@ffiffiffiffi
`
`Advsncing Technology
`for Humanity
`
`DECLARATION OF GERARD P. GRENIER
`
`I, Gerard P. Grenier, am over twenty-one (21) years of age. I have never been convicted
`of a felony, and I am fully competent to make this declaration. I declaie the following to be true
`to the best of my knowledge, information and belief:
`
`1. I am Senior Director of Publishing Technologies of the Institute of Electrical and
`Electronics Engineers, Inc. ("IEEE").
`
`2.
`
`aJ.
`
`4.
`
`5.
`
`6.
`
`7.
`
`IEEE is a neutral third party in this dispute.
`\
`Neither I nor IEEE itself is being compensated for this declaration.
`
`Among my responsibilities as Senior Director of Publishing Technologies, I act as a
`custodian of certain records for IEEE.
`
`I make this declaration based on my personal knowledge and information contained
`in the business records of IEEE.
`
`As part of its ordinary course of business IEEE publishes and makes available
`technical articles and standards. These publications are made available for public
`download through the IEEE digital library, IEEE Xplore.
`
`It is the regular practice of IEEE to publish articles and other writings including
`article abstracts and make them available to the public through IEEE Xplore. IEEE
`maintains copies of publications in the ordinary course of its regularly conducted
`activities.
`
`8.
`
`The article below has been attached as Exhibits A to this declaration:
`
`t
`
`A.
`
`J.C. Rault, et al. "The coded orthogonal frequency division multiplexing
`(COFDM) technique, and its application to digital radio broadcasting
`towards mobile receivers" IEEE Global Telecommunications Conference
`and Exhibition oCommunications Technology for the 1990s and Beyond
`(GLOBECOM), November 27 -30, 1989.
`
`9. I obtained a copy of Exhibit A through IEEE Xplore, where it is maintained iri the
`ordinary course of IEEE's business. Exhibit A is a true and correct copy of the
`Exhibit as it existed on or about December 10,2016.
`
`10. The article abstracts from IEEE Xplore shows the date of publication. IEEE Xplore
`populates this information using the metadata associated with the publication
`
`445 Hoes Lane PiscatawdY, NJ 08854
`
`Aruba Networks et al. Exhibit 1004 Page 00001
`
`

`
`11. J.C. Rault, et al. "The coded orthogonal frequency division multiplexing (COFDM)
`technique, and its application to digital radio broadcasting towards mobile receivers"
`was published in IEEE Global Telecommunications Conference and Exhibition
`oCommunications Technology for the 1990s and Beyond (GLOBECOM). IEEE
`Global Telecommunications Conference and Exhibition' Communications
`Technology for the 1990s and Beyond (GLOBECOM) was held from November2T-
`30, 1989. Attendees of the conference were provided copies of the publication no
`later than the last day of the conference. The article is currently available for public
`download from the IEEE digital library, IEEE Xplore.
`
`12. I hereby declare that all statements made herein of my own knowledge are true and
`that all statements made on information and belief are believed to be true, and further
`that these statements were made with the knowledge that willful false statements and
`the like are punishable by fine or imprisonment, or both, under 18 U.S.C. $ 1001.
`
`I declare under penalty of perjury that the foregoing statements
`Executed on, ! ,l "DeC F#i&
`
`.correc
`
`Page 00002
`
`

`
`EXHIBIT A
`EXHIBIT A
`
`Page 00003
`
`Page 00003
`
`

`
`12/10/2016 The coded orthogonal frequency division multiplexing (COFDM) technique, and its application to digital radio broadcasting towards mobile receivers - 1.. .
`
`IEEE.org I IEEE Xplore Digital Library I IEEE-SA I IEEE Spectrum I More Sites
`
`Create Account 1 Personal Sign In
`
`Institutional Sign In
`
`BROWSE
`
`MY SETTINGS
`
`GET HELP
`
`WHAT CAN I ACCESS ?
`
`SUBSCRIBE
`
`Browse Conferences> Global Telecommunications Con_.
