`
`(12) Ulllted States Patent
`Narasimhan et a].
`
`(10) Patent N0.:
`(45) Date of Patent:
`
`US 7,796,681 B1
`*Sep. 14, 2010
`
`(54) MIMO-OFDM RECEIVER PROCESSING
`
`(75) Inventors: Ravi Narasimhan, Los Altos, CA (US);
`Hemanth sampaths Sunnyvale’ CA
`(US); Hsiao-Cheng Tang, Milpitas, CA
`(Us)
`(73) Assjgnee; Marvel] International Limited,
`H am?ton (BM)
`
`( * ) Notice:
`
`Subject to any disclaimer the term of this
`atent is extended or ada'usted under 35
`p
`J
`U'S'C' 15403) by 39 days‘
`_
`_
`_
`_
`_
`Thl_s Patent 15 sublect to a termlnal dls'
`Clalmer-
`
`(21) Appl. No.: 12/179,830
`
`(22) Filed:
`
`Jul. 25, 2008
`
`Related U_s_ Application Data
`_
`_
`_
`_
`(63) COmlnuaUOn 0f apphcatlon NO- 10/912,829, ?led On
`Aug. 5, 2004, HOW Pat. NO. 7,408,976.
`(60) Provisional application No. 60/ 572,934, ?led on May
`19, 2004.
`
`7/2004 Goldstein et a1.
`2004/0136313 A1
`8/2005 Walton et al.
`2005/0180312 A1
`9/2005 Walton et a1~
`2005/0195733 A1
`2006/0252386 A1 11/2006 Boer et al.
`OTHER PUBLICATIONS
`
`“Information
`Society,
`Computer
`IEEE
`TechnologyiTelecommunications and Information Exchange
`Between SystemsiLocal and Metropolitan Area N etworksiSpeci?c
`RequirementsiPart 11.‘ Wireless LAN Medium Access Control
`(MAC) and Physical Layer (PHY) specl?cations ”’ IEEE Std 802'
`lliFirst Edition’ 1999'
`IEEE Computer Society, “Supplement to IEEE Standard for Infor
`mation
`TechnologyiTelecommunications
`and Information
`Exchange Between SystemsiLocal and Metropolitan Area
`NetWorks?Speci?c RequirementsiPart 11.‘ Wireless LAN Medium
`Access Control (MAC) and Physical Layer (PHY) Speci?cations.‘
`High-speed Physical Layer in the 5 GHZ Band”, IEEE Std 802.
`llail999 (Supplement to IEEE Std 802.llil999).
`IEEE Computer Society, “Supplement to IEEE Standard for Infor
`mation
`TechnologyiTelecommunications
`and Information
`Exchange Between SystemsiLocal and Metropolitan Area
`NetWorks?Speci?c RequirementsiPart 11.‘ Wireless LAN Medium
`Access Control (MAC) and Physical Layer (PHY) Speci?cations.‘
`Higher-speed Physical LayerExtension in the 2.4 GHZ Band ”, IEEE
`Std 802.llbil999 (Supplement to IEEE Std 802.llil999).
`(Continued)
`_
`Primary ExamineriKhal Tran
`
`(51) Int. Cl.
`H04B 1/69
`
`(2006.01)
`
`(57)
`
`ABSTRACT
`
`1e
`
`375/148;337755//1l:80
`.....
`.... ... ....
`455/132’
`0 3 721557314 ; as‘); 233124
`’ 45’5/1 0 3 2 4 / 6 14’
`’
`’
`’
`h h, ’
`1
`?l f
`_
`1,
`S
`ee app lcanon e or Comp ete Seam lstory'
`References Cited
`
`(
`
`)
`
`(56)
`
`U.S. PATENT DOCUMENTS
`
`A receiver in a MIMO_OFDM System may process OFDM
`symbols received on a number (MR) of receive antennas. The
`system may utiliZe a MIMO-OFDM frame format that
`includes additional long training OFDM symbols, for train
`ing additional antennas and for link adaptation, and a header
`With an additional SIGNAL symbol to indicate MIMO
`OFDM-speci?c information.
`
`7,463,577 B2 * 12/2008 Sudo et a1. ................ .. 370/208
`
`40 Claims, 8 Drawing Sheets
`
`FIRST TRANSCEIVER
`"TRANSMITTER" 102
`
`110
`H TRANSMIT
`SEC'HON
`T
`112
`\__ RECEIVE
`SECTION
`
`F_.___—_]
`
`Frame
`Formatter
`\
`120
`
`‘i
`
`.
`
`-
`I
`
`i
`
`1 0O
`/
`
`i
`
`.
