United States Patent [19]
`Fattouche et a1.
`
`IIlllllllllllllIll!llllllllLwltllllllllllllllllllllllllllllllllll
`
`[11] Patent Number:
`[45] Date of Patent:
`
`5,555,268
`Sep. 10, 1996
`
`1541
`
`MULTICODE DIRECT SEQUENCE SPREAD
`SPECTRUM
`
`[751
`
`[211
`122}
`[511
`1521
`
`Inventors: Michel Fattouehe, 156 Hawkwood
`Blvd N.W., Calgary, Alberta, Canada,
`T3G 2T2; Hatim Zaghloul, 402 a 1st
`Ave. NE, Calgary, Alberta, Canada,
`T213 034
`
`Appl. No.: 186,784
`Filed:
`Jan. 24, 1994
`
`Int. Cl.6 ............................ .. H04B 1/707; H0413 1/69
`US. Cl. .......................... ._ 375/206; 375/200; 380/34;
`370/19; 370/21; 370/22
`Field of Search .............................. .. 375/1, 200-210;
`380/34, 46; 370/18, 19, 20, 21, 22, 23,
`24; 331/78; 364/717
`
`[561
`
`References Cited
`
`U.S. PATENT DUCUMENTS
`
`4,928,310
`4,933,952
`4,944,009
`
`5/1990 Goutzoulis et a1. .................... .. 380/46
`375/1
`6/1990 Albrieux et a1.
`'
`7/1990 Micali et al.
`.. 380/46
`
`5,235,614
`8/1993 Bruckertet a1. .......................... .. 375/1
`Primary Examiner—Bernarr E. Gregory
`Attorney, Agent, or Firm-Frank J. Dykas
`[57]
`ABSTRACT
`In this patent, we present MultiCode Direct Sequence
`Spread Spectrum (MODSSS) which is a modulation
`scheme that assigns up to N DSSS codes to an individual
`user where N is the number of chips per DSSS code. When
`viewed as DSSS, MC-DSSS requires up to N correlators (or
`equivalently up to N Matched Filters) at the receiver with a
`complexity of the order of N2 operations. In addition, a non
`ideal communication channel can cause InterCode Interfer
`ence (ICI), i.e., interference between the N DSSS codes. In
`this patent, we introduce new DSSS codes, which we refer
`to as the “MC” codes. Such codes allow the information in
`a MC-DSSS signal to be decoded in a sequence of low
`complexity parallel operations which reduce the ICI. In
`addition to low complexity decoding and reduced 1C1.
`MC-DSSS using the MC codes has the following advan
`tages: (1) it does not require the stringent synchronization
`DSSS requires, (2) it does not require the stringent carrier
`recovery DSSS requires and (3) it is spectrally ei?cient.
`
`f3O
`
`32 Claims, 20 Drawing Sheets
`
`(32 x<1J<> J“,
`
`___> MC DSSS
`ZFHORDER
`A ING
`‘
`—-> SH ‘P -> TIME
`sym?g) TRANSMITTER dam
`Wk) DIVERSITY
`
`:
`-
`
`; ?g
`nu, __>
`is‘;
`X(k)
`<1
`
`Q4
`
`ERIC-1005
`Ericsson v. IV
`Page 1 of 26
`
`

`
`U.S. Patent
`
`Sep. 10, 1996
`
`Sheet 1 0f 20
`
`5,555,268
`
`'ZVDIHS
`
`@ziw
`
`Page 2 of 26
`
`

`
`US. Patent
`
`Sep. 10, 1996
`
`Sheet 2 of 20
`
`5,555,268
`
`Q58
`Alan!
`
`HGlDHlHG
`
`N HMDUH h
`
`Page 3 of 26
`
`

`
`U.S. Patent
`
`Sep. 16, 1996
`
`Sheet 3 of 20
`
`5,555,268
`
`mmMV.
`
`V.3V11Hn.in_u
`
`U
`
`
`
`
`.Q$..u.~2mommz<.m.i._2mo%z<.meN...:2,mo%z§:._..
`
`
`
`..,:,mom..2.5N5....:,=o._-z._ezwonvzH
`
`.3n.5mzeumm.5.NEma..
`
`
`
`
`
`mofimazoumm.mufimemmfi
`
`Page 4 of 26
`
`

`
`U.S. Patent
`
`Sep. 10, 1996
`
`Sheet 4 of 20
`
`5,555,268
`
`._.uU5HVWU31l0mMV
`
`I/‘\I§iO:iSNV}§i
`.LNIOd'
`‘I37
`
`mm/«em:5
`
`I/\?I0:ISNV’8i.
`'
`.I.NIOd' N
`GNOQEIS
`
`I/\i‘€{O:1SNVH.L
`.LNIOd‘ N
`._LSH§:i
`
`E{TIV}IVcI‘OJ:"1VIH3 ‘
`
`wmmnwnm
`
`Page 5 of 26
`
`