`
`The coded orthogonal frequency division multiplexing
`(COFDM) technique, and its application to digital radio
`broadcasting towards mobile receivers
`
`Sign In or Purchase
`to V1ew Full Text
`
`16
`Paper
`Citations
`
`7
`Patent
`Citations
`
`205
`Full
`Text Views
`
`3
`Author(s)
`
`J C Raul! : D. Castelain ; B.L Le Floch
`
`Related Articles
`
`Frequency-domain equalization of mobile radio
`and terrestri al broadcast channels
`
`Interpolation-based OR decomposition in MIMO(cid:173)
`OFDM systems
`
`High performance Viterbi decoder for OFDM
`systems
`
`View All
`
`View All Authors
`
`Abstract
`
`Authors
`
`Figures
`
`References
`
`Citations
`
`Keywords
`
`Metrics
`
`Media
`
`Abstract:
`A system which is able to broadcast at high data rates in a selective Rayleigh channel is presented. This technique combines an orthogonal frequency
`division multiplexing technique and a convolutional coding scheme associated with a Viterbi decoding algorithm. This system is able to benefit from
`wideband transmission by employing the information contained in all the echoes of the multipath channel while having a very good spectral efficiency
`and a low computational complexity. The authors present the theoretical principles of the system. They describe the realization of a complete COFDM
`(coded orthogonal frequency division multiplexing) system, designed within the framework of the DAB (digital audio broadcasting) EUREKA 147 project,
`which is able to broadcast at 5.6 Mb/s in a bandwidth of 7 MHz. At present, this rate corresponds to 16 high-quality stereophonic programs. Network
`aspects are pointed out as far as the introduction of a new radio broadcasting service is concerned.
`
`Published in: Global Telecommunications Conference and Exhibition 'Communications Technology for the 1990s and Beyond' (GLOBECOM), 1989.
`IEEE
`
`Date of Conference: 27-30 Nov. 1989
`
`INSPEC Accession Number: 3719102
`
`Date Added to IEEE Xp/ore: 06 August 2002
`
`DOl: 10.1109/GLOCOM.1989.64008
`
`Publisher: IEEE
`
`Download PDF
`
`Download Citations
`
`This article is only available in PDF.
`Read document
`
`View References
`
`Keywords
`
`Emal
`
`Print
`
`Request Permissions
`
`Export to Collabratec
`
`Alerts
`
`IEEE Keywords
`OFDM, Digital audio broadcasting, Radio broadcasting, Rayleigh channels, Convolutional codes ,
`Viterbi algorithm, Decoding, Wideband , Multipath channels, Computational complexity
`
`INSPEC: Controlled Indexing
`telecommunication channels, digital radio systems, frequency division multiplexing , mobile radio
`systems , radio broadcasting, radio receivers
`
`INSPEC: Non-Controlled Indexing
`7 MHz, coded orthogonal frequency division multiplexing, digital radio broadcasting, mobile receivers ,
`
`Abstract
`
`Authors
`
`Figures
`
`References
`
`Citations
`
`Keywords
`
`Back to Top
`
`http://ieeexplore.ieee.org/document/64008/
`
`1/3
`
`Page 00004
`
`

`
`12/10/2016 The coded orthogonal frequency division multiplexing (COFDM) technique, and its application to digital radio broadcasting towards mobile receivers - I
`selective Rayleigh channel, convolutional coding scheme, wideband transmission, EUREKA 147
`project, 5.6 Mbit/s
`
`..