`
`I
`I
`
`SECOND TRANSCEIVER
`"RECEIVER" 1_0_§
`
`1 ‘IO
`\m TRANSMIT
`SECTION
`
`_
`
`'l ‘I 2
`
`RECEIVE
`SECTION
`
`I__
`Frame
`Formatter
`R
`120
`
`Exhibit 2004
`IPR2015-00221
`Page 1 of 15
`
`
`
`US 7,796,681 B1
`Page 2
`
`OTHER PUBLICATIONS
`
`IEEE Computer Society, “IEEE Standard for Information
`Technology~Draft Supplement to Standard [for] Information
`Technology~Telecommunications and Information Exchange
`Between SystemsiLocal and Metropolitan Area NetworksiSpeci?c
`RequirementsiPart 11.‘ Wireless LAN Medium Access Control
`(MAC) and Physical Layer (PHY) Speci?cations.‘ Further Higher
`Data Rate Extension in the 2.4 GHZ Band”, IEEE P8021 lg/D8.2,
`Apr. 2003.
`
`IEEE Std 802.lla-l999(R2003), Part 11: Wireless LAN Medium
`Access Control (MAC) and Physical Layer (PHY) speci?cations,
`High-speed Physical Layer in the 5 GHZ Band, Jun. 12, 2003, pp.
`7-8.
`van Nee, Richard, A new OFDM standard for high rate Wireless LAN
`in the 5 GHZ band; Vehicular Technology Conference, 1999. VTC
`l999iFall. IEEE VTS 50”‘ vol. 1, Sep. 19-22, 1999 pp. 258-262
`discloses a system With relevance to claims 1-60.
`
`* cited by examiner
`
`Exhibit 2004
`IPR2015-00221
`Page 2 of 15
`
`
`
`U.S. Patent
`
`Sep.14
`
`,2010
`
`Sheet10f8
`
`US 7,796,681 B1
`
`
`
`mm>m0w2<NF0200mm
`
`wow.mewomm,
`
`
`
`mm>mowz<E5mm
`
`
`
`Mod=mmE2mz<mt
`
`oEmE
`
`bmumczou
`
`_‘.O_n_
`
`mEm:u
`
` hut—muteu.
`
`Exhibit 2004
`|PR2015-00221
`Page 3 of 15
`
`Exhibit 2004
`IPR2015-00221
`Page 3 of 15
`
`
`
`
`
`
`U.S. Patent
`
`Sep. 14, 2010
`
`Sheet 2 of 8
`
`US 7,796,681 B1
`
`9N
`
`-wO<&m
`
`”fix/(w.—
`
`MDa.
`
`@2500
`
`mw4m5<mowwo
`
`gm
`
`hwNvrN
`
`m2
`
`Wm“dwEmac
`
`dmmu-mQEw
`
`.zoFomEozofiowmmoo
`
`
`AMZZ<IO
`
`29.522me
`
`m>02wm
`
`030>U
`
`Kimmn.
`
`w>0§wm
`
`O_I_O>O
`
`XEwmm
`
`rxx
`
`Ly
`
`:2xx
`
`5r
`
`EN
`
` oow
`
`
`
`owzmommomm032
`
`mtmmmwjgcmN0—H—
`
`Exhibit 2004
`IPR2015-00221
`Page 4 of 15
`
`Exhibit 2004
`IPR2015-00221
`Page 4 of 15
`
`
`
`
`
`
`
`US. Patent
`
`Sep. 14, 2010
`
`Sheet 3 of8
`
`US 7,796,681 B1
`
`m .@_n_
`
`
`
`1% > 1% r “mm > am \
`
`a? 1 a? R 3w + 3w r
`in). 2m N; / ?nd
`
`
`
`
`. . . Q55 6 ._<zow .@ 81R 6V3“ mime“ 6% 8&5N 818% a‘ 6% 81%
`
`Exhibit 2004
`IPR2015-00221
`Page 5 of 15
`
`
`
`U.S. Patent
`
`Sep.14,2010
`
`Sheet40f8
`
`US 7,796,681 B1
`
`___.._._,_.._._..._._.._.._,._.._
`
` Nov
`
`Exhibit 2004
`|PR2015-00221
`Page 6 of 15
`
`Exhibit 2004
`IPR2015-00221
`Page 6 of 15
`
`
`
`US. Patent
`
`Sep. 14, 2010
`
`Sheet 5 0f 8
`
`US 7,796,681 B1
`
`Perform ACR LPF and frequency
`offset correction on received signal
`
`502
`
`y
`
`l
`
`Estimate symbol timing from short
`training symbols received on MR
`antennas
`
`l
`
`Estimate ?ne frequency offset from
`MIT long preambles
`
`l
`
`Correct samples of long preambles
`using ?ne frequency offset
`
`l
`
`Calculate relative CPEs of MT long
`preambles
`
`l
`
`Perform channel estimation for pilot
`and data tones
`
`508
`
`510
`
`512
`
`FIG. 5A
`
`Exhibit 2004
`IPR2015-00221
`Page 7 of 15
`
`
`
`US. Patent
`
`Sep. 14, 2010
`
`Sheet 6 0f 8
`
`US 7,796,681 B1
`
`Perform MIMO equalization
`
`514
`T/
`
`516
`l
`Calculate CPE for an OFDM data J
`symbol