`
`U.S. Patent
`
`Sep. 10, 1996
`
`Sheet 5 of 20
`
`5,555,268
`
`LTHESTIMATO'ESTIMATO’
`LTHPILOTSECOND_—————>
`
`
`SECONDPILO
`
`FIRSTESTIMATOR
`
`FIRSTPILOT
`
`H':!2'i’IVflDE:'
`.1.m0a- N
`mt:
`
`aaz1'wnDa
`m10&- N
`.1S‘$:1
`
`I/\IHOdSNV‘d.L
`HSHEIANI
`
`LthSTAGE
`
`$
`:55
`*0
`D
`Z
`8
`m
`U)
`
`
`
`FIRSTSTAGE
`
`Ln
`an
`a:
`D
`(5
`H
`in
`
`Page 6 of 26
`
`

`
`5,555,268
`
`FIGURE 6
`
`US. Patent
`
`Sep. 10, 1996
`
`Sheet 6 of 20
`
`)
`
`N V
`
`n+2“)
`__+> x ,
`
`92 ( x(1,k ZT H ORDER ) TIME
`SHAPING % v vac) DIVERSITY
`MC-DSSS TRANSMITTER d(k
`
`
`
`f :30
`
`------> S ym(k)
`
`Page 7 of 26
`
`

`
`US. Patent
`
`Sep. 10, 1996
`
`Sheet 7 of 20
`
`5,555,268
`
`8:08
`MEBMUQMAI!
`wmmnTUE
`UZE<EMZD
`
`3%
`
`g;
`
`h WMDUHb
`
`'IHTIVHVd '
`
`oxwma
`
`'HDNAS
`HINVHd
`
`Page 8 of 26
`
`

`
`US. Patent
`
`Sep. 10, 1996
`
`Sheet 8 of 20
`
`5,555,268
`
`
`Amy @353 a ANVSQEQQ
`SW A 0‘
`
`
`
`Qébagg @SEQEQ
`
`‘ A A
`
`w WMDGHM
`
`
`EMOHHWZ<E .HZHOL 2
`
`2:353 QKREE
`
`
`
`A t WA .1
`2%
`
`Page 9 of 26
`
`

`
`US. Patent
`
`Sep. 10, 1996
`
`Sheet 9 of 20
`
`5,555,268
`
`553%
`
`@SQSQ
`
`gyma?g
`
`AZKHEE
`
`‘'0
`
`III
`
`Cy BQEE.
`
`
`
`EMOQmZAQMF PZHOQ 2
`
`JUMREE
`
`m HmDUHh
`
`Page 10 of 26
`
`

`
`US. Patent
`
`Sep. 10, 1996
`
`Sheet 10 of 20
`
`5,555,268
`
`mmEE
`5% amaze 5% g.“ Q:
`
`
`
`A zoE<26ma A.|iw almwmw??ma All!
`
`3
`
`OH MMDQHh
`
`3V
`
`
`5 mmo?omv 55E
`
`
`
`All 882; Aulll zoi?o??é All...’
`
`
`
`Q: 23 3% $50 E: 3%
`
`Page 11 of 26
`
`

`
`US. Patent
`
`Sep. 10, 1996
`
`Sheet 11 of 20
`
`5,555,268
`
`+ ‘0- MNTU +3) ~*
`
`/
`f
`
`l
`
`FIGURE 11
`
`Page 12 of 26
`
`

`
`U.S. Patent
`
`Sep. 10, 1996
`
`02fl021tee..nS
`
`5,555,268
`
`NHmmawam
`
`ZOH.H4ZH.Hw.mQ
`
`mMQOZ
`
`mumbom
`
`5%WDOZ
`
`Page 13 of 26
`
`

`
`US. Patent
`
`Sep. 10, 1996
`
`Sheet 13 of 20
`
`5,555,268
`
`Lookup sha in
`
`I'TEIbIC c p g V
`
`FIGURE 13
`
`Page 14 of 26
`
`

`
`US. Patent
`
`Sep. 10, 1996
`
`Sheet 14 0f 20
`
`5,555,268
`
`
`
`imam BMQQAAII
`
`A||
`ESBQQ
`E25
`
`@333:
`
`=6 QQQNB
`
`
`
`“In WmDUHh
`
`Page 15 of 26
`
`

`
`‘ US. Patent
`
`Sep. 10, 1996
`
`Sheet 15 of 20
`
`5,555,268
`
`mimmsw All. mswwowa All
`
`u o m ww3w©<
`
`
`
`
`
`
`
`m H mmDwH .m
`
`
`
`éouguwmw 60:53
`
`
`
`@255 NED mo #
`
`Page 16 of 26
`
`

`
`US. Patent
`
`Sep. 10, 1996
`
`Sheet 16 of 20
`
`5,555,268
`
`WM WMDUHb
`
`AIIQAFIJ
`
`£23m
`662m F
`
`Page 17 of 26
`
`

`
`U.S. Patent
`
`Sep. 10, 1996
`
`Sheet 17 0f 20
`
`5,555,268
`
`FIGURE 17
`
`Page 18 of 26
`
`