`
`Authors
`
`J.C. Rault
`
`CCE'l'l', Cesson Sevigne, France
`
`D. Castelain
`
`CCE'|‘l', Cesson Sevigne, France
`
`B.L. Le Floch
`
`CCE'i‘|', Cesson Sevigne, France
`
`Related Articles
`
`Frequency-domain equalization of mobile radio and terrestrial broadcast channels
`H. Sari; G. Karam; I. Jeanclaud
`
`Interpolation-based QR decomposition in MIMO-OFDM systems
`D. Cescato; M Borgmann; H. Bolcskei; J. Hansen; A. Burg
`
`High performance Viterbi decoder for OFDM systems
`E. Akay; E. Ayanoglu
`
`Joint coding and decision feedback equalization for broadband wireless channels
`8. Ariyavisitakul; Ye Li
`
`Theoretical bit error rate for uncoded and coded data transmission in digital audio broadcasting
`G. Zimmermann; M. Rosenberger; S. Dostert
`
`Optimal detection when combining OFDM-CDMA with convolutional and turbo channel coding
`5. Kaiser; L. Papke
`
`Low complexity implementation of a downlink CDMA generalized RAKE receiver
`G. Kutz; A. Chass
`
`An adaptive diversity receiver for OFDM in fading channels
`SB. Bulumulla; S.A. Kassam; S.S. Venkatesh
`
`Iterative interference cancellation and decoding for a coded UWB—TH-CDMA system in multipath
`channels using MMSE filters
`A. Bayesteh; M. Nasiri-Kenari
`
`Different channel coding strategies for OFDM-CDMA
`R.A. Stirling-Gallacher; G.J.R. Povey
`
`httpzl/ieeexplore.ieee.org/documentI64008/
`
`Personal Sign In 1 Create Account
`
`Page 00005
`
`

`
`12/10/2016 The coded orthogonal frequency division multiplexing (COFDM) technique, and its application to digital radio broadcasting towards mobile receivers - I...
`
`IEEE Account
`
`Purchase Details
`
`Profile Information
`
`Need Help?
`
`» Change UsemameIPassword
`» Update Address
`
`» Payment Options
`» Order History
`» View Purchased Documents
`
`» Communications Preferences
`» Profession and Education
`» Technical Interests
`
`» US 8. Canada: +1 800 678 4333
`» Worldwide: +1 732 981 0060
`» Contact & Support
`
`About lEEEXpiore; Cuntactus l Helpl Terms of Usei Nondiscrimination Policy Silemapl Privacy &Opting Out of Cookies
`
`A not—for-profit organization, IEEE is the world's largest technical professional organization dedicated to advancing technology iorthe benefit ofhumanity
`© Copyright 2016 IEEE - All rights reserved. Use oi this web site signifies your agreement to the terms and conditions
`
`http://ieeexplore.ieee.org/documentJ64008/
`
`Page 00006
`
`

`
`THE CODED ORTHOGONAL FREQUENCY DIVISION MULTIPLEXING
`
`(COFDM) TECHNIQUE, AND ITS APPLICATION TO DIGITAL RADIO
`BROADCASTING TOWARDS MOBILE RECEIVERS
`
`J.C. RAULT, D. CASTELAIN, B. LE FLOCH
`
`CCETT (Centre Commun d'Etudes de Téléditfusion et
`Télécommunications) 35512 CESSON SEVIGNE - FRANCE
`
`Abstract
`
`towards mobile receivers,
`The broadcasting channel
`especially in a dense urban area, is particularly h0stile,which
`makes the transmission oi high bit rate very challenging.
`The conjunction of an Orthogonal Frequency Division
`Multiplexing (OFDM technique and at convolutional coding
`scheme (associate
`to a Vilerbi decoding at orithm) is a
`promising solution (COFDM) studied at CC TT.
`that
`is
`suitable to cope with such a channel.
`
`In the that part or thgcpa er, the theoretical principles oi
`the system are detail
`.
`he second part concernslthe
`realization 01 a complete COFDM system, designed within
`the framework oi t e DAB (Di
`ita Audio Broadcasting)
`EUFIEKA 14? project. which is ab e to broadcast 5.6 Mbilrs in
`a bandwidth of
`1' MHz. For the time being,
`this rate
`corresponds to 16 high quality stereophonic
`rograms.
`Finali
`. network aspects are pointed out as
`air as the
`intro uclion oi a new radio broadcasting service is
`concerned.
`
`1. Introduction
`towards mobile receivers,
`The broadcasting channel
`especially in a dense urban area. is parlicularl hostile {1} in
`tact, the presence oi muttipath propagation us to multiple
`reflections by buildings and other scattering structures
`around the vehicle. together with the electrical interferences
`arising trom domestic and industrial sources makes the
`diiiiouity to taco to is the continual variation 0 the channel
`transmission oi high bit data rate very challenging. Another
`characteristics as a result 0! the changing environment of
`the vehicle.
