INTERNATIONAL ORGANIZATION FOR STANDARDIZATION
`
`ORGANISATION INTERNATIONALE DE NORMALISATION
`
`ISO/IEC JTC1 /SC2IWG8
`
`CODED REPRESENTATION OF PICTURE AND AUDIO INFORMATION
`
`1SOIEC/JTCI/SC2/WG8
`MPEG 89/206
`October
`18 1989
`
`Source
`
`Title
`
`CCE
`IRT
`
`Philips
`
`Universal Subband Coding
`Detailed description of sound coding algorithm
`
`Status
`
`Draft
`
`Universal Subband Coding
`
`submitted on October 18 1989
`
`1
`
`HP 1026
`
`
`
`ORGANISATION INTERNATIONALE DE NORMALISATION
`
`ISO/IEC JTC1 /SC2IWG8
`
`CODED REPRESENTATION OF PICTURE AND AUDIO INFORMATION
`
`1SOIEC/JTCI/SC2/WG8
`MPEG 89/206
`October
`18 1989
`
`Source
`
`Title
`
`CCE
`IRT
`
`Philips
`
`Universal Subband Coding
`Detailed description of sound coding algorithm
`
`Status
`
`Draft
`
`Universal Subband Coding
`
`submitted on October 18 1989
`
`1
`
`HP 1026
`
`
`
`
`
`---- -- ----------1
`
`MS-M PT_ 0008722
`
`2
`
`
`
`Ii
`
`IRT
`
`PHILIPS
`
`ISO/IEC JTCi/SC2/G8
`MPEG 89/206
`
`Dear Sir
`
`Please find enclosed herewith the coimuon proposal of CCETT IRT
`scheme suitable
`for high
`bitrate reduction
`and PHILIPS for
`quality audio signals
`
`As
`explained
`corresponding
`technique
`
`individual
`in our
`algorithm is based
`
`the
`preregistration papers
`on
`32
`subband
`coding
`
`bearing
`
`in mind the requirements
`
`of
`
`The scheme has been developed
`ISO/NPEG Audio Group
`on the design of
`We have put special
`emphasis
`decoder with the following features
`low complexity
`short decoding delay
`compatibility with
`
`universal simple
`
`broad range of bitrates
`
`Additionally the compressed sound format allows
`small granularity
`random access having
`for ancillary information
`supports data capacity
`to bit errors
`and has high robustness
`
`This scheme will be supported by hardware
`having in mind
`intend to demonstrate
`integration
`
`implementation which we
`further step towards VSLI
`
`does not
`that
`follow
`We will
`policy
`licensing
`standardisation of this scheme as an ISO standard
`
`hinder
`
`the
`
`Best regards
`
`on behalf of
`
`-ee -i
`
`Ccirr Y.F Dehery
`
`IRT
`
`Stall
`
`PHILIPS
`
`Lokhoff
`
`3
`
`
`
`imr
`
`IRT
`
`PHILIPS
`
`UNIVERSAL
`
`SUBBAND CODING
`
`SYSTEM DESCRIPTION
`
`CENTRE
`
`COMMUN DTUDES
`
`DE TELEDIFFtJS ION ET TLCOMMUNICATIONS
`
`INSTITUT
`
`F(cid:220)R
`
`RUNDFUNKTECHNIK
`
`PHILIPS
`
`G.L
`
`1982-10-11
`
`CCETTIRTPHILIPS
`
`frontpage
`
`4
`
`
`
`UNIVERSAL
`CODING
`SUBBAND
`system description
`
`INTRODUCTION
`
`an audio Subband Coding system developed
`This document describes
`be used
`between CCETT IRT and Philips which can
`in cooperation
`to transfer high quality digital audio information on channels
`with limited capacity It is designed for use in different
`by ISO/MPEG Audio group e.g
`as considered
`applications
`Digital Audio Broadcasting DAB
`ISDN systems
`Audio storage
`systems etc
`recording
`digital
`System design constraints were
`layered structure
`three sample frequencies
`stereo bilingual or mono mode
`range of bitrates
`one decoder
`for
`short decoding delay
`random access capability with small granularity
`low decoder complexity
`high bit error robustness
`for ancillary information
`datacapacity
`
`BASIC PARAMETERS
`
`Audio Format
`Sample Frequency Fs
`
`48 44.1 or 32 kHz
`
`Quantization
`