`
`US. Patent
`
`Sep. 10, 1996
`
`Sheet 18 0f 20
`
`5,555,268
`
`
`
`All 165 BE“ QAIQAIIwEQEWHB
`
`2%‘ .SwQQQQ _ + a‘
`
`
`
`
`
`
`
`
`
`w." WMDUHRH
`
`Page 19 of 26
`
`

`
`Sep. 10, 1996
`
`Sheet 19 of 20
`
`5,555,268
`
`US. Patent
`
`855mg:
`wugmwmm
`
`Kiowa
`
`‘11...!
`
`d @2298
`
`mmull-unonnnmmnva.
`
`
`
`@235 Rama 13m 5360B
`
`
`
`
`
`m H HMDU H .m
`
`
`
`m @othNUwOo OmWKNOQKQ
`AlaWAi .?wpmwmwmo . ESE AA! 13% TWA‘ 3x23
`
`
`
`
`
`
`SASMEQ wmwmwwmmm 3GP BEEES:
`d 950% ii
`
`*
`
`
`
`. . E8233
`
`Page 20 of 26
`
`

`
`US. Patent
`
`Sep. 10, 1996
`
`Sheet 20 of 20
`
`5,555,268
`
`5380.0NRumEEGmo_§<M
`
`303
`
`
`wcmnommn?.w.¢..w.oH.a:<mwmun%uw~-%.wmAmmmmucw._umwnwmmm
`wmfiuoomHmwfimmwnwvoBEE52$@0388
`
`
`
`
`M62203wcsobmExomm
`
`Owamomoo
`
`Page 21 of 26
`
`