`the
`the paper, we recall
`In the lirst section of
`characteristics of the urban radio channel and introduce the
`problems which have to be solved in order to ensure the
`transmission of high bit rates.
`
`The second section deals with the general principles of the
`Coded Orthogonal Frequency
`ivision Multiplexing
`{COFDMl technique that we propose in order to cope with
`the multipalh propagation.
`The demodulation process and the decoding rules are
`developed in section 3, while section 4
`ives the
`perlormances oi the COFDM modulation. paiticu arty in the
`case oi the selective Flayleigh channel.
`In section 5, we present a hardware realization ol at
`corn tale CDFDM system. designed within the lrameworlr oi
`the AB (Digital Audio Broadcasting} EUREKA 14?‘
`roiecl.
`which is able to broadcast 5.6 Mbitts in a oandwi
`th of Y
`MHz. For the time being, this rate corresponds to 16 high
`quality stereophonic programs Finally, network aspects are
`oinled out as tar as the introduction of a new radio
`roadcasling service is concerned.
`
`2. Characteristics oi the urban radio channel
`
`ills previously said. the _two main characteristics of the
`radio channel towards mobile receivers are the presence of
`c anne.
`rrli1u|tipa!th propagation and the continual changing of the
`in tact, studies led to a channel model in two parts:
`- the lirst one gives the average received ener y in an
`area ol small dimensions I a few hundreds of wave englh ).
`Experimental studies have shown that in an urban area, the
`received energy lollows a log-normal distribution, of which
`the mean value is a simple iunction oi the received power,
`deduced lrom the tree space propagation.
`- the second one takes into account the combination of
`several waves, arising from specular rellections and
`received alter scattering by material structures near the
`vehicle. that cannot be considered as simple reflectors.
`
`A mathematical modeilin of this second part leads to the
`block diagram of llgure 1. w ioh represents t a channel [2].
`The terms A,lt} represent Ihe Ftayieigti process associated
`with the path i. delayed by T, and produced by scattering on
`material structure near the vehicle. Am) describes a process
`of which the spectrum is limited to the band [i°vrc,+l,;vrg]. The
`least favourable model
`that we have to consider in our
`application corresponds to the absence of a constant
`amplitude path (Tim),
`Furthermore, depending on the relation between the
`delay spread (range of values oi T over which A,{t)
`is
`essentially nonvzero) and the bandriridih considered for the
`transmission, lrequency seieclivily will or will not affect the
`received signal.
`in urban reception.
`this delay usually
`extends over several microseconds. Therefore.
`the non-
`selectivity concerns only low bit rates which cannot be
`assumed tor high quality sound broadcasting.
`ii is
`Taking into acount the above channel modelling.
`possible to represent the eilects oi the transmission by
`cornbining the channel frequency response and time
`variation til we 2). This
`two—dimerision
`iunction
`characterizes t e "selective Fla leigh channel‘ and admits a
`decomposition in surlaces oi di terent sizes :
`- the small surfaces represent the lrequency—time areas
`where the channel can be considered as locally invariant.
`
`—_ the large surtaces indicate the minimum separation lot
`which two small surlaces are statistically independent.
`This decomposition constitutes the basis of the channel
`modulation and coding mthod described in the lollowing
`sections.
`
`12.3.1.
`CH2682-3/89/0000-0428 $1.00 © 1939 IEEE
`
`Page 00007
`
`

`
`3. General principles of the COFDM system
`
`3.1. The OFDM solution to cope with the frequency
`selectivity
`
`_ The first idea of the a stern is to suppress the iniersymboi
`interference due to the requency selectivity of the channel.
`For this purpose. the information to be transmitted is split
`into a lar e number of modulated carriers. The effect of this
`process s to decrease the frequency selectivity of the
`channel on each carrier of which the bit
`late is reduced. This
`technique. called OFDM.
`is equivalent to split the time
`frequency domain into small surfaces with a dimension is
`(symbol duration) on the time axis and 1/t. on the frequency
`axis [3].
