst and seoond ~omponen~s are demoduhted and ma~dmum ~tio oombhed 3U. In addition, an estimated sbnal is ~enerated 306 by spreading the known sbnal at the second component received phase '' (~2) with the known ~bnal at the first component received phase (~), 25 spreadin~ the known sbnal at the second component received phase (~2) with a channel sebc~ing spreading code at the first component receh~ed phase (~), inte~ting the spread known signal over a predetermlned time (T), and adjusting a gain ot the integrated spread known sbnal as a function of the received amplih~des of the ~irst (A1) 30 and the second oomponents (A2). Subsequentb, a portion of a spread-spectnJm noise signal is cancebd 310 by pr~c~3~s7..~ the known si~nal out ot the received spread-spectrum signal th~ugh subtracting the - estimated signal from the demodulsted received spread-spectrum signal onb it 308 the adjusted ~ain of the integrated fonn of the spread 35 known signal (9192) iS greater than a predetennined threshold.
Subsequently, the spr~spectrum signal receiving pr~cess may be completed by ~ bling 312 the ph~s~l demodulated forrn of WO 94/00917 2 1 1 6 1 2 7 PCl'/US93/05622 the received spr~..~spectrum signal by utilizing a known spr~aJi..~
code. In ~-J-);t;o,), the descrambled received spr~ spectrum signal is deinterieaved 314 within a predete....;n~i ske block Finally, at least one esti.-~te~l data bit is ~enerated 316 by utiiizin~ maximum likelihood 5 dec~ding techniques which are su~' ~ntially similar to the Viterbi dec~ii..y algorithm to derive the at least one estimated data bit from the deinterleaved received spread-spectrum si~nal.
Although the invention has been described and illustrated with a certain degree of particularity, it is understood that the present lisdQsure of e--~L--ents has been made by way of example only and that numerous changes in the arrangement and co.-lbin~ti~,- of parts as well as steps may be I~SGrl~ ~0 by lhose shlled in the art without departin~ from the spirit and scope of the invention as claimed. For example, it will be appreciated by those s~illed in the art that the above described noise cancellation techniques can be performed in the intennediate or baseband frequencies without departing from the spirit and scope of the present invention as claimed. In addition, the modulator, antennas and demodulator poltions of the preferred embodiment communication system as described were directed to - 20 spread spec~rum signals transmitted over a radio communicatiion d~annel. However, as will be understood by those sldlled in the art, the communication channel could altematively be an electronic data bus, wireline, optical tiber link, or any other type of communication channel.

.

Claims

Claims What is claimed is:

1. An apparatus comprising a spread-spectrum noise canceller, the spread-spectrum noise canceller comprising:
(a) determining means for determining a received phase and a received amplitude for a first and a second component of a received spread-spectrum signal, the first component being different from the second component, the spread-spectrum signal including a first and a second known signal; and (b) noise canceling means, operatively coupled to the determining means, for canceling a portion of a spread-spectrum noise signal in the received spread-spectrum signal by:
(i) generating an estimated signal by spreading the second known signal at the second component received phase with the first known signal at the first component received phase and adjusting a gain of an integrated form of the spread second known signal as a function of the received amplitudes of the first and the second components; and (ii) processing the second known signal out of the received spread-spectrum signal by subtracting the estimated signal from a demodulated form of the received spread-spectrum signal.

2. The apparatus of claim 1 wherein the spread-spectrum noise canceller canceling means processes the second known signal out of the received spread-spectrum signal only if the adjusted gain of the integrated form of the spread second known signal is greater than a predetermined threshold.

3. The apparatus of claim 1 wherein the spread-spectrum noise canceller canceling means generates the estimated signal by further spreading the second known signal at the second component received phase with a channel selecting spreading code at the first component receive phase.

4. The apparatus of claim 1 further comprising:
(a) descrambling means, operatively coupled to the spread-spectrum noise canceller, for descrambling the processed demodulated form of the received spread-spectrum signal by utilizing a known spreading code;
(b) deinterleaving means, operatively coupled to the descrambling means, for deinterleaving the descrambled received spread-spectrum signal within a predetermined size block; and (c) decoding means, operatively coupled to the deinterleaving means, for generating at least one estimated data bit by utilizing maximum likelihood decoding techniques to derive the at least one estimated data bit from the deinterleaved received spread-spectrum signal.

5. A spread-spectrum signal processing method, comprising:
(a) receiving a spread-spectrum signal having a first and a second component from over a communication channel, the first component being different from the second component, the spread-spectrum signal including a first and a second known signal;
(b) determining a received phase and a received amplitude for the first and the second components of a received spread-spectrum signal;
(c) demodulating the received spread-spectrum signal; and (d) canceling a portion of a spread-spectrum noise signal in the received spread-spectrum signal by:
(i) generating an estimated signal by spreading the second known signal at the second component received phase with the first known signal at the first component received phase, integrating the spread second known signal over a predetermined time, and adjusting a gain of the integrated spread second known signal as a function of the received amplitudes of the first and the second components;
and (ii) processing the second known signal out of the received spread-spectrum signal by subtracting the estimated signal from the demodulated received spread-spectrum signal.

6. The method of claim 5 wherein the step of canceling comprises processing the second known signal out of the received spread-spectrum signal only if the adjusted gain of the integrated form of the spread second known signal is greater than a predetermined threshold.

7. The method of claim 5 wherein the step of canceling comprises generating the estimated signal by further spreading the spread known signal at the second component received phase with a channel selecting spreading code at the first component received phase.

8. The method of claim 5 further comprising:
(a) descrambling the processed demodulated form of the received spread spectrum signal by utilizing a known spreading code;
(b) deinterleaving the descrambled received spread-spectrum signal within a predetermined size block; and (c) generating at least one estimated data bit by utilizing maximum likelihood decoding techniques to derive the at least one estimated data bit from the deinterleaved received spread-spectrum signal.