`
`5,555,268
`
`1
`MULTICODE DIRECT SEQUENCE SPREAD
`SPECTRUM
`
`FIELD OF THE INVENTION
`
`The invention deals with the field of multiple access
`communications using Spread Spectrum modulation. Mul-
`tiple access can be classified as either random access,
`polling, TDMA, FDMA, CDMA or any combination
`thereof. Spread Spectrum can be classified as Direct
`Sequence, Frequency-Hopping or a combination of the two.
`
`BACKGROUND OF THE INVENTION
`
`Commonly used spread spectrum techniques are Direct
`Sequence Spread Spectrum (DSSS) and Code Division
`Multiple Access (CDMA) as explained in Chapter 8 of
`“Digital Communication” by J. G. Proakis, Second Edition,
`l99l, McGraw Hill, DSSS is a communication scheme in
`which information bits are spread over code bits (generally
`called chips). It is customary to use noise-like codes called
`pseudo random noise (PN) sequences. These PN sequences
`have the property that their auto-correlation is almost a delta
`function and their cross-correlation with other codes is
`almost null. The advantages of this information spreading
`are:
`
`l. The transmitted signal can be buried in noise and thus
`has a low probability of intercept.
`_2. The receiver can recover the signal from intcrfercrs
`(such as other transmitted codes) with a jamming
`margin that
`is proportional
`to the spreading code
`length.
`3. DSSS codes of duration longer than the delay spread of
`the propagation channel can lead to multipath diversity
`implementable using a Rake receiver.
`4. The FCC and the DOC have allowed the use of
`unlicensed low power DSSS systems of code lengths
`greater than or equal to 10 in some frequency bands
`(the ISM bands).
`It is the last advantage (i.e., advantage 4. above) that has
`given much interest recently to DSSS.
`An obvious limitation of DSSS systems is the limited
`throughput they can offer. In any given bandwidth, B, a code
`of length N will reduce the eifective bandwidth to B/N. To
`increase the overall bandwidth efiiciency, system designers
`introduced Code Division Multiple Access (CDMA) where
`multiple DSSS communication links can be established
`simultaneously over the same frequency band provided each
`link uses a unique code that is noise-like. CDMA problems
`are:
`
`l. The near—far problem: a transmitter “near” the receiver
`sending a different code than the receiver's desired
`code produces in the receiver a signal comparable with
`that of a “ ar” transmitter sending the desired code.
`2. Synchronization of the receiver and the transmitter is
`complex (especially) if the receiver does not know in
`advance which code is being transmitted.
`
`SUMMARY OF THE INVENTION
`
`We have recognized that low power DSSS systems corn-
`plying with the FCC and the DOC regulations for the ISM
`bands would be ideal communicators provided the problems
`of CDMA could be resolved and the throughput could be
`enhanced. To enhance the throughput, we allow a single link
`(i.c., a single transceiver) to use more than one code at the
`
`2.
`same time. To avoid the near—far problem only one trans-
`ceiver transmits at a time. In this patent, we present Multi-
`Code Direct Sequence Spread Spectrum (MC-DSSS) which
`is a modulation scheme that assigns up to N codes to an
`individual transceiver where N is the number or” chips per
`DSSS code. When viewed as DSSS, MC-DSSS requires up
`to N correlators (or equivalently up to N Matched Filters) at
`the receiver with a complexity of the order of N2 operations.
`When N is large, this complexity is prohibitive. In addition,
`a nonideal communication channel can cause lnterCode
`Interference (ICI), i.c,, interference between the N DSSS
`codes at the receiver. In this patent, we introduce new codes,
`which we refer to as “MC” codes. Such codes allow the
`information in a MC-DSSS signal
`to be decoded in a
`sequence of low complexity parallel operations while reduc-
`ing the ICI. In addition to low complexity decoding and ICI
`reduction, our implementation of MC-DSSS using the MC
`codes has the following advantages:
`l. It does not require the stringent synchronization DSSS
`requires. Conventional DSSS systems requires syn-
`chronization to within a fraction of a chip whereas
`MC-DSSS using the MC codes requires synchroniza-
`tion to within two chips.
`2. It does not require the stringent carrier recovery DSSS
`requires. Conventional DSSS requires the carrier at the
`receiver to be phase locked to the received signal
`whereas MC-DSSS using the MC codes does not
`require phase locking the carriers. Commercially avail-
`able crystals have sufiicient stability for MC-DSSS.
`3. It is spectrally eflicicnt.
`
`BRIEF DESCRIPTION OF THE DRAWINGS
`
`FIG. 1 is a schematic showing for the Bascband Trans-
`mitter fer the xth MC-DSSS frame: d(k)=[d(l,x) d(2,x) .
`.
`.
`d(.\I,k)] where c(i)=[c(l ,i) c(2,i)] is the ith code and Sym(k)=
`[sym(l,k) sym(N,k)] is the kth information-bearing vector
`containing N symbols.
`FIG. 2 is a schematic showing a Baseband Receiver for
`the kth received MC-DSSS frame: d'(k)=[d‘(l,k) cl'(2,k) .
`.
`.
`d‘(N,k)] where c(i)=[c(l,i) c(2,i) .
`.
`. c(N,i)] is the ith code,
`SyrhCk)={sym(l,k) syri1(2,k) .
`.
`. syri1(N,k)] is the estimate of
`the Kth information-bearing vector Sym(k) and
`
`a"(k) ~;1Q——> is a dot product defined as
`c(i)
`d’(k)G ch) = c(l,i)d’(l,k) = c(2,i)d'(2,k) +. .
`
`. + c(N.i)d'(N,N).
`
`FIG. 3 is a schematic showing of the ith MC code
`c(i)=[c(i,l) c(i,2) .
`.
`. c(i,NO) where i can take one of the N
`values: 1,2, .
`.
`. N corresponding to the position of the single
`‘1’ at the input of the first N-point transform.
`FIG. 4 is a schematic showing the alternate transmitter for
`the kth MC-DSSS frame: d(k)=[d(l,k), d(2,k) .
`.
`. d(N,k)]
`using the MC codes generated in FIG. 3 where Sym(l<)=
`[Sym(1,k)Sym(2k) .
`.
`. Sym(N,k)] is the kth information-
`bcaring vector contacting N symbols.
`FIG. 5 is the alternate receiver for the kth received
`MC-DSSS frame d’(k)=[d'(lk)d'(2,K) .
`. .d'(N,k)] using MC
`codes generated in FIG. 3 where Syri1(lc)=[syr’h(l,l<) sym(2l<)
`.
`. sy1i1(N,l<)] is the estimate of the information-bearing
`vetor Sym(k).
`FIG. 6 is a schematic showing the Baseband Transmitter
`of the kth Data Frame X(l<) where Sym(N)=[sym(l,k)
`
`Page 22 of 26
`
`