`
`518
`Calculate resldual frequency J
`tracking offset for d-th OFDM data
`symbol
`l
`Compensate for CPE
`
`520
`J
`
`22
`Perform space-frequency j
`deinterleaving and decoding
`
`l
`
`524
`
`Descramble data bits
`
`FIG. 5B
`
`Exhibit 2004
`IPR2015-00221
`Page 8 of 15
`
`
`
`US. Patent
`US. Patent
`
`Sep. 14, 2010
`Sep. 14, 2010
`
`Sheet 7 of8
`Sheet 7 of8
`
`US 7,796,681 B1
`US 7,796,681 B1
`
`I:
`
`
`
`FIRSTSAMPLE
`
`Exhibit 2004
`|PR2015-00221
`Page 9 of 15
`
`E
`
`
`
`(FIRSTSAMPLE)
`
`604
`
`(
`
`'Bewl
`I
`‘991 I
`.1.CIL I m L)
`OS JGJJnS
`
`('Bew! llq L ‘IeeJ uq I)
`085mg
`
`602
`
`
`
`Exhibit 2004
`IPR2015-00221
`Page 9 of 15
`
`
`
`US. Patent
`
`Sep. 14, 2010
`
`Sheet 8 of8
`
`US 7,796,681 B1
`
`-—>
`
`Mnmax : O,MnlOcalmax= O?ll =ornr : 0,"; = 0,"; =(Lnmax :01 c” = 0
`
`"I = 0
`and Mn >1'1A
`
`N0 ‘4
`
`Mfl> Mnmax ’
`NO
`
`M-l 2 Mn-Z
`and
`Mn‘! > M"
`
`N0 +
`
`Yes
`
`Set )1; = n
`
`I
`
`Set Mum“: Mypnm =n
`Store Pnmax
`
`M]
`S tM
`=
`I
`e
`7"
`localmax
`
`Yes
`
`M” <T2Mnlocal max
`MP1 2T2lm'local max
`
`Yes
`
`_
`Set "8 _ "J
`
`No ‘L
`
`"I > 0
`and M,,_1> r3 Mnmax
`
`Yes
`
`Set ctr = 0
`
`No
`
`"I > 0
`No
`
`Yes '
`
`617+- ctr + 1
`
`L— n<-n+1
`
`ctr = B?
`
`V
`
`Yes
`
`Set n, = n - B
`
`Coarse Frequency
`‘
`Estimates
`0% = tan 401mm,‘ magma ) Km) “A
`
`p z 2(1 '
`
`7
`
`wcarrier)
`
`Stan of Long Training
`Symbol is Estimated by
`_ L
`J
`n3 " ("8 41A)”
`
`Symbol Timing
`Estimates
`
`FIG 7
`'
`
`L——--—~—--—-->To Removal of Cyclic Pre?x
`
`Exhibit 2004
`IPR2015-00221
`Page 10 of 15
`
`
`
`US 7,796,681 B1
`
`1
`MIMO-OFDM RECEIVER PROCESSING
`
`CROSS-REFERENCE TO RELATED
`APPLICATIONS
`
`This application is a continuation application of US.
`patent application Ser. No. 10/912,829, ?led Aug. 5, 2004
`Which claims priority to US. Provisional Application Ser.
`No. 60/572,934, ?led on May 19, 2004, Which are hereby
`incorporated by reference.
`
`BACKGROUND
`
`Wireless phones, laptops, PDAs, base stations and other
`systems may Wirelessly transmit and receive data. A single
`in-single-out (SISO) system may have tWo single-antenna
`transceivers in Which one predominantly transmits and the
`other predominantly receives. The transceivers may use mul
`tiple data rates depending on channel quality.
`An MR><MT multiple-in-multiple-out (MIMO) Wireless
`system uses multiple transmit antennas (MT) and multiple
`receive antennas (MR) to improve data rates and link quality.
`The MIMO system may achieve high data rates by using a
`transmission signaling scheme called “spatial multiplexing,”
`Where a data bit stream is demultiplexed into parallel inde
`pendent data streams. The independent data streams are sent
`on different transmit antennas to obtain an increase in data
`rate according to the number of transmit antennas used. Alter
`natively, the MIMO system may improve link quality by
`using a transmission signaling scheme called “transmit diver
`sity,” Where the same data stream (i.e., same signal) is sent on
`multiple transmit antennas after appropriate coding. The
`receiver receives multiple copies of the coded signal and
`processes the copies to obtain an estimate of the received data.
