US007408976B1
`
`(12) United States Patent
`Narasimhan et a].
`
`(10) Patent N0.:
`(45) Date of Patent:
`
`US 7,408,976 B1
`Aug. 5,2008
`
`(54) MIMO-OFDM RECEIVER PROCESSING
`WITH FREQUENCY AND CHANNEL
`ESTIMATION
`
`(75) Inventors: Ravi Narasimhan, Los Altos, CA (US);
`Hemanth Sampath, Sunnyvale, CA
`(US); Hsiao-Cheng Tang, Milpitas, CA
`(Us)
`
`(73)
`
`Assignee:
`
`Marvell International Ltd., Hamilton
`(BM)
`
`Notice:
`
`Subject to any disclaimer, the term of this
`patent is extended or adjusted under 35
`U.S.C. 154(b) by 617 days.
`
`(21)
`
`(22)
`
`(60)
`
`(51)
`
`(52)
`(58)
`
`Appl. N0.: 10/912,s29
`
`Filed:
`
`Aug. 5, 2004
`
`Related US. Application Data
`
`Provisional application No. 60/572,934, ?led on May
`19, 2004.
`
`Int. Cl.
`(2006.01)
`H04B 1/69
`(2006.01)
`H04B 1/04
`US. Cl. ...................... .. 375/148; 375/299; 455/132
`Field of Classi?cation Search ............... .. 375/148,
`375/347, 349, 299, 346, 348; 324/614; 455/132,
`455/ 101
`See application ?le for complete search history.
`
`(56)
`
`References Cited
`
`U.S. PATENT DOCUMENTS
`
`2004/0136313 A1*
`2005/0180312 A1*
`2005/0195733 A1*
`2006/0252386 A1*
`
`7/2004 Goldstein et al. ......... .. 370/203
`
`8/2005 Walton et al. . . . . . . .
`
`. . . .. 370/208
`
`9/2005 Walton et al. ............. .. 370/208
`
`11/2006 Boer et al. ................ .. 455/101
`
`OTHER PUBLICATIONS
`
`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
`1999iFall. IEEE VTS 50th vol. 1, Sep. 19-22, 1999 pp. 258-262
`discloses a system with relevance to claims 1-60.*
`“Information
`IEEE
`Computer
`Society,
`TechnologyiTelecommunications and Information Exchange
`Between SystemsiLocal and Metropolitan Area N etworksiSpeci?c
`RequirementsiPart 11.‘ Wireless LAN Medium Access Control
`(MAC) and Physical Layer (PHY) Speci?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.
`11ai1999 (Supplement to IEEE Std 802.11-1999)
`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.‘
`Higher-speed Physical Layer Extension in the 2.4 GHZ Band’ ’, IEEE
`Std 802.11b-1999 (Supplement to IEEE Std 80211-1999).
`
`(Continued)
`Primary ExamineriKhai Tran
`
`(57)
`
`ABSTRACT
`
`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.
`
`IEEE Std 802.11a-1999 (R2003), Part 11: Wireless LAN Medium
`Access Control (MAC) and Physical Layer (PHY) speci?cations,
`
`50 Claims, 8 Drawing Sheets
`
`100
`./
`
`FIRSTTRANSCENER
`"TRANSMITI'ER’ 192 j
`
`110
`
`TRANSMIT
`SECTION
`I
`
`112
`\ REcEIvE
`SECTION
`
`'
`I
`
`I
`
`l
`Frame
`Formatter
`(
`120
`
`SECOND TRANSCEIVER
`Y__ “RECEIVER" 1L6
`110
`' \ TRANSMIT
`I
`SECTION
`I
`I
`
`112
`\ RECEIVE
`Y__
`SECTION
`
`l
`Frame
`Formatter
`g
`120
`
`08
`A
`
`Exhibit 2003
`IPR2015-00221
`Page 1 of 16
`
`

`

`US 7,408,976 B1
`Page 2
`
`OTHER PUBLICATIONS
`
`IEEE Computer Society, “IEEE Standard for Information
`Technology~DRAF T 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.
`
`* cited by examiner
`
`Exhibit 2003
`IPR2015-00221
`Page 2 of 16
`
`

`

`US. Patent
`
`Aug. 5, 2008
`
`Sheet 1 of8
`
`US 7,408,976 B1
`
`
`
`25E 296%
`
`_ , tzwzé J
`
`a? .mwzwowma
`
`o:
`
`oow
`
`5:95am _
`ON; w>momm J
`205%
`
`N:
`
`TUE
`
`0&9“. 296mm
`
`_ :22»: W;
`
`hozmEhom _
`ON? w>mowm J J N:
`
`, 205%
`
`Exhibit 2003
`IPR2015-00221
`Page 3 of 16
`
`