`A simple Frequency Division Multiplexing lFDMi
`technique.
`in which the spectra of the N carriers are
`separate. has two main drawbacks. the first one being a low
`spectral efficiency andthe second one bein atechnol
`ical
`difficulty in implementing a large numbr a matched ii leis
`tone for each carrier).
`As a consequence. we propose to use another solution
`which consists in tolerating an overlappin in the spectra of
`the emitted signals (figure 3]. provi ed that certain
`orthogonality conditions are satislied. which guarantee the
`absence of interference between the different carriers.[«it}_
`We define a base of N elementary orthogonal signals
`gkll), lor k=0 to N—f:
`
`gk(t) = e2"'(fo*l‘hs)‘
`for 0 S t < t
`gk(t) = 0
`otherwise
`The OFDM transmitted signal can be written :
`909 N-I
`Xi!) = Fb( 2 2 Cl.|< Qiilfitsl
`i- -nit-e
`Cjkrepresents the emitted information. having complex
`values taken from a linile alphabet depending on the chosen
`modulation. For the time being, we have concentrated our
`efforts‘ on 4rF'Sl< modulated carriers because of
`the
`sim licity and elficiency of this modulation. Nevertheless. the
`OF M system allows the use oi more sophisticated
`modulations ii the application requires it.
`The s ectrurn of the signal lends asymptotically towards
`an idea rectangular spectrum. which corresponds to a
`spectral efficiency of 2 biisisrHz for a 4-Psi-C modulation.
`Nevertheless.
`the conditions of orthci onality are no
`longer maintained at the receiver input.
`ecause of the
`inlersymbol interference in the time domain, resulting from
`the multiple paths of the channel. The implementation of ‘a
`safeguard interval before each useful symbol solves this
`prob em. The OFDM technique. by increasing the symbol
`duration proponionally to the number of carriers. permits to
`choose a safeguard interval A longer than the data spread.
`with an acceptable loss in the spectral eliicincy. herefore.
`the useful period oi the signal remains free of interference
`and the orthogonality remains perfect.
`lead to
`_ Finally.
`the OFDM solution does not
`implementation problems since the modulation and
`demodulation processin s can be carried out by fast Fourier
`transform algorithms. w ich can easily be performed with
`the avalaible digital technology [rt].
`
`3.2. The temporal coherence
`The channel variations in time are essentially due to the
`Doppiet effect. which is characterized b
`its maximum
`frequency ti”, = luvlc . The temporal co erence of the
`channel implies that the symbol must be much shorter than
`
`. The minimum value of the symbol duration being
`it it"...
`1
`the delay spread. and the speed v of the vehicle
`fixed I)
`being ixed by the service (about 200 kmrhi.
`la rriusthe
`chosen low enough to keep the Doppler distortion negligible
`(lg < 2 GHZ. see section 6 ).
`
`3.3. The channel coding
`
`The second main principle of the system is to use channel
`coding. In fact. we have yusi demonstrated that the OFDM
`technique wipes out the inicrsymool
`interference in the
`multlpath channel.
`However.
`the OFDM technique does not suppress
`ladings. As a manor of fact. the amplitude of each carrier
`generally follows a Rayleigh law.
`in such a channel. the
`decrease of the error rate as a function of Ehihla is extremely
`slow. This is why an effective channel coding system is
`essential.
`Let Us notice that the evolution of the Rayleigh law as a
`liequency~time function is relatively slow when compared to
`the density of transmitted samples on the time-frequency
`domain.
`It means that the value of one received cam ie is
`correlated to the values of
`II5
`l'I9lQl1DOLll"S.
`s a
`consequence.
`in the case of a fading. all
`the received
`samples taken into account by the decoder (which takes its
`decisions by observin a finite number of samples). can be
`considered as eras
`by the channel. Of course. this will
`lead to decision errors. whatever the coding efficiency may
`be.
`Nevertheless. an ellicient interleaving s stem (working on
`both time and frequency dimensions] aiows the received
`amplitudes to be independent from one sample to another.