`The number of independent data streams transmitted is
`referred to as the “multiplexing order” or spatial multiplexing
`rate (rs). A spatial multiplexing rate of 15:1 indicates pure
`diversity and a spatial multiplexing rate of rS:min(MR, MT)
`(minimum number of receive or transmit antennas) indicates
`pure multiplexing.
`
`20
`
`25
`
`30
`
`35
`
`40
`
`BRIEF DESCRIPTION OF THE DRAWINGS
`
`45
`
`FIG. 1 is a block diagram of a Wireless MIMO-OFDM
`communication system according to an embodiment.
`FIG. 2 is a block diagram of a receive section in a trans
`ceiver in the MIMO-OFDM communication system.
`FIG. 3 illustrates an IEEE 802.11a frame format.
`FIG. 4 illustrates a frame format for the MIMO-OFDM
`communication system.
`50
`FIGS. 5A and 5B shoW a ?owchart describing a MIMO
`OFDM receiver processing operation according to an
`embodiment.
`FIG. 6 is a block diagram of a portion of a time/frequency
`synchronization module in the receive section of the trans
`ceiver.
`FIG. 7 is a ?owchart describing a symbol timing estimation
`operation according to an embodiment.
`
`55
`
`DETAILED DESCRIPTION
`
`FIG. 1 illustrates a Wireless multiple-in-multiple-out
`(MIMO) communication system 100, Which includes a ?rst
`transceiver 102 With MT transmit (TX) antennas 104 and a
`second transceiver 106 With MR receive (RX) antennas 108,
`forming an MR><MT MIMO system. For the description
`beloW, the ?rst transceiver 102 is designated as a “transmit
`
`60
`
`65
`
`2
`ter” because the transceiver 102 predominantly transmits sig
`nals to the transceiver 106, Which predominantly receives
`signals and is designated as a “receiver”. Despite the desig
`nations, both “transmitter” 102 and “receiver” 106 may
`include a transmit section 110 and a receive section 112 and
`may transmit and receive data.
`The transmitter 100 and receiver 102 may be implemented
`in a Wireless local Area NetWork (WLAN) that complies With
`the IEEE 802.1 1 standards (including IEEE 802.1 1, 802.11a,
`802.11b, 802.11g, and 802.11n). The IEEE 802.11 standards
`describe
`orthogonal
`frequency-division multiplexing
`(OFDM) systems and the protocols used by such systems. In
`an OFDM system, a data stream is split into multiple sub
`streams, each of Which is sent over a different subcarrier
`frequency (also referred to as a “tone”). For example, in IEEE
`802.11a systems, OFDM symbols include 64 tones (With 48
`active data tones) indexed as {-32,-31, .
`. ., —1,0, 1, .
`.
`. , 30,
`31}, Where 0 is the DC tone index. The DC tone is not used to
`transmit information.
`The antennas in the transmitter 102 and receiver 106 com
`municate over channels in a Wireless medium. In FIG. 1, H
`represents the re?ections and multi-paths in the Wireless
`medium, Which may affect the quality of the channels. The
`system may perform channel estimation using knoWn train
`ing sequences Which are transmitted periodically (e. g., at the
`start of each frame). A training sequence may include one or
`more pilot symbols, i.e., OFDM symbols including only pilot
`information (Which is knoWn a priori at the receiver) on the
`tones. The pilot symbol(s) are inserted in front of each trans
`mitted frame. The receiver 106 uses the knoWn values to
`estimate the medium characteristics on each of the frequency
`tones used for data transmission. For example, on the receiver
`side, the signalYk for tone k in an SISO system can be Written
`as,
`
`Where Hk is the channel gain for the k-th tone, Xk is the
`symbol transmitted on the k-th tone, and Nk is the additive
`noise. An estimate of the channel may be determined at the
`receiver by dividing Yk by Xk.
`The number of independent data streams transmitted by the
`transmit antennas 104 is called the “multiplexing order” or
`“spatial multiplexing rate” (rs). A spatial multiplexing rate of
`rS:1 indicates pure diversity, and a spatial multiplexing rate
`of rS:min(MR, MT)(minimum number of receive or transmit
`antennas) indicates pure multiplexing.
`In an embodiment, the MIMO system 100 may use com
`binations of diversity and spatial multiplexing, e.g.,
`1§rS§min(MR, MT) . For example, in a 4x4 MIMO system,
`the system may select one of four available multiplexing rates
`(rSe[1, 2, 3, 4]) depending on the channel conditions. The
`system may change the spatial multiplexing rate as channel
`conditions change.
`FIG. 2 shoWs a block diagram of the receive section 112.
`The receive section 112 includes stages similar to those in the
`receive section of an IEEE 802.11a receiver, but With some
`modi?cations to account for the multiple receive antennas.