`

`U.S. Patent
`
`Aug. 5, 2008
`
`Sheet 2 of 8
`
`US 7,408,976 B1
`
`-mDa
`
`02.000
`
`mi
`
`amimszmowwo
`
`m_.N
`
`\.FNer
`
`9N
`
`-w0<n_w
`
`mi
`
`.men—IWU<QW
`
`
`
`«WWI—finZOFOwhmoZOFmemOO
`-mo8%was
`
`
`
`Dm>_wommCNN
`
`05:2
`
`wtm«EN—.23Ow
`
`4mzz<IO
`
`29.2.2.me
`
`N.OE
`
`w>02mm
`
`o_._o>o
`
`van—mum
`
`w>02mm
`
`O_._O>o
`
`van—wan.
`
`DON
`
`
`
`:9.
`
`L7
`
`L7
`
`SN
`
`Exhibit 2003
`IPR2015-00221
`Page 4 of 16
`
`Exhibit 2003
`IPR2015-00221
`Page 4 of 16
`
`
`
`
`
`

`

`U.S. Patent
`
`Aug. 5, 2008
`
`Sheet 3 of 8
`
`US 7,408,976 B1
`
`
`
`mom3»rllJwaII'IILSm
`
`
`
`m.OE
`
`3v¢3x3m
`
`
`
`...Engagfimafiaafiflggafi
`
`SmSmN58m
`
`Exhibit 2003
`|PR2015-00221
`Page 5 of 16
`
`Exhibit 2003
`IPR2015-00221
`Page 5 of 16
`
`
`

`

`U.S. Patent
`
`Aug. 5, 2008
`
`Sheet 4 of 8
`
`US 7,408,976 B1
`
`
`
`Exhibit 2003
`|PR2015-00221
`Page 6 of 16
`
`Exhibit 2003
`IPR2015-00221
`Page 6 of 16
`
`

`

`US. Patent
`
`Aug. 5, 2008
`
`Sheet 5 of8
`
`US 7,408,976 B1
`
`502
`Perform ACR LPF and frequency J
`offset correction on received signal
`
`504
`i7
`Estimate symbol timing from short /
`training symbols received on MR
`antennas
`
`506
`J7
`Estimate ?ne frequency offset from J
`MT long preambles
`
`Ell
`Correct samples of long preambles
`using ?ne frequency offset
`
`508
`
`510
`V
`Calculate relative CPEs of MT long J
`preambles
`
`512
`3'7
`Perform channel estimation for pilot J
`and data tones
`
`FUG, 5A
`
`Exhibit 2003
`IPR2015-00221
`Page 7 of 16
`
`

`

`US. Patent
`
`Aug. 5, 2008
`
`Sheet 6 of8
`
`US 7,408,976 B1
`
`Perform MlMO equalization
`
`514
`‘J
`
`516
`l
`,
`Calculate CPE for d-th OFDM data J
`symbol
`
`518
`Calculate resldual frequency J
`tracking offset for d-th OFDM data
`symbol
`l
`Compensate for CPE
`
`520
`'/
`
`Perform space-frequency
`deinterleaving and decoding
`
`J22
`
`l
`
`Descramble data bits
`
`524
`_/
`
`FIG. 5B
`
`Exhibit 2003
`IPR2015-00221
`Page 8 of 16
`
`

`

`U.S. Patent
`
`Aug. 5, 2008
`
`Sheet 7 of 8
`
`US 7,408,976 B1
`
`+8mm—+:vmmvcmm
`
`8E:8E_Emm._
`
`
`
`In...)A“ill
`2kIiw
`a8a.EamwImyI.o
`
`w
`
`m.9“.
`
`m._n=2<m5a.“:
`
`...o
`
`SusanaSEa!nnu“vv“;.oamIa8
`.u.EHmm”
`
`Exhibit 2003
`|PR2015-00221
`Page 9 of 16
`
`Exhibit 2003
`IPR2015-00221
`Page 9 of 16
`
`
`
`

`

`US. Patent
`
`Aug. 5, 2008
`
`Sheet 8 of8
`
`US 7,408,976 B1
`
`"I = 0
`and Mn >114
`
`No 4
`
`Yes
`
`Set n; = n
`
`I
`
`M" > Mnmax
`No
`
`Mn-] 2 Mn-Z
`and
`Mn4> Mn
`
`No 4
`
`Mun-‘ax: Mn," max = H
`Store Pnmax
`
`l
`S
`= M _,
`at M,
`local max
`n
`
`Yes
`
`M" <T2Mnlocal max
`M4 ZTZMnIocal max
`
`YES
`
`_
`Set ng _ "-1
`
`No ‘
`
`n, > O
`and MM > '3 M"max
`
`Y es
`
`Set ctr = 0
`
`-— n<--n+1
`
`No
`
`No
`
`"I > 0
`
`No
`
`ctr = B?
`
`Yes
`
`Yes
`
`crr<-— ctr + 1
`
`i
`
`Set n,- = n - B
`
`Coarse Frequency
`Estimates
`‘
`40,3 = tan "(lnrlPnmax H1013,max ) /(L/2) , “A I
`
`p = 2(1 - GJA/ 02cm") L’
`
`Start of Long Training
`Symbol is Estimated by
`n _ L
`J
`s — (ng -nA)/2
`
`Symbol Timing
`Estimates
`
`FIG 7
`-
`
`l_——-—->To Removal of Cyclic Pre?x
`
`Exhibit 2003
`IPR2015-00221
`Page 10 of 16
`
`