`
`5,555,268
`
`3
`. syrn(N,k)] is the kth information-bearing vector
`.
`sym(2,k) .
`d(k)=[e(l,k) d(2,k) .
`.
`. d(N,k)] is the kth MC-DSSS frame
`v(k)=[v(l,k) v(2,k) .
`.
`. v((l+[3)MN,k)], [5e(0,l), M=l,2,3 .
`.
`. and X(k)=[x(lk) x(2,k)], Z=Z=l, 2, 3, .
`.
`.
`.
`FIG. 7 is a schematic showing the Baseband Receiver for
`the kth received Data Frame X'(k) where Syrf1(N)=
`[syrii(l,k)] syri1(2,k) .
`.
`. syri1(N,k)] is the estimate of the kth
`information-bearing vector d'(k)=[d'(1,l-:) d'(2k) .
`.
`. d'(N,K)]
`is the kth received MC-DSSS frame v'(k)=[v'(l,k) v(2k) .
`.
`. v'((l+B) MN,k)], Be(O,1), M=l,2,3, .
`.
`. and X'(l<)=[x'(1,k)
`x'(2,k) .
`.
`. r'(Z,k)], Z=1,2,3, .
`.
`.
`.
`FIG. Sis a schematic showing the Randomizer Transform
`(RT) where a (1) a (2) .
`.
`. a (N) are complex constants
`chosen randomly.
`FIG. 9 is a schematic showing the Permutation Transform
`(PT).
`FIG. 10 is a schematic showing (a) the shaping of a
`MC-DSSS frame and (b) the unshaping of a MC-DSSS
`frame where d(k):[d(l,k) d(2,k) .
`.
`. d(N,k)j
`is the kth
`.
`MC-DSSS frame g(k)=[g(1,k) g(2k) .
`. g(MN,k)], M=1,2,3,
`.
`.
`.
`, v(k)=[v(1,k) v(2,k) .
`.
`. v((l+B) MN,k)l, Be(0,1)
`d'(k):[d(l,k) d(2,k) .
`.
`. d(N,K)] is the kth received MC-
`DSSS frame g'(k)=[g'(l,k) g‘(2,k) .
`.
`. g'(M'N,k)] and v'(k)=
`fv(1,k) v'(2,k) .
`.
`. v‘((1+B) M'N,k)], M‘=l,2,3, .
`.
`.
`FIG. 11 is a schematic showing (a) Description of the
`alias/window operation (b) Description of dealias/dewindow
`operation, where 1/1‘ is the symbol rate.
`FIG. 12 is a schematic showing the frame structure for
`data transmission from source (Node A) to destination
`(Node B).
`FIG. 13 is a schematic showing the baseband transmitter
`for one request frame v where c:[c(l) c(2) .
`.
`. c(1)] is the
`DSSS code, v=[v(1) v(2) .
`.
`. v((l+[3)MI)], l3e(0,l), M=l,2,
`.
`.
`. and I is the length of the DSSS code.
`FIG. 14 is a schematic showing the baseband receiver for
`the received request frame where c=[c(1) c(2) .
`.
`. c(1)] is the
`DSSS code for the request frame, d':[d(1) d(2) .
`.
`. d(1)] is
`the received request frame, v'=[v'(1) v'((l+[3) MI)], [3e(0,l),
`M=l,2, .
`.
`. and l is the length of the DSSS code.
`FIG. 15 is a schematic showing the baseband transmitter
`for one address frame where <,=[c(1) c(2) .
`.
`. c(1)] is the
`CDMA code for the address frame, v=[v(l) v(2) .
`.
`. v(l+B)
`MD], [3e(0,l), M=l,2, .
`.
`. and 1' is the length of the CDMA
`code.
`
`FIG. 16 is a schematic showing the baseband receiver the
`address where <‘:[c(1) c(2) .
`.
`. c(I')] is the CDMA code for
`the address frame, d'=[d(1) d(2) .
`.
`. d(I)] is the received
`address frame, v‘=[v’(1) v’(2) .
`.
`. v'((l+[3) MI')], [5e(0,l),
`M=l,2, .
`.
`. and I‘ is the length of the CDMA code.
`FIG. 17 is a schematic showing the baseband transmitter
`for Ack. Frame where c=[c(1) c(2) .
`.
`. c(I')] is the DSSS
`code for the Ack. frame, v=[v(1) v(2) .
`.
`. v((I+B) MI')]
`Be(0,l), M=l,2,3, .
`.
`. and I‘ is the length of the DSSS code.
`FIG. 18 is a schematic showing the baseband receiver for
`the ack. frame where c=[c(l) c(2) .
`.
`. c(I")] is the DSSS code
`.
`for the Ack. frame, d'=[d(1) d(2) .
`. d'(I")] is the received
`Ack. frame, v'=[v'(1) v(2) .
`.
`. v'(1+B) MI")], Be(0,l), M=1,2,
`.
`.
`. and l" is the length of the DSSS code.
`FIG. 19 is a schematic showing the passband transmitter
`for a packet where fa is the IF frequency and l'0+l‘c is the RF
`frequency.