`16 bits/sample
`
`Subband
`
`filter transforms
`
`Number of subbands
`
`32
`
`Sample Frequency
`
`Fs/32
`
`Bandwidth
`
`of subbands
`
`Fs/64
`
`Subband Quantiation
`
`24 bits/sample
`
`Headroom
`
`Coder
`
`Modes
`
`Scaling
`
`Quantization
`
`dB
`
`bit
`
`stereo bilingual mono
`
`range of
`to 15 bits/sample NoB
`
`to -118 dB in
`
`dB steps
`
`Quantization
`
`levels
`
`2N0B
`
`G.L
`
`1989-1011
`
`CCETTIRTPHILIPS
`
`page
`
`5
`
`
`
`UNIVERSAL
`CODING
`SUBBAND
`system description
`
`THE THREE
`
`LEVELS
`
`An application
`to the system may be composed by
`according
`grouping items from three levels Each of
`these levels contains
`algorithms for
`special purpose
`
`this is the basic level which will be used in all
`LEVEL
`the filter
`applications It contains the desctiption of
`and decoder
`transforms as well as requirements for the encoder
`the Level
`The definition of
`is also included
`format
`signal
`
`this contains additional algorithms that can be used to
`II
`LEVEL
`reduce the bitrate of
`the Level
`further
`even
`signal
`
`error detection and/or correction codes
`III
`LEVEL
`applications
`
`for different
`
`levels II and III are optional This means that
`Algorithms of
`application may include besides the level
`algorithm only
`III algorithm or both Also
`level II or level
`combination
`different algorithms from the same level may be used
`
`an
`
`of
`
`Examples may be found on the following page
`
`G.L
`
`1989-10-.l
`
`CCETTIRTPHILIPS
`
`page
`
`6
`
`
`
`basic configuration
`
`level
`
`only
`
`UNIVERSAL
`CODING
`SUBBAND
`system description
`
`reduced bitrate configuration
`
`levels
`
`and II
`
`error protected configuration
`
`levels
`
`and III
`
`full configuration
`
`levels
`
`II and III
`
`Fig
`
`examples of system configurations
`
`application specific
`
`19891011
`
`CCETTIRTPHILIPS
`
`page
`
`7
`
`
`
`UNIVERSAL
`
`SUBBAND
`
`CODING
`level
`
`LEVEL
`
`BASIC LEVEL
`
`NOTATION
`
`audio and subband sample values are considered
`All broadband
`i.e an
`be twos complement
`bit sample has
`numbers
`
`to
`
`maximum value
`
`minimum value
`
`Stated differently
`
`the separate bits have the following values
`
`MSB 4SB-l MSB2 MSB-3 MSB4
`
`etc
`
`21
`
`22
`
`The quantized samples in the signal according
`different way
`format will be expressed in
`
`to the level
`
`INPUT AND OUTPUT
`
`DATA
`
`FORMAT
`
`and output data are digital samples of an audio signal
`The input
`represented in 16 bit twos complement notation The sample
`frequency may be 32 44.1 or 48 kHz
`
`FILTER TRANSFORMS
`
`filter transforms are used to split the broadband
`signal
`Special
`at sample frequency
`Fs into 32 equal bandwidth
`subbands at sample
`frequency Fs/32 and to join the 32 subband signals to regenerate
`the broadband signal The subband samples will be represented in
`24 bit twos complement notation
`
`G.L
`
`1989-10-il
`
`CCETTIRTPHILIPS
`
`page
`
`8
`
`
`
`SPLIT TRANSFORH
`
`UNIVERSAL
`
`SUBBAND
`
`CODING
`level
`
`or every calculation
`samplas one
`of one set of 32 subband
`vector of
`input samples is multiplied
`subband
`sample for each
`transform matrix
`with
`The input sample vector
`
`is
`
`Mi Element
`
`column vector of
`elements
`is the top element element Hi is
`the bottom element Every calculation requires 32
`input samples