`

`US 7,408,976 B1
`
`1
`MIMO-OFDM RECEIVER PROCESSING
`WITH FREQUENCY AND CHANNEL
`ESTIMATION
`
`CROSS-REFERENCE TO RELATED
`APPLICATIONS
`
`This application claims priority to US. Provisional Appli
`cation Ser. No. 60/572,934, ?led on May 19, 2004.
`
`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 rfl indicates pure
`diversity and a spatial multiplexing rate of rS:min(MR, MT)
`(minimum number of receive or transmit antennas) indicates
`pure multiplexing.
`
`BRIEF DESCRIPTION OF THE DRAWINGS
`
`20
`
`25
`
`30
`
`35
`
`40
`
`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.
`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.
`
`50
`
`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
`ter” because the transceiver 102 predominantly transmits sig
`
`60
`
`65
`
`2
`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 rSImin (M R, 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
`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,
`
`Exhibit 2003
`IPR2015-00221
`Page 11 of 16
`
`

`

`US 7,408,976 B1
`
`3
`fast Fourier transform (FFT) 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 sub streams 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 is, 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 302 includes of a short training
`symbol 308 With a duration of 0.8 ps 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 508 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 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 (MTI3), 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 406, a header 408 including a SIGNAL symbol 410,
`and a data ?eld 412. In addition, the header 408 may include
`a second SIGNAL symbol (SIGNAL2) 414, 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 Mfr-1 additional long training symbols
`to train the other antennas.
`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:
`
`The multiplication by ‘/13/6 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_26,26:{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,—
`1,1,1,1,1 }
`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 (rm) may be used by a computation module 604 to
`generate a quantity qim using the folloWing equation:
`
`20
`
`25
`
`30
`
`35
`
`40
`
`45
`
`The quantity qim calculated for each of the MR antennas
`may be used by a summing module 606 to generate a metric
`P” for the n-th sample using the folloWing equation:
`
`50
`
`Ms
`
`55
`
`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:
`
`60
`
`65
`
`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 2003
`IPR2015-00221
`Page 12 of 16
`
`

`

`US 7,408,976 B1
`
`TABLE 1
`
`Parameter
`
`Exemplary value
`
`Range
`
`L
`
`1:1
`A
`1:2
`
`'l:3
`
`B
`
`nD
`
`32
`
`0.375
`64*MR
`0.890625
`
`0.5
`
`15
`
`25
`
`(0, .
`
`.
`
`. ,255/256)
`
`(0, .
`
`(0, .
`
`.
`
`.
`
`. ,255/256)
`
`. ,255/256)
`
`(0, .
`
`.
`
`(0,. .
`
`.
`
`.
`
`, 63)
`
`, 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:
`
`Mlrl Nil
`
`C: 2 ZrInriJwN,
`
`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:
`
`Am)
`hok
`
`qrsgl)
`h0,k
`
`Hk :
`
`.
`
`.
`
`Am)
`hMRglk
`
`qrsgl)
`hMRgrk
`
`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)
`Af:
`27rNTS ’
`
`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:
`
`kg kpum [: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:
`
`where half") is the channel estimate for the k-th tone, i-th
`receive antenna, and p-th preamble. For the pilot tones, which
`
`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:
`
`kg K -
`ptlOlS
`CPE‘d) =
`
`zgo
`
`MRgr
`Am) 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):
`
`30
`
`35
`
`40
`
`45
`
`50
`
`55
`
`60
`
`The space-frequency detection module 212 may then com
`pensate for CPE (block 520) using the following equations:
`
`65
`
`X1501) : X1501) /CPE(J);
`
`Exhibit 2003
`IPR2015-00221
`Page 13 of 16
`
`