`FIG. 20 is a schematic showing the passband receiver for
`a packet where fa is the IF frequency and f,,+fc is the RF
`frequency.
`
`4
`DETAILED DESCRIPTION OF THE
`PREFERRED EMBODIMENTS OF THE
`INVENTION
`
`FIG. 1 illustrates the transmitter of the MC-DSSS modu-
`lation technique generating the kth MC-DSSS frame bearing
`N symbols of information. The symbols can be either analog
`or digital.
`A converter 10 converts a stream of data symbols into
`plural sets of N data symbols each. A computing means 12
`operates on the plural sets of N data symbols to produce
`modulated data symbols corresponding to an invertible
`randomized spreading of the stream of data symbols. A
`combiner 14 combines the modulated data symbols for
`transmission. The computing means shown in FIG. 1
`includes a source 16 of N direct sequence spread spectrum
`code symbols and a modulator 18 to modulate each ith data
`symbol from each set of N data symbols with the 1 code
`symbol from the N code symbolto generate N modulated
`data symbols, and thereby spread each I data symbol over a
`separate code symbol.
`FIG. 2 illustrates the receiver of the MC-DSSS modula-
`tion techniques accepting the kth MC-DSSS frame and
`generating estimates for the corresponding N symbols of
`information. The dot product in FIG. 2 can be implemented
`as a correlator. The detector can make either hard decisions
`or soft decisions.
`
`A sequence of modulated data symbols is received at 22
`in which the sequence of modulated data symbols has been
`generated by the transmitter such as is shown in FIG. 1 or
`4. A second computing means 24 operates on the sequence
`of modulated data symbols to produce an estimate of the
`second string of data symbols. The computing means 24
`shown in FIG. 2 includes a correlator 26 for correlating each
`I modulated data symbol from the received sequence of
`modulated data symbols with the I code symbol from the set
`of N code symbols and a detector 28 for detecting an
`estimate of the data symbols from output of the correlator
`26.
`
`FIG. 3 illustrates the code generator of the MC codes. Any
`one of the P N-point transforms in FIG. 3 consists of a
`reversible transform to the extent of the available arithmetic
`precision. In other words, with finite precision arithmetic,
`the transforms are allowed to add a limited amount of
`irreversible error.
`One can use the MC-DSSS transmitter in FIG. 1 and the
`MC-DSSS receiver in FIG. 2 together with the MC codes
`generated using the code generator in FIG. 3 in order to
`implement MC-DSSS using the MC codes.
`An alternative transmitter to the one in FIG. 1 using the
`MC codes in FIG. 3 is shown in FIG. 4.
`The alternative transmitter shown in FIG. 4 includes a
`transformer 20 for operating on each set of N data symbols
`to generate N modulated data symbols as output. A series of
`transforms are shown.
`
`An alternative receiver to the one in FIG. 2 using the MC
`codes in FIG. 3 is shown in FIG. 5. L pilots are required in
`FIG. 5 for equalization.
`Both transmitters in FIGS. 1 and 4 allow using shaper 30
`in diversity module 32 shaping and time diversity of the
`MC-DSSS signal as shown in FIG. 6. We will refer to the
`MC-DSSS frame with shaping and time diversity as a Data
`frame.
`
`Both receivers in FIGS. 2 and 5 allow diversity combining
`followed by the unshaping of the Data frame as shown in
`FIG. 7. A Synch. is required in FIG. 7 for frame synchro-
`nization.
`
`Page 23 of 26
`
`