`to 31 while the 32 bottom
`to be shifted in at positions
`elements are shifted out The vector
`thus operates as
`shifting
`block on the inputs samples with the eldest sample at position
`Hi and the newest at position
`32 by matrix with elements
`The transform matrix
`Hi These elements are constants
`is
`O..31
`samples are collected in vector B11
`The subband
`th lower
`index elements representing
`with the lower
`index elements
`and the higher
`representing
`components
`components
`frequency
`
`rc
`31
`frequency
`the higher
`
`S.A
`
`H-i
`
`cO
`
`Sk
`
`in matrix notation or
`
`for
`
`31
`
`Bk
`
`JOIN TRANS FORN
`
`is
`
`transform
`
`from the subband signals
`can be regenerated
`The broadband
`signal
`samples is performed
`of 32 broadband
`by
`calculation
`vector of subband samples with
`multiplication of
`matrix
`column vector of
`elements
`sample vector
`The subband
`Hi Element
`is the top element element Mi
`is
`te bottom element Every calculation requires 32 subband samples
`to 31 while the 32 bottom
`to be shifted in at positions
`elements are shifted out The vector
`of sets
`combination
`thus is
`components at
`of 32 subband samples with the lowest
`frequency
`and the highest
`mod 32
`frequency
`element positions
`31
`mod 32
`components at element positions
`32 by matrix with elements
`The transform matrix
`is
`M13 These elements are constants
`.31
`samples are collected in vector E11
`The broadband
`is frm to 31
`the signal
`sequence of
`The correct output
`
`31
`
`S.D
`
`M-1
`
`c0
`
`in matrix notation or
`
`.D
`
`for
`
`31
`
`G.L
`
`19891011
`
`CCETTIRTPHILIPS
`
`page
`
`9
`
`
`
`UNIVERSAL
`
`SUBBAND
`
`CODING
`level
`
`ENCODING
`
`DECODING
`
`THE ENCODING PROCESS
`
`is fixed except
`for the allocation
`The level
`encoding process
`of bits to suiDbands Obviously the encoding for bilingual or mono
`mode differs in details from the stereo encoding
`
`the mono
`In this chapter we only give
`simple description of
`In this description and in the chapters on scaling and
`encoding
`quantizing we try to provide enough
`information to enable
`to evaluate the encoding for the other modes
`reader
`
`signal carries information on 384
`the level
`One frame of
`stthband samples 12 samples of each subband For every subband
`the sample with the highest amplitude is used to find the
`for that subband
`scalefactor
`Then the samples are scaled with
`the scalefactor
`so that
`they are normalized with respect
`to the
`highest amplitude
`
`to provide the
`the encoding algorithm will have
`One part of
`information table We do not consider
`this here as it
`allocation
`this table in various ways The simplest
`is possible to generate
`algorithm uses an allocation information table in which the
`fixed allocation However
`entries are fixed
`this information
`may also change with every frame adaptive allocation
`
`reduces the number
`the samples are quantized The quantizer
`Next
`of bits of
`in the allocation
`the samples to the number
`information table Its output
`special notation
`is in
`
`Finally the sync and system information is added
`information is formatted into the right sequence
`signal
`
`and all
`for
`Level
`
`allocation
`info
`
`subband saiqples
`24 bit
`
`scale
`samples
`
`quan
`tizer
`
`_____________
`find
`scalefactor
`
`L.