`which will lead the decoder with a set of independent
`Rayleigh samples.
`In such conditions the probaoillt
`of
`receiving a group of "erased" samples at the input to
`the
`decoder decreases considerably.
`The inlerieavin depth is relative to the dimensions of the
`large surfaces o the channel re resentation of figure 2,
`Such a coding system is design
`to take benefit from the
`wideband transmission. and it can be pointed out
`that
`muliipaths as a source of frequency diversity can be
`considered as an advantage.
`
`4. Demodulation and decoding processes
`
`The signal transmitted during the time interval T‘ = t, + A
`can be written:
`
`X0) = Re Ck 92”")
`vi-ii
`where f“: fD+k/is
`.tsis the duration of the useful symbol
`and A is the duration of the safeguard interval.
`
`Ck takes its complex value in the alphabet [t+i. 1-i. -1+i,
`»1—i} for a 4-PSK modulation.
`Assuming that in safeguard interval duration at is Ion er
`than the impulse response of the channel. we can say t at
`the received si
`rial will not be affected by the lnlersymbol
`interference an thus can be written
`
`te[0.T;]
`Y(t) =Re(EHkOke2""‘|‘)
`where HR = pk ei“’ii stands for the channel frequency
`response at the frequency fk.
`The received signal ‘((1) is translated in baseband by the
`mean of protection on two quadrature carriers of frequency
`to + Ni'2t5 and then sampled at the lrequency N its =-
`l r '1‘
`
`12.3.2.
`
`Page 00008
`
`

`
`
`
`The obtained complex samples areN I
`
`I-n
`yinT) = (-1)" Z HK C.(e2"*"’"
`
`Let us write
`
`y“ = (-1 )"y(irr) I N
`
`thus we have y.. = in
`
`i Htcteztwm
`ii.o
`
`(yn) appears as the inverse discrete Fourier transform of
`(HkCk].
`(HKCK) can thus be calculated by using a Fl-T algorithm.
`we can see that.
`in the absence of noise. the emitted
`symbols can be recontructed Wlli'|0Ui‘BifOf,
`if the irequency
`response Hk of the channel can be estimated.
`As far as the choice of the coding scheme is concerned,
`convolutional codes can be considered as very interesting ii
`the condition of independence can be ensured at the input oi
`the decoder. Such a code family can take into account the
`Rayleigh law, with an aoceptab a increase in the decoding
`complexity.
`the convolutional code is associated with a
`In tact,
`maximum likelihood decoding illiterbi soil decision
`algorithm]. The required conditions of independence are
`carried out by an interleaving arren emerlt, in the time and
`frequency domains. the irequency omain being necessary
`for a fixed reception.
`Let us detail the deco-din rules. in fact, the received signal
`Y k at the instant] is distur ed by a complex gaussian noise
`hit‘ [which is not necessary white) and can be expressed as :
`Y” = H“ CH‘ + N”
`The Viterbl decoder im laments the criterion ci a
`pcsieriori maximum Iikelihoo . which consists in maximizing
`with respect to {CM} and under the code constraint the
`probability density 2
`
`H H P ({Yl.k}/ {Hi.i<i.iCi.ti)
`I
`K
`
`This leads to minimize 2 E “Y”.
`i
`k
`
`-H“. C,Vk||2 /20?”
`
`G2” being the variance of the real and imaginary parts of
`the noise.
`
`in the case oi a -tAPSK modulation. when Crka A“ + iBl.k
`{All and Blkbeing equal to +r- 1).
`the decoder ‘has to
`maximize with respect or all the values (AH, rel hi or the code:
`
`2 Z R9 (Vii Hit’ °ii<l Aik + ‘"1 (Yik H'i.k’°iiil Err
`I
`k
`
`YWH Hie?” appears to be the weighting at A _ and B” in
`the computation of the branch metrics in the \frte'rbi decoder.
`The estimation oi the channel lrequency response H H could
`be carried out by a coherent demodulation schemefbut we
`have developed another solution based on dilterentlal
`demodulation. The implementation oi this solution is very
`simple. The perlorrnance degradation is in reality slight ti
`account is taken oi
`the practicei
`limitation oi coherent
`demodulation in such a hostile channel.