`
`5,555,268
`
`5
`In addition to the Data frames, we need to transmit (1) all
`of the L pilots used in FIG. 5 to estimate and equalize for the
`various types of channel distortions, (2) the Synch. signal
`used in FIG. 7 for frame synchronization, and (3) depending
`on the access technique employed,
`the source address,
`destination address and number of Data frames. We will
`refer to the combination of all
`transmitted frames as a
`packet.
`
`PREFERRED EMBODIMENTS OF THE
`INVENTION
`
`transforms in FIG. 3 are a
`Examples of the N—point
`Discrete Fourier Transform (DFT), a Fast Fourier Transform
`(FFT), :1 Walsh Transform (WT), a Hilbert Transform (HT),
`21 Randornizer Transform (RT) as the one illustrated in FIG.
`8, a Permutator Transform (PT) as the one illustrated in FIG.
`9, an Inverse DFT (IDFT), an Inverse FFT (IFFT), an
`Inverse WT (IWT), an Inverse HT (IHT), an Inverse RT
`(IRT), an Inverse PT (IPT), and any other reversible trans-
`form. When L=2 with the first N-point transform being a
`DFF and the second being a RT, we have a system identical
`to the patent: “Method and Apparatus for Multiple Access
`between Transceivers in Wireless Communications using
`OFDM Spread Spectrum” by M. Fattouche and H. Zaghloul,
`filed in the U.S. Pat. Oliice in Mar. 31, 1992, Ser. No.
`O7/86l ,725.
`Preferred shaping in FIG. 6 consists of an Mth order
`interpolation filter followed by an alias/window operation as
`shown in FIG. 10a. The Alias/window operation is described
`in FIG. 11a where a raised-cosine pulse of rolloff B is
`applied. The interpolation filter in FIG. 10:: can be imple-
`mented as an FIR filter or as an NM-point IDFT where the
`first N(M—l)/2 points and the last N(M-1)/2 points at the
`input of the IDFF are zero. Preferred values of M are 1,2,3
`and 4.
`
`Preferred unshaping in FIG. 7 consists of a dealiasl
`dewindow operation followed by a decimation filter as
`shown in FIG. 1%. The dealias/dewindow operation is
`described in FIG. 113:.
`
`Time Diversity in FIG. 6 can consist of repeating the
`MC-DSSS frame several times. It can also consist of repeat-
`ing the frame several times then complex conjugating some
`of the replicas, or shifting some of the replicas in the
`frequency domain in a cyclic manner.
`Diversity combining in FIG. 7 can consist of cophasing,
`selective combining, Maximal Ratio combining or equal
`gain combining.
`In FIG. 5, L pilots are used to equalize the effects of the
`channel on each information-bearing data frame. The pilot
`frames can consist of Data frames of known information
`symbols to be sent either before, during or after the data, or
`of a number of samples of known values inserted within two
`transformations in FIG. 4. A preferred embodiment of the
`pilots is to have the first pilot consisting of a number of
`frames of known information symbols. The remaining pilots
`can consist of a number of known information symbols
`between two transforms. The L estimators can consist of
`averaging of the pilots followed by either a parametric
`estimation or a nonparametric one similar to the channel
`estimator in the patent: “Method and Apparatus for Multiple
`Access between Transceivers in Wireless Communications
`using OFDM Spread Spectrum" by M. Fattouche and H.
`Zaghloul, filed in the U.S. Pat. Oflice in Mar. 31, I992, Ser.
`No. 07/861,725.
`When Node A intends to transmit information to Node B,
`a preferred embodiment of a packet is illustrated in FIG. 12:
`
`6
`a Request frame 40, anAddress frame, an Ack. frame, a Pilot
`frame 36 and a number of Data frames 38. The Request
`frame is used (I) as a wake-up call for all the receivers in the
`band, (2) for frame synchronization and (3) for packet
`synchronization. It can consist of a DSSS signal using one
`PN code repeated a number of times and ending with the
`same PN code with a negative polarity. FIGS. 13 and 14
`illustrate the transmitter and the receiver for the Request
`frame respectively. In FIG. 14, the dot product operation can
`be implemented as a correlator with either hard or soft
`decision (or equivalently as a filter matched to the PN code
`followed by a sample/hold circuit). The Request frame
`receiver is constantly generating a signal out of the correla-
`tor. When the signal is above a certain threshold using the
`level detector, (l) a wake-up call signal is conveyed to the
`portion of the receiver responsible for the Address frame and
`(2) the frames are synchronized to the wake-up call. The
`packet
`is then synchronized to the negative differential
`correlation between the last two PN codes in the Request
`frame using a decoder as shown in FIG. 14.
`The Address frame can consist of a CDMA signal where
`one out of a number of codes is used at a time. The code
`consists of a number of chips that indicate the destination
`address,
`the source address and/or the number of Data
`frames. FIGS. 15 and I6 illustrate the transmitter and the
`receiver for the Address frame respectively. Each receiver
`differentially detects the received Address frame, then cor-
`relates the outcome with it is own code. If the output of the
`coirelator is above a certain threshold, the receiver instructs
`its transmitter to transmit an Ack. Otherwise, the receiver
`returns to its initial (idle) state.
`The Ack. frame is a PN code reflecting the status of the
`receiver, i.e. whether it is busy or idle. When it is busy, Node
`A aborts its transmission and retries some time later. When
`it is idle, Node A proceeds with transmitting the Pilot frame
`and the Data frames. FIGS. 17 and 18 illustrate the trans-
`mitter and the receiver for the Address frame respectively.
`An extension to the MC-DSSS modulation technique
`consists of passband modulation where the packet is up-
`converted from baseband to RF in the transmitter and later
`down-converted from RF to baseband in the receiver. Pass-
`band modulation can be implemented using IF sampling
`which consists of implementing quadrature modulationl
`demodulation in an intermediate Frequency between base-
`band and RF, digitally as shown in FIGS. 19 and 20 which
`illustrate the transmitter and the receiver respectively. IF
`sampling trades complexity of the analog RF components (at
`either the transmitter, the receiver or both) with complexity
`of the digital components. Furthermore, in passband systems
`carrier feed-through is often a problem implying that the
`transmitter has to ensure a zero dc component. Such a
`component reduces the usable bandwidth of the channel. In
`IF sampling the usable band of the channel does not include
`do and therefore is the dc component is not a concern.
`A further extension to the MC-DSSS modulation tech-
`nique consists of using antenna Diversity in order to
`improve the Signal-to-Ratio level at the receiver. Apreferred
`combining technique is maximal selection combining based
`on the level of the Request frame at the receiver.
`We claim:
`I. A transceiver for transmitting a first stream of data
`symbols, the transceiver comprising:
`a converter for converting the first stream of data symbols
`into plural sets of N data symbols each;
`first computing means for operating on the plural sets of
`N data symbols to produce modulated data symbols
`
`Page 24 of 26
`
`