`
`sync
`
`sys info
`
`Fig
`
`the encoding process
`
`for
`mat
`ter
`
`output
`
`G.L
`
`198910-11
`
`CCETTIRTPHILIPS
`
`page
`
`10
`
`
`
`UNIVERSAL
`
`SUBBAND
`
`CODING
`level
`
`THE DECODING PROCESS
`
`in the same way The
`is always performed
`The decoding process
`and the
`information is carried in the scalefactors
`essential
`subband samples All other information is just included in order
`to be able to decode correctly
`
`frame After the
`The decoding process will be repeated with every
`information
`and System info the allocation
`the Sync
`reception of
`is used to fill the scalefactor array correctly
`array is an array with two columns of 32 rows
`The scalefactor
`bits wide Each column corresponds with one of
`the output
`the 32 subbands The received
`row with one of
`each
`channels
`are stored in this array The location of
`scalefactors
`transferred will become 111111
`that has not been
`scalefactor
`for that subband will all be
`the outputsamples
`that
`indicating
`101
`
`for the two
`In stereo and bilingual mode the scalefactors
`the array In mono mode
`channels are stored in the two columns of
`the array
`is stored in both columns of
`the same scalefactor
`
`fixed table This
`are used as an index into
`The scalefactors
`factors with which the subband
`the multiplication
`table contains
`to the join
`samples have to be multiplied before they are sent
`the ones in the
`factors are the reciprocals of
`transform These
`encoder
`
`input
`
`de
`for
`mat
`ter
`
`sync
`
`sys info
`
`alloc.info
`RAN
`
`scale fac
`array
`
`quantized
`samples
`
`de-
`quantizer
`
`DECODER
`CONTROL
`
`to .P
`
`out
`
`multiplication
`factor
`table
`
`multiplier
`
`Fig
`
`the decoding process
`
`out
`
`G.L
`
`1989-1011
`
`CCETTIRTPHILIPS
`
`page
`
`11
`
`
`
`UNIVERSAL
`
`SUBBAND
`
`CODING
`level
`
`LEVEL
`
`SIGNAL FORMAT
`
`SLOTS and FRAMES
`
`The format supplied to the decoder will be constructed using
`slots each slot will carry 32 data bits
`
`III1tIIIII.IIIIIBIIIIFIIIIIIIIIiJIIIIIItIf
`
`11111111
`
`IIIIIIl1II11
`
`slot
`32 bits
`
`Fig
`
`slot definition
`
`The bits in
`first
`
`slot are numbered
`
`to 31 Bit
`
`is transferred
`
`1111111111222222222233
`01234567890123456789012345678901
`
`Fig
`
`numbering of bits in
`
`slot
`
`time
`
`information is transferred in frames The
`subband
`The encoded
`frame varies with the sample frequency Fs
`number of slots in
`and bitrate Each frame contains information on 384 samples of
`the original
`input signal Thus
`the frame rate is Fs/384
`frame size msec
`
`Fs kHz
`
`48
`44.1
`32
`
`8.7
`12
`
`frame may carry audio information from one or two channels
`to mono or stereo/bilingual
`modes
`corresponding
`
`The number of slots in
`formula
`
`of slots/frame
`
`frame can be found with the following
`
`bitrate
`
`Fs
`
`12
`
`If this does not give an integer number
`the result is truncated
`and padding is required This means that