`Signals received on the multiple receive antennas are input
`to corresponding processing chains 200. Each processing
`chain includes a radio-frequency (RF) module 201 for RF-to
`baseband and analog-to-digital (A/ D) conversion. The
`receiver may have a common automatic gain control (AGC)
`for all antennas to provide minimal gain across all the receive
`
`Exhibit 2004
`IPR2015-00221
`Page 11 of 15
`
`
`
`US 7,796,681 B1
`
`3
`antennas. A time/ frequency synchronization module 202 per
`forms synchronization operations and extracts information
`from the multiple substreams (for rS>1) for channel estima
`tion 203. Each processing chain 200 includes a cyclic pre?x
`removal module 204, serial-to-parallel (S/P) converter 206,
`fast Fourier transform (EFT) module 208, a common phase
`error (CPE) correction module 210, a space-frequency detec
`tion module 212, and a parallel-to-serial (P/ S) converter 214.
`The multiple substreams are input to a space-frequency
`deinterleaver and decoding module 216 Which de-interleaves
`the substreams into a single data stream 217 andperforms soft
`Viterbi decoding. The single stream is then input to a
`descrambler 218.
`The MIMO-OFDM system may be compatible With IEEE
`802.11a systems, and consequently may have many similari
`ties to an IEEE 802.11a system. For example, like IEEE
`802.11a systems, the MIMO-OFDM system may use 52
`tones (48 data tones and 4 pilot tones), 312.5 kHz subcarrier
`spacing, an EFT/inverse FFT (IFFT) period of 3 .2 us, a cyclic
`pre?x With a duration of 0.8 us, and an OFDM symbol dura
`tion of 4.0 us. The MIMO-OFDM system may also use a
`frame format 300 similar to that speci?ed by IEEE 802.11a,
`Which is shoWn in FIG. 3. In addition, variations of the
`MIMO-OFDM systems are also possible, including using
`different numbers of tones, different guard intervals, different
`forWard error correction codes, and different constellations.
`An IEEE 802.11a frame 300 includes a short preamble
`301, a long preamble 302, a header 304, and a DATA ?eld
`306. The short preamble 301 includes a short training symbol
`308 With a duration of 0.8 is repeated ten times. The short
`preamble may be used for signal detection, AGC, coarse
`frequency offset estimation, and symbol timing estimation.
`The long preamble 302 includes tWo long training symbols
`310, each of duration 3.2 us, Which are separated from the
`short training symbols 308 by a long guard interval (1.6 us)
`312. The long preamble is used for ?ne frequency offset
`estimation and channel estimation.
`The header 304 includes a SIGNAL symbol 314, Which is
`encoded at 6 Mbps. The SIGNAL symbol 314 is 12 bits in
`length and includes 4 bits for the data rate, 1 reserved bit, 1
`parity bit, and 6 tail bits (set to “0” to return the convolutional
`decoder to State 0). The DATA ?eld 306 includes OFDM
`45
`symbols including the data bits to be transmitted. The data
`bits are prepended by a 16-bit SERVICE ?eld and are
`appended by 6 tail bits. The resulting bits are appended by a
`number of pad bits needed to yield an integer number of
`OFDM symbols.
`The MIMO-OFDM system 100 may use a similar frame
`format 400, as shoWn in FIG. 4. The illustrated frame format
`400 is for systems With three transmit antennas (MT:3), but
`can be modi?ed for other MT. Each transmit antenna trans
`mits a different MIMO-OFDM frame 400. Like the IEEE
`802.11a frame 300, the MIMO-OFDM frames 400 include a
`short preamble 402 With a series of short training symbols
`404, a long preamble 405 With a set of tWo long training
`symbols 606, a header 408 including a SIGNAL symbol 610,
`and a data ?eld 412. In addition, the header 408 may include
`a second SIGNAL symbol (SIGNAL2) 614, Which may be
`used to transmit MIMO-OFDM-speci?c information, such as
`the number of transmit antennas and the spatial multiplexing
`rate. The frame may also include a supplemental long pre
`amble 416 including MT-l additional long training symbols
`to train the other antennas.
`
`50
`
`60
`
`65
`
`20
`
`25
`
`30
`
`35
`
`40
`
`4
`As in IEEE 802.11a, a short OFDM training symbol con
`sists of 12 tones, Which are modulated by the elements of the
`folloWing frequency-domain sequence:
`
`546,26 =
`
`The multiplication by is in order to normalize the average
`poWer of the resulting OFDM symbol. The short training
`symbol has a duration of 0.8 ps and is repeated 10 times.