`

`US 7,408,976 B1
`
`20
`
`7
`The CPE compensated Weight value for the l-th substream
`(VAVLk) 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 frequency estimation module to estimate a ?ne frequency
`offset from a plurality of long preambles in a multiple
`in-multiple-out-or‘thogonal frequency-division multi
`plexing (MlMO-OFDM) frame, the MlMO-OFDM
`frame including at least one long preamble for each of a
`25
`plurality of transmit antennas,
`Wherein the frequency estimation module is operative to
`correlate a plurality of received M R><1 vectors from the
`plurality of long preambles to produce a correlation
`result, Where MR is the number of receive antennas.
`2. The apparatus of claim 1, Wherein the frequency estima
`tion module is operative to use an angle of the correlation
`results to estimate the ?ne frequency offset.
`3. 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.
`4. 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; and
`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 multiplexing rate corresponding to a number of
`substreams, and
`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 trans
`form value.
`5. The apparatus of claim 4, Wherein the equaliZer is opera
`tive to generate a bit-metric Weight for each of the substreams
`from the effective channel matrix.
`6. The apparatus of claim 4, Wherein the equaliZer is opera
`tive to generate a Zero forcing equaliZer value for the data tone
`from the effective channel matrix.
`7. The apparatus of claim 6, 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.
`
`8
`
`8. A method comprising:
`receiving orthogonal frequency-division multiplexing
`(OFDM) symbols on a plurality of receive antennas; and
`estimating a ?ne frequency offset from a plurality of long
`preambles in a multiple-in-multiple-out (MIMO)
`OFDM frame, the MlMO-OFDM frame including at
`least one long preamble for each of a plurality of trans
`mit antennas,
`Wherein said estimating comprises correlating a plurality
`of received MR><1 vectors from the plurality of long
`preambles to produce a correlation result, Where MR is
`the number of receive antennas.
`9. The method of claim 8, further comprising:
`calculating an angle of the correlation results; and
`using said angle to estimate the ?ne frequency offset.
`10. The method of claim 8, 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.
`11. A method comprising:
`receiving orthogonal frequency-division multiplexing
`(OFDM) symbols on a plurality of receive antennas;
`determining a relative common phase error value for each
`of a plurality of long preambles in a multiple-in-mul
`tiple-out (MlMO-OFDM) frame using a fast Fourier
`transform value calculated for each of the plurality of
`receive antennas;
`determining subcarrier channel estimates for each data
`tone from the relative common phase error and the fast
`Fourier transform value; and
`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. An apparatus comprising:
`a receive section including:
`means for processing signals received on a plurality of
`receive antennas;
`means for estimating a ?ne frequency offset from a plural
`ity of long preambles in a multiple-in-multiple-out
`orthogonal frequency-division multiplexing (MIMO
`OFDM) frame, the MlMO-OFDM frame including at
`least one long preamble for each of a plurality of trans
`mit antennas, and
`means for correlating a plurality of received MR><1 vectors
`from the plurality of long preambles to produce a corre
`lation result, Where MR is the number of receive anten
`nas.
`16. The apparatus of claim 15, means for estimating the
`?ne frequency offset from an angle of the correlation result.
`17. The apparatus of claim 15, further comprising means
`for 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.
`
`30
`
`35
`
`40
`
`45
`
`50
`
`55
`
`60
`
`65
`
`Exhibit 2003
`IPR2015-00221
`Page 14 of 16
`
`

`

`US 7,408,976 B1
`
`9
`18. The apparatus of claim 15, further comprising means
`for determining a relative common phase error value for each
`of the plurality of long preambles using a fast Fourier trans
`form value calculated for each of the plurality of receive
`antennas.
`19. The apparatus of claim 15, 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.
`20. The apparatus of claim 19, further comprising means
`for generating an MR><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 anten
`nas and rs a spatial multiplexing rate corresponding to a num
`ber of substreams.
`21. The apparatus of claim 20, further comprising means
`for generating a bit-metric Weight for each of the substreams
`from the effective channel matrix.
`22. The apparatus of claim 20, further comprising means
`for generating a zero forcing equalizer value for the data tone
`from the effective channel matrix.
`23. The apparatus of claim 22, further comprising means
`for 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.
`24. A computer-readable medium having instructions
`stored thereon, Which, When executed by a processor, causes
`the processor to perform operations comprising:
`receiving orthogonal frequency-division multiplexing
`(OFDM) symbols from a plurality of receive antennas;
`and
`estimating a ?ne frequency offset from a plurality of long
`preambles in a multiple-in-multiple-out (MIMO)
`OFDM frame, the MlMO-OFDM frame including at
`least one long preamble for each of a plurality of trans
`mit antennas,
`Wherein said estimating comprises correlating a plurality
`of received MRxl vectors from the plurality of long
`preambles to produce a correlation result, Where MR is
`the number of receive antennas.
`25. The computer-readable medium of claim 24, further
`comprising:
`calculating an angle of the correlation results; and
`using said angle to estimate the ?ne frequency offset.
`26. The computer-readable medium of claim 24, further
`comprising:
`estimating symbol timing using samples in a short pre
`amble in the MlMO-OFDM frame, Wherein each
`sample is received on each