`the number of slots
`and
`may vary between
`
`G.L
`
`1989-1011
`
`CCETTIRTPHILIPS
`
`page
`
`12
`
`
`
`UNIVERSAL SUBBAND
`
`CODING
`level
`
`the number of
`For four bitrates and three sampling frequencies
`be found in the following table
`slots in
`frame can
`
`itrate
`kbits/s
`
`sample freq
`kHz
`
`slots in
`
`frame
`
`96
`
`128
`
`192
`
`256
`
`384
`
`32
`44.1
`48
`
`32
`44.1
`48
`
`32
`44.1
`48
`
`32
`44.1
`48
`
`32
`44
`48
`
`36
`26
`24
`
`48
`34
`32
`
`72
`52
`48
`
`96
`69
`64
`
`144
`104
`96
`
`padding
`
`padding
`
`padding
`
`padding
`
`padding
`
`Table
`
`examples of bitrates and number of slots
`
`consists of adding
`The padding for the 44.1 kHz sample frequency
`frame This slot contains no
`an extra dummy slot to
`frame rate to
`is added
`to adjust
`the average
`information but
`rate of Fs/384 For the different bit rates the following rules
`apply
`
`bitrate
`kbits/s
`
`frames in
`total
`padding sequence
`
`frames with
`dummy slot
`
`96
`
`128
`
`192
`
`256
`
`384
`
`49
`
`147
`
`49
`
`147
`
`49
`
`122
`
`12
`
`97
`
`24
`
`Table
`
`examples of padding sequence variables
`
`G.L
`
`19891011
`
`CCETTIRTPHILIPS
`
`page
`
`13
`
`
`
`UNIVERSAL
`
`SUBBAND
`
`CODING
`level
`
`the decoder will be able to process signals with the
`However
`following bitrates most of which are an integer multiple of 32
`kbits/sec
`
`bitrate index sample frequenc
`48 kHz
`kbits/
`sec
`slots
`
`44.1 kHz
`slots
`
`32 kHz
`slots
`
`seetext
`
`32
`64
`96
`128
`160
`192
`224
`256
`288
`320
`352
`384
`416
`448
`
`16
`24
`32
`40
`48
`56
`64
`72
`80
`88
`96
`104
`112
`
`10
`11
`12
`13
`14
`
`17
`26
`34
`43
`52
`60
`69
`78
`87
`95
`104
`113
`121
`
`12
`24
`36
`48
`60
`72
`84
`96
`108
`120
`132
`144
`156
`168
`
`padding
`
`Table
`
`bitrates index and number of slots in
`
`frame
`
`The index indicates the bitrate Together with the sample
`indication
`it is used as an address
`table to find
`in
`frequency
`the number of slots in the frame
`indicated
`the exact bitrate is not
`If the bitrate index is
`So other bitrates can be used This is called free format
`
`frame are numbered
`The slots in
`and bitrate
`frequency
`
`to
`
`varies with sample
`
`SLOT
`SLOTS
`SLOTS Q1
`
`SYNC and SYSTEM INFO
`ALLOCATION
`INFORMATION
`SCALE FACTORS and SAMPLES
`2-Q
`
`Q1
`
`SYNC
`ALLOCATION
`SYS INFO INFORMATION
`
`SCALE
`FACTORS
`
`SAMPLES
`
`DUMMY
`SLOT
`
`fixed length
`
`variable length
`
`Fig
`
`USCS
`
`frame format
`
`is
`
`function of
`
`the operation mode
`
`G.L
`
`1989-10-11
`
`CCETTIRTPHILIPS
`
`page
`
`10
`
`-I-
`
`14
`
`
`
`SYNC and SYSTEM INFO
`
`UNIVERSAL SUBBAND
`
`CODING
`level
`
`and system information
`Slot will carry the synchronization
`signal
`the bits in the slot
`of
`the allocation
`figure shows
`The next
`1111111
`IllIlIllilIlIll
`bitrate Fs
`SIGNAL
`Program
`information
`111111
`index
`1111111
`III
`...311
`...15116...19120...23124..
`
`SYNCHRONIZATION
`16 bits
`II III
`
`0..