`As in IEEE 802.11a, a long training OFDM symbol
`includes 52 tones, Which are modulated by the folloWing
`frequency-domain BPSK training sequence:
`L,26726:{1,1,—1,—1,1,1,—1,1,—1,1,1,1,1,1,1,—1,—1,1,1,—
`1,1,-1,1,1,1,1,0,1,-1,-,1,1,1,-1,1,-1,1,-1,-1,
`1,-1,-1,1,1,-1,1,-1,1,-1,1,1,1,1}
`The number of sets of long training symbols (or “long
`preambles”) may be MTfor all spatial multiplexing rates. The
`additional long training symbols may be used to estimate the
`full M R><MT channel matrix. This estimation may be used for
`link adaptation, in Which modulation, coding rate, and/or
`other signal transmission parameters may be dynamically
`adapted to the changing channel conditions.
`FIGS. 5A-5B shoW a ?owchart describing a MIMO
`OFDM signal processing operation 500 performed by the
`receiver 106. FIG. 6 shoWs a portion of the time/frequency
`synchronization module 202. The module 202 may perform
`adjacent channel rejection (ACR) and loW pass ?ltering
`(LPF) operations (using ACR LPF module 602) and fre
`quency offset correction on the received signals (block 502).
`The time/ frequency synchronization module 202 may use
`the short training symbols to estimate symbol timing (block
`504). The received signal for the i-th receive antenna and n-th
`sample (rim) may be used by a computation module 604 to
`generate a quantity ql-m using the folloWing equation:
`
`The quantity ql-m calculated for each of the MR antennas
`may be used by a summing module 605 to generate a metric
`P” for the n-th sample using the folloWing equation:
`
`1
`
`Where L denotes the number of samples in one short train
`ing symbol. A value Mn for the n-th sample may then be
`computed using the folloWing equation:
`
`Where the parameter as may have a value in the range of
`(0, .
`.
`.
`, 31/64), e.g., 3/32 for a 40 MHz analog-to-digital
`(A/ D) conversion rate.
`The value Mn may be used to estimate the symbol timing as
`in IEEE 802.1 1a, With the exception of using a more ?exible
`threshold '53 to check for n, (right endpoint of the plateau), as
`shoWn in FIG. 7. Typical parameter values for a 40 MHz
`analog-to-digital (A/D) conversion rate are given in Table 1.
`
`Exhibit 2004
`IPR2015-00221
`Page 12 of 15
`
`
`
`US 7,796,681 B1
`
`TABLE 1
`
`Parameter
`
`Exemplary value
`
`Range
`
`L
`
`1:1
`A
`1:2
`
`'l:3
`
`B
`nD
`
`32
`
`0.375
`64 * MRR
`0.890625
`
`0.5
`
`15
`25
`
`(0, .
`
`.
`
`. ,255/256)
`
`(0, .
`
`(0, .
`
`.
`
`.
`
`. ,255/256)
`
`. ,255/256)
`
`, 63)
`.
`.
`(0, .
`(0,. .. ,63)
`
`The time/ frequency synchronization module 202 may esti
`mate the ?ne frequency offset (block 506) by correlating the
`received M R><1 vectors from the tWo long training symbols in
`a long preamble using the following equation:
`
`6
`are alWays sent on the same tone, the subcarrier channel
`estimates may be calculated using the folloWing equation:
`
`An equalizer 220 may perform MIMO equalization (block
`514) by forming an M R><rs effective channel matrix for data
`tone k:
`
`The equalizer 220 may use a zero forcing equalizer per
`tone:
`
`20
`
`Gk:(Hk*Hk)TlHk*
`The equalizer 220 may then compute a bit-metric Weight
`for the l-th substream, Which equals the normalized post
`processing signal-to-noise ratio (SNR) of the l-th substream:
`
`Where N is the FFT size. The angle of the correlation result
`(Af) may be used to estimate the ?ne frequency offset using
`the following equation:
`
`25
`
`arg( C)
`27rN TS ’
`
`Where T S is the sampling period at the FFT 208 output. The
`time/frequency synchronization module 202 may correct
`samples of long training symbols using the estimated ?ne
`frequency offset (block 508).
`The channel estimation module 203 may perform channel
`estimation by averaging the corrected samples corresponding
`to the tWo long training symbols in a long preamble for all
`receive antennas 108. The channel estimation module 203
`may compute the relative CPEs of the MT long preambles
`(block 510) using the folloWing equation:
`
`MRil
`
`Z 2 IR???
`
`kekpilom [:0
`
`Where CPEOI1, Ralf") is the k-th FFT output for the i-th
`receive antenna and p-th preamble, and Kim-lots are the indices
`ofthe pilot tones. As in IEEE 802.1 1a, tones k:—2 1, —7, 7, and
`21 are used for pilot tones in each data MIMO-OFDM sym
`bol.
`The channel estimation module 203 may generate channel
`estimates for the pilot tones and data tones using the fre
`quency domain BPSK (Biphase Shifting Key) long training
`symbols (Lk) (block 512). For the data tones, the subcarrier
`channel estimates may be calculated using the folloWing
`equation:
`
`30
`
`35
`
`40
`
`45
`
`50
`
`55
`
`60
`
`Where l,l represents the diagonal element.