`The diiterentlal demodulation consists in estimating the
`frequency response H“ by using the values oi the signal at
`the instant 1-1. which means :
`l.|(
`H - Yl_1‘k/Cm
`
`The differential encoding is written :
`
`a1_k+ i in” = (1 +i)Cl'k/ cm
`where a._ and bit equal to +r— 1 are the outputs of the
`convolutiorliat coder. The weightings of am‘ and bu in the
`branch metric oi the decoder are thus
`
`Re my v',.1,i/(1-ii of.)
`
`and
`
`Im my Y‘;.1.k/(1-i)C|EiJ
`
`5. Performances of the COFDM system
`
`Figure 4 represents the evolution oi the Bit Error iflate as a
`tunctron of the Ebililo ratio. The modulation scheme is 4-PSK
`(either coherent or differential}. We have used 3
`convolutional code with a constraint length oi ‘i’ and a rate oi
`H2. Figure 4 also(points out the results obtained when using
`a concateried co e {convolutional code + Cyclotomaticaily
`Shortened Reed Solomon (USES) code). Instead ot a
`regressive degradation. we can observe_a "virtually error
`ree“ channel. Especially it the system is used for data
`broadcasting,
`this approach is all the more interesting
`because the CSFIS decoder is able to indicate a decoding
`failure. Let Us remark that the CSRS codes are Maximum
`Distance separable [ om... - N - _K +_t ). but that they only
`require processing on the Gators Field GF(2) instead or
`GF(23) in the case of a Reed Solomon code. which reduces
`considerably their complexity. The chosen parameters In our
`application are N s 336 and N - K H 48, with 12 bit words.
`
`6. A hardware realization
`
`A hardware realization was implemented in the UHF band.
`in order to validate the COFDM principles [5]. The system is
`able to process 16 stereo programs of 283 ltBiti's (4.6 iiilbius
`including additional data].
`in a bandwidth
`of
`7 MHz in
`The lniorrnailon is transmitted on 443 carriers s aced by
`15625 Hz, each carriers being 4—PSK modulated. he total
`duration OI‘ the symbol
`is short enough to ensure the
`tent
`rat coherence oi the channel. even at a speed of 200
`it
`. for a carrier frequency of 1.5 GH:.
`in each trame ol 24 ms (300 symbols). one s mbol is
`forced to zero. which allows the noise analysis and i e frame
`synchronization, and one symbol corresponds to a fixed
`sine-sweep. which constitutes a
`hase reierence tor the
`diiierential demodulation oi the 44 carriers. Moreover. this
`sine-sweep signal allows the computation oi the impulse
`response oi the channel, which is very useful for improving
`the accuracy of the receiver synchronization.
`The binary inlorrnation to be transmitted was previously
`processed by a convolutional code oi constraint length 7 and
`rate 1J2. associated with a irenuenoy interleaving over 448
`carriers together with a lime interleavin over ‘.334 ms. and
`then differentially encoded. The ma ulation of the 448
`carriers is achieved by means oi a Fi-'|"1 algorithm.
`The receiver architecture is described in iigure 5. The
`analog i’-tFFparl is conventionai. the channel si net being
`littered in I
`by a SAW illtsr bandwidth of i'.5 Hz). Alter
`demodulation, the '|' and "
`“ si
`rials are sampled and
`converted to a digital
`term. T e FFT algorithm are
`pEIr'loIT!19€| by a ZOR N digital signal
`rccessor which is able
`to compute a 512 complex oints FF in about 1.1 ms. The
`ZOFMN DSP also deals wit
`the differential demodulation of
`the 448 carriers. Alter the desinterieaving arrangements. the
`Vireihi and the CSRS decodtn s are carried out by two
`Aslcs. deveioped b
`the SOFI P French company under
`CCETT contracts.
`he number oi operations per useiul
`transmitted hit remains relatively low when compared to
`other techniques such as equalization.
`
`12.3.3.
`
`Page 00009
`
`

`
`Additive guussion
`HDIEDC
`nld.