`
`Fig
`
`Bits
`
`information carried in slot
`
`to 15
`
`synchronization
`
`signal
`
`Bits 16 to 23 contain frame format
`
`information
`
`Bits 16 to 19
`
`bitrate index
`
`this index indicates the bitrate It is
`in Table
`As shown
`represented in binary notation NSB first
`index indicates the free format condition
`in which
`The all
`indicated and the decoder must rely only on
`the bitrate is not
`new frame
`to find the start of
`the synchronization
`signal
`
`Bits 20 and
`
`21
`
`Sample Frequency
`
`00
`01
`10
`
`44.1kHz
`kHz
`48
`kHz
`32
`reserved
`
`Bit 22
`
`padding bit
`
`This bit
`is
`it will be
`
`if the frame contains
`
`dummy slot Otherwise
`
`Bit 23
`
`Additional
`
`info
`
`info
`no additional
`info
`additional
`format tb.f
`
`G.L
`
`19891011
`
`CCETTIRTPHILIPS
`
`page 11
`
`15
`
`
`
`Bits 24 to 31 contain information related to the program
`
`UNIVERSAL
`
`SUBBAND
`
`CODING
`level
`
`MODE
`
`Switch
`
`24
`
`25
`
`26
`
`27
`
`28
`
`0/al Emh
`29
`30
`31
`
`Fig
`
`information carried in slot
`
`Bits 24 and 25
`
`Mode
`
`indication
`
`stereo
`stereo irrelevance
`bilingual
`mono
`
`11
`
`These modes will be explained in the following chapters
`
`Bits 26 and 27
`
`stereo irrelevance coding
`
`in which an additional
`These bits are used to indicate the modes
`bitrate reduction
`can be obtained by using stereophonic
`irrelevance
`If the bits 24 and 25 are not equal
`
`these bits will be
`
`to
`
`Bit 28
`
`copyright
`
`Bit 29
`
`Original/Home
`
`Copy
`
`Bits 30 and 31
`
`Emphasis
`
`no copyright
`copyright protected
`
`Copy
`original
`
`no emphasis
`50/15
`jsec emphasis
`reserved
`CCITT J.17
`
`G.L
`
`198910il
`
`CCETTIRTPHILIPS
`
`page
`
`12
`
`16
`
`
`
`UNIVERSAL
`
`SUBBAND
`
`CODING
`level
`
`ALLOCATION
`
`INFORMATION
`
`to 31 Subband
`will be
`The system uses 32 subbands numbered
`the original
`components of
`the subband with the lowest
`frequency
`broadband signal subband 31 with the highest
`frequency
`components
`The number of bits allocated to the samples of each
`bit binary number MSB first
`as
`encoded
`
`subband will be
`
`The number of slots for allocation information depends on the
`operation mode
`
`slots
`
`mode
`
`stereo
`
`bilingual
`
`mono
`
`Table
`
`number of allocation slots
`
`G.L
`
`19891011
`
`CCETTIRTPHILIPS
`
`page
`
`13
`
`17
`
`
`
`UNIVERSAL
`
`SUBBAND
`
`CODING
`level
`
`SCALE FACTORS and SANPLES
`
`For the transfer of scalefactors and subband
`samples the bits in
`pool of bits As the number of
`the slots are considered
`to be
`sample may vary it is possible that
`bits in
`sample will be
`carried in the last bits of one slot and the first of
`the next
`slot
`
`bits MSB first All
`Scale factors will be represented with
`scale factors are transferred before the transfer of
`the samples
`starts
`
`The maximum number of scalefactors is 64 However if the
`subband is 0000 indicating no
`allocation
`information of
`of any bits to the subband there is no need
`allocation
`for the
`transfer of
`the number of
`the scalef actor and samples Then
`scalefactors will be less
`
`The order in which the scalefactors are transferred is the same
`as the allocation
`that scalefactors
`information sequence
`except
`information equal
`for subbands with allocation
`to 0000 are not
`transferred
`
`The samples will be transferred in the same order as the
`time
`allocation
`subband at
`information one sample of each
`This sequence will be repeated
`12 times in order to transfer all
`subband
`12 samples of
`
`the frame is used the
`the sample capacity of
`If not all of
`in the case of additional
`remaining bits will be
`except
`is indicated by the Al bit in the system info
`information
`which
`of slot
`
`G.L
`
`1989-10-il
`
`CCETTIRTPHILIPS
`
`page 14
`
`18
`
`
`
`UNIVERSAL
`
`SUBBAND
`
`CODING
`level
`
`SCALING
`
`to scale the 12 samples of one
`In order to find the scalefactor
`subband the following algorithm is used
`
`At first one needs to find the sample with the maximum amplitude
`This is done by comparing the absolute values of