`The CPE correction module 210 may generate a scalar
`CPE estimate for the d-th data symbol (block 518) using the
`folloWing equation:
`
`CPEM) : kekpilm [:0
`MRil
`A10) 2
`lhi,kl
`
`Where Yhkw is the k-th output for the i-th receive antenna
`and d-th data symbol, and Pke{1,—1} is the BPSK pilot sym
`bol for tone k.
`Using this CPE value, the CPE correction module 210 can
`determine the residual frequency offset tracking for the d-th
`OFDM data symbol (Afw) (block 518) using the folloWing
`equation:
`
`The space-frequency detection module 212 may generate
`the k-th output for the d-th data symbol by concatenating the
`corresponding outputs of the MR receive antennas:
`
`The space-frequency detection module 212 may then form
`an equalized signal for data tone k using the zero forcing
`equalizer for the tone (Gk):
`
`The space-frequency detection module 212 may then com
`pensate for CPE (block 520) using the folloWing equations:
`
`Where half") is the channel estimate for the k-th tone, i-th
`receive antenna, and p-th preamble. For the pilot tones, Which
`
`65
`
`X1501) : X1501) /c P E01);
`
`WM: 1 CPE“) 12 WM.
`
`Exhibit 2004
`IPR2015-00221
`Page 13 of 15
`
`
`
`US 7,796,681 B1
`
`7
`The CPE compensated Weight value for the l-th substream
`(WLk) may be used to obtain log-likelihood ratios (LLRs) for
`soft Viterbi decoding. As in IEEE 802.11a, the space-fre
`quency deinterleaving and decoding module 216 may con
`catenate LLRs for each substream into a single sequence,
`deinterleave the LLR sequence, and decode data bits using
`soft Viterbi decoding (block 522). The descrambler may then
`descramble data bits using scrambler state estimation
`obtained from the SERVICE ?eld (block 524).
`A number of embodiments have been described. Neverthe
`less, it Will be understood that various modi?cations may be
`made Without departing from the spirit and scope of the
`invention. For example, blocks in the ?oWcharts may be
`skipped or performed out of order and still produce desirable
`results. Accordingly, other embodiments are Within the scope
`of the folloWing claims.
`
`The invention claimed is:
`1. An apparatus comprising:
`a receive section to process signals received on a plurality
`of receive antennas, the receive section including:
`a channel estimation module to determine a relative com
`mon phase error value for each of a plurality of long
`preambles in a multiple-in-multiple-out-orthogonal fre
`quency-division multiplexing (MlMO-OFDM) frame
`using a fast Fourier transform value calculated for each
`of the plurality of receive antennas.
`2. The apparatus of claim 1, Wherein the channel estimation
`module is operative to determine subcarrier channel estimates
`for each data tone from the relative common phase error and
`the fast Fourier transform value.
`3. The apparatus of claim 2, further comprising:
`an equalizer to generate an M R ><rs effective channel matrix
`for a data tone using the subcarrier channel estimates
`from the channel estimation module, Where MR is the
`number of receive antennas and rs is a spatial multiplex
`ing rate corresponding to a number of substreams.
`4. The apparatus of claim 3, Wherein the equaliZer is opera
`tive to generate a bit-metric Weight for each of the substreams
`from the effective channel matrix.
`5. The apparatus of claim 3, Wherein the equaliZer is opera
`tive to generate a Zero forcing equaliZer value for the data tone
`from the effective channel matrix.
`6. The apparatus of claim 5, further comprising:
`a detection module to generate an equaliZed signal for a
`data tone from an output of each of the plurality of
`receive antennas and the Zero forcing equaliZer value.
`7. The apparatus of claim 3, further comprising:
`a detection module to:
`generate an equaliZed signal for a data tone from an
`output of each of the plurality of receive antennas and
`a Zero forcing equaliZer value determined from the
`effective channel matrix, and
`compensate for the relative common phase error value
`for each of a plurality of long preambles using the
`equaliZed signal.
`8. The apparatus of claim 1, further comprising:
`a timing module to estimate symbol timing using samples
`in a short preamble in the MlMO-OFDM frame, Wherein
`each sample is received on each of the plurality of
`receive antennas.
`9. A method comprising:
`receiving orthogonal frequency-division multiplexing
`(OFDM) symbols on a plurality of receive antennas; and
`determining a relative common phase error value for each
`of a plurality of long preambles in a multiple-in-mul
`
`20
`
`25
`
`30
`
`35
`
`40
`
`45
`
`50
`
`55
`
`60
`
`65
`
`8
`tiple-out (MlMO-OFDM) frame using a fast Fourier
`transform value calculated for each of the plurality of
`receive antennas.