`
`I
`
`Hr.-ccalion_ ....
`
`Wuvc of constant amplitude
`(may not exist)
`
`Fl ure 1
`
`: Modellin oi the transmission channel
`
`cannyI'll:
`
`elementary ‘,{,“'f"f',fi,-LI,‘ V
`..f.',7.1 :..-.-.efi\ .
`In J’ 4 P."
`U
`5.5;.-. .1 5:’
`Ir .avnn".'-'-‘p 5'
`I
`.';'eJr.- ..
`
`frequency
`independence
`
`frequency
`
`aIIII
`
`‘!IIfd
`
`5I
`
`F
`
`k:
`
`Figure 3 : Spectrum 01 gk signals
`
`7. Network aspects
`
`s when
`Flexibility of a system is an important advant
`OFDM
`frequency planning must be considered. The
`system can easily be adapted to different configurations and
`several possibilities are investigated like UHF Iocai
`broadcasting (bandwidth of 4 to 7 MHz) by using TV
`channels in adjacent regions or satellite radio broadcastin
`in a frequency range of 0.5 to 2 GHz. with a nations
`coverage.
`Another configuration which is very promising is at sin Ia
`frequency network (nations: or regional coverage) in the 0-
`200 Ml-lz band. using only one 4 Hz channei. it consists in
`a‘ network of synchronized transmitters working on the same
`signal. each transmitter being considered as an active echo
`is related to the distances
`etween the transmitters.
`It is
`by any receiver. The delay sgread of the equivalent channel
`therefore necessary to impiement very long syrribots {about
`1 ms). with a saleouarcl interval able to do away with echoes
`from ice km distant transmitters. The number of carriers in a
`given bandwidth is increased but the complexity is hardly
`greater because of the efficiency of the FFT algorithms.
`
`8. Conclusion
`
`In this paper. we have presented an original 5 stern, which
`is able to broadcast high data rates in a select e Rayleigh
`channel. This technique. called COFDM.
`implements
`sophisticated processes such as Onhogonal Frequency
`Division Multiplexing and Vlterbi decodin .The s stem is
`thus able to lake benefit from the wideban transm ssion by
`turning to account
`the information contained in all
`the
`echoes of the multipath channel while having a very good
`spectral ellioiency and a tow computation complexity. The
`flexibility of the system is atso a very prornissing point. as far
`as frequency planning is concerned.
`A field demonstration of digital sound broadcasting at the
`WARC ORB 88 conference in Geneva has stron ly proved
`the complete feasibility of the GOFDM technique{t-:3.
`
`9. References
`
`1] Lee W.C.Y : Mobile communications engineering.
`Pu lished by McGraw-Hill 1982
`[2] Pommisr D., Wu Yi
`: interleaving or s eclrum
`spreading in digital radio intended for want: as. EBU
`review N°217, June 1986
`: Date transmission by
`[3] weinstain S.B., Ebert D.M.
`Frequency Division Multiplexing using the discrete
`Fourier Transform. lEEE trans. on comm. technology. V03-
`COM 19. N°15. October 1971
`[4] Alard M.. l-iaibert R.
`: Principles oi modulation and
`channel coding tor digitat broadcasting for mobile
`receivers. EBU review t~io224. August 198?
`2 5 new
`[5] Alard M.. Halbert Fl., Le Floch B,.Pommier D.
`system of sound broadcasting to mobile receivers.
`Eurocon conference 1988
`[6] Sound broadcasting lobby proves a point on a
`bus.
`Financial Times, October 7. 1988
`
`Page 00010
`
`

`
`Performances of 4PSK-COFDM
`
`with coherent demodulation
`
`: No coding
`
`: Convolutional and algebraic coding
`
`: Convolutional coding only
`2 Gaussian channel
`
`: Rayleigh channel
`
`Pertormances of 4PSK-COFDM
`
`with differential demodulation
`
`DZCDDESI.
`
`ll? CHANNEL
`SELECTION
`
`Fiure 5 : S notic diararn of the receiver
`
`12.3.5.
`
`Page 00011