`the 12 samples
`Lets call
`this maximum value
`
`table of values at
`This maximum value is compared with
`table
`that
`dB intervals The smallest value of
`approximately
`is used to normalize the samples
`than ISmaxi
`that
`is larger
`
`SI
`
`max 2dB
`
`for the remainder of
`
`this text SF
`
`indicates that
`
`this value is different
`
`max 12dB
`for all subbands
`
`QUANTI
`
`ING
`
`value between
`to have
`The scaled subband samples are considered
`The samples will be quantized linearly to the number
`and
`information table
`of bits as specified in the allocation
`
`DE-SCALING
`
`to be
`the dequantized samples will have
`In the decoder
`the scalefactor
`multiplied using the reciproke value of
`reconstruct
`the original signal
`
`to
`
`SF
`
`GL
`
`198910li
`
`CCETTIRTPHILIPS
`
`page
`
`15
`
`19
`
`
`
`UNIVERSAL
`
`SUBBAND
`levels II
`
`CODING
`III
`
`LEVEL
`
`II
`
`ADDITIONAL
`
`BITRATE
`
`REDUCTION
`
`are available to reduce the bitrate of LeveJ
`Several
`techniques
`further Each of
`these can be used independently or
`signals even
`in combination with others
`
`these techniques
`Some of
`are based on
`further
`reduction of subband sample flow e.g using Huffman
`coding to the price of an additional delay and error
`sensitivity
`reduction of allocation information and scalefactors by using
`the sound
`the stationarity of
`
`LEVEL
`
`III
`
`ERROR DETECTION AND CORRECTION
`
`The binary information generated by the algorithms described in
`levels
`and
`II can be protected by appropriate error detection
`and correction schemes according
`to the error distribution of
`channel or recording medium
`
`the
`
`information is
`limited amount of additional
`Nevertheless
`only
`level
`to efficiently protect
`coded signals In
`necessary
`effect
`subband coding is characterized
`low
`by an inherent
`sensitivity to transmission errors because
`the spectrum of
`the
`noise enery is confined to
`single subband and is limited by the
`scalefactor
`if this one and the allocation
`information are
`satisfactorily protected
`
`G.L
`
`19891011
`
`CCETTIRTPHILIPS
`
`page 16
`
`20
`
`
Accessing this document will incur an additional charge of $.
After purchase, you can access this document again without charge.
Accept $ Charge
Still Working On It
This document is taking longer than usual to download. This can happen if we need to contact the court directly to obtain the document and their servers are running slowly.
Give it another minute or two to complete, and then try the refresh button.
A few More Minutes ... Still Working
It can take up to 5 minutes for us to download a document if the court servers are running slowly.
Thank you for your continued patience.
This document could not be displayed.
We could not find this document within its docket. Please go back to the docket page and check the link. If that does not work, go back to the docket and refresh it to pull the newest information.
Your account does not support viewing this document.
You need a Paid Account to view this document. Click here to change your account type.
Your account does not support viewing this document.
Set your membership
status to view this document.
With a Docket Alarm membership, you'll
get a whole lot more, including:
- Up-to-date information for this case.
- Email alerts whenever there is an update.
- Full text search for other cases.
- Get email alerts whenever a new case matches your search.
One Moment Please
The filing “” is large (MB) and is being downloaded.
Please refresh this page in a few minutes to see if the filing has been downloaded. The filing will also be emailed to you when the download completes.
Your document is on its way!
If you do not receive the document in five minutes, contact support at support@docketalarm.com.
Sealed Document
We are unable to display this document, it may be under a court ordered seal.
If you have proper credentials to access the file, you may proceed directly to the court's system using your government issued username and password.
Access Government Site