`10. The method of claim 9, further comprising:
`determining subcarrier channel estimates for each data
`tone from the relative common phase error and the fast
`Fourier transform value.
`11. The method of claim 10, further comprising:
`generating an MR><rS effective channel matrix for a data
`tone using the subcarrier channel estimates from the
`channel estimation module, Where MR is the number of
`receive antennas and rs is a spatial multiplexing rate
`corresponding to a number of substreams.
`12. The method of claim 11, further comprising:
`generating a bit-metric Weight for each of the substreams
`from the effective channel matrix.
`13. The method of claim 11, further comprising:
`generating a Zero forcing equaliZer value for the data tone
`from the effective channel matrix.
`14. The method of claim 13, further comprising:
`generating an equaliZed signal for a data tone from an
`output of each of the plurality of receive antennas and
`the Zero forcing equaliZer value.
`15. The method of claim 11, further comprising:
`generating an equaliZed signal for a data tone from an
`output of each of the plurality of receive antennas and a
`Zero forcing equaliZer value determined from the effec
`tive channel matrix, and
`compensating for the relative common phase error value
`for each of a plurality of long preambles using the equal
`iZed signal.
`16. The method of claim 9, further comprising:
`estimating symbol timing using samples in a short pre
`amble in the MlMO-OFDM frame, Wherein each
`sample is received on each of the plurality of receive
`antennas.
`17. A transceiver comprising:
`a plurality of receive antennas; and
`a receive section to process signals received on the plurality
`of receive antennas, the receive section including:
`a channel estimation module to determine a relative com
`mon phase error value for each of a plurality of long
`preambles in a multiple-in-multiple-out-or‘thogonal fre
`quency-division multiplexing (MlMO-OFDM) frame
`using a fast Fourier transform value calculated for each
`of the plurality of receive antennas.
`18. The transceiver of claim 17, Wherein the channel esti
`mation module is operative to determine subcarrier channel
`estimates for each of the data tones from the relative common
`phase errors and the fast Fourier transform values.
`19. The transceiver of claim 18, further comprising:
`an equaliZer to generate an M R ><rs effective channel matrix
`for a data tone using the subcarrier channel estimates
`from the channel estimation module, Where MR is the
`number of receive antennas and rs is a spatial multiplex
`ing rate corresponding to a number of substreams.
`20. The transceiver of claim 19, Wherein the equaliZer is
`operative to generate a bit-metric Weight for each of the
`substreams from the effective channel matrix.
`21. The transceiver of claim 19, Wherein the equaliZer is
`operative to generate a Zero forcing equaliZer value for the
`data tone from the effective channel matrix.
`22. The transceiver of claim 21, further comprising:
`a detection module to generate an equaliZed signal for a
`data tone from an output of each of the plurality of
`receive antennas and the Zero forcing equaliZer value.
`
`Exhibit 2004
`IPR2015-00221
`Page 14 of 15
`
`
`
`US 7,796,681 B1
`
`9
`23. The transceiver of claim 19, further comprising:
`a detection module to:
`generate an equalized signal for a data tone from an
`output of each of the plurality of receive antennas and
`a zero forcing equalizer value determined from the
`effective channel matrix, and
`compensate for the relative common phase error value
`for each of a plurality of long preambles using the
`equalized signal.
`24. The transceiver of claim 17, further comprising:
`a timing module to estimate symbol timing using samples
`in a short preamble in the MlMO-OFDM frame, Wherein
`each sample is received on each of the plurality of
`receive antennas.
`25. An system comprising:
`a receive section including:
`means for processing signals received on a plurality of
`receive antennas; and
`means for determining a relative common phase error
`value for each of a plurality of long preambles in a
`multiple-in-multiple-out-orthogonal frequency-divi
`sion multiplexing (MlMO-OFDM) frame using a fast
`Fourier transform value calculated for each of the
`plurality of receive antennas.
`26. The system of claim 25, further comprising:
`means for determining subcarrier channel estimates for
`each of the data tones from the relative common phase
`errors and the fast Fourier transform values.
`27. The system of claim 26, further comprising:
`means for generating an M R><rs effective channel matrix for
`a data tone using the subcarrier channel estimates from
`the channel estimation module, Where M R is the number
`of receive antennas and rs is a spatial multiplexing rate
`corresponding to a number of substreams.
`28. The system of claim 27, further comprising:
`means for generating a bit-metric Weight for each of the
`substreams from the effective channel matrix.
`29. The system of claim 27, further comprising:
`means for generating a zero forcing equalizer value for the
`data tone from the effective channel matrix.
`30. The system of claim 29, further comprising:
`means for generating an equalized signal for a data tone
`from an output of each of the plurality of receive anten
`nas and the zero forcing equalizer value.
`31. The system of