`
`5,555,268
`
`7
`corresponding to an invertible randomized spreading of
`the first stream of data symbols; and
`means to combine the modulated data symbols for trans-
`mission.
`2. The transceiver of claim 1 in which the first computing
`means includes:
`
`5
`
`a source of N direct sequence spread spectrum code
`symbols; and
`a modulator to modulate each ith data symbol from each
`set of N data symbols with the ith code symbol from the
`N code symbol to generate N modulated data symbols,
`and thereby spread each ith data symbol over a separate
`code symbol.
`3. The transceiver of claim 2 in which the code symbols
`are generated by operation of a non-trivial N point transform
`on a sequence of input signals.
`4. The transceiver of claim 1 in which the first computing
`means includes:
`
`a transformer for operating on each set of N data symbols
`to generate N modulated data symbols as output, the N
`modulated data symbols corresponding to spreading of
`each ith data symbol over a separate code symbol.
`5. The transceiver of claim 4 in which the transformer
`effectively applies a first transform selected from the group
`comprising a Fourier transform and a Walsh transform to the
`N data symbols.
`6. The transceiver of claim 5 in which the first transform
`is a Fourier transform and it is followed by a randomizing
`transform.
`7. The transceiver of claim 6 in which the first transform
`is a Fourier transform and it is followed by a randomizing
`transform and a second transform selected from the group
`comprising a Fourier transform and Walsh transform.
`8. The transceiver of claim 4 in which the transformer
`effectively applies a first inverse transform selected from the
`group comprising a randomizer transform, a Fourier trans-
`form and a Walsh transform to the N data symbols, followed
`by a first equalizer and a second inverse transform selected
`from the group comprising a Fourier transform and a Walsh
`transform.
`9. The transceiver of claim 8 in which the second trans-
`form is followed by a second equalizer.
`10. The transceiver of claim 1 further including:
`means for receiving a sequence of modulated data sym-
`bols, the modulated data symbols having been gener-
`ated by invertible randornized spreading of a second
`stream of data symbols
`second computing means for operating on the sequence of
`modulated data symbols to produce an estimate of the
`second stream of data symbols.
`11. The transceiver of claim 10 further including means to
`apply diversity to the modulated data symbols before trans-
`mission, and means to combine received diversity signals.
`12. The transceiver of claim 10 in which the second
`computing means includes:
`a corrclator for correlating each ith modulated data sym-
`bol from the received sequence of modulated data
`symbols with the ith code symbol from the set of N
`code symbols; and
`a detector for detecting an estimate of the data symbols
`from output of the corrclator.
`13. The transceiver of claim 10 in which the second
`computing means includes an inverse transformer for regen-
`erating an estimate of the N data symbols.
`14. The transceiver of claim 1 f11rthcr including a shaper
`for shaping the combined modulated data symbols for
`transmission.
`
`8
`15. The transceiver of claim 1 further including means to
`apply diversity to the combined modulated data symbols
`before transmission.
`16. The transceiver of claim 1 in which the N data
`symbols include a pilot frame and a number of data frames,
`and is preceded by a request frame, wherein the request
`frame is used to wake up receiving transceivers, synchronize
`reception of the N data symbols and convey protocol infor-
`mation.
`17. A transceiver for transmitting a first stream of data
`symbols and receiving a second stream of data symbols, the
`transceiver comprising:
`a converter for converting the first stream of data symbols
`into plural sets of N data symbols each;
`first computing means for operating on the plural sets of
`N data symbols to produce sets of N modulated data
`symbols corresponding to an invertible randomized
`spreading of each set of N data symbols over N code
`symbols;
`means to combine the modulated data symbols for trans-
`mission;
`means for receiving a sequence of modulated data sym-
`bols, the modulated data symbols having been gener-
`ated by an invertible randorrrized spreading of a second
`stream of data symbols over N code symbols;
`second computing means for operating on the sequence of
`modulated data symbols to produce an estimate of the
`second stream of data symbols; and
`means to combine output from the second computing
`means.
`18. The transceiver of claim 17 in which the first com-
`puting means includes:
`a source of N direct sequence spread spectrum code
`symbols; and
`a modulator to modulate each ith data symbol from each
`set of N data symbols with the ith code symbol from the
`N code symbol to generate N modulated data symbols,
`and thereby spread each ith data symbol over a separate
`code symbol.
`19. The transceiver of claim 18 in which the code symbols
`are generated by operation of plural non-trivial N point
`transforms on a random sequence of input signals.
`20. The transceiver of claim 17 in which the first com-
`puting means includes:
`a transformer for operating on each set of N data symbols
`to generate N modulated data symbols as output, the N
`modulated data symbols corresponding to spreading of
`each ith data symbol over a separate code symbol.
`21. The transceiver of claim 17 in which the second
`computing means includes:
`a correlator for correlating each ith modulated data sym-
`bol from the received sequence of modulated data
`symbols with the ith code symbol from the set of N
`code symbols; and
`a detector for detecting an estimate of the data symbols
`from the output of the corrclator.
`22. The transceiver of claim 17 in which the second
`computing means includes an inverse transformer for regen-
`erating an estimate of the N data symbols.
`23. A method of exchanging data streams between a
`plurality of transceivers, the method comprising the steps of:
`converting a first stream of data symbols into plural sets
`of N data symbols each;
`operating on the plural sets of N data symbols to produce
`modulated data symbols corresponding to a spreading
`
`Page 25 of 26
`
`

`
`5,555,268
`
`9
`of the first stream of data symbols over N code sym-
`hols;
`combining the modulated data symbols for transmission;
`and
`
`transmitting the modulated data symbols from a first
`transceiver at a time when no other of the plurality of
`transceivers is transmitting.
`24. The method of claim 23 in which the spreading is an
`invertible randomized spreading and operating on the plural
`sets of N data symbols includes modulating each ith data
`symbol from each set of N data symbols with the ith code
`symbol from the N code symbols to generate N modulated
`data symbols, and thereby spread each ith data symbol over
`a separate code symbol.
`25. The method of claim 23 in which the spreading is an
`invertible ra