Note : Les descriptions sont présentées dans la langue officielle dans laquelle elles ont été soumises.
PU070283. CA 02704485 2013-11-01
1
CODE ENHANCED STAGGERCASTING
Field of the Invention
The present invention relates to transmitting data in a multimode
transmission system. In particular, the present invention relates to a
transmission
system wherein multiple code rates can be used in data transmission within a
single standard transmission protocol, such as ATSC.
Background of the Invention
Over the past decades, video transmission systems have migrated from
analog to digital formats. In the United States, broadcasters are in the final
stages
of completing the switch from the National Television System Committee (NTSC)
analog television system, to the Advanced Television Systems Committee (ATSC)
A/53 digital television system. The A/53 standard provides "specification of
the
parameters of the system including the video encoder input scanning formats
and
the preprocessing and compression parameters of the video encoder, the audio
encoder input signal format and the pre-processing and compression parameters
of the audio encoder, the service multiplex and transport layer
characteristics and
normative specifications, and the VSB RF/Transmission subsystem." The A/53
standard defines how source data (e.g., digital audio and video data) should
be
processed and modulated into a signal that is to be transmitted over the air.
This
processing adds redundant information to the source data so that a receiver
may
recover the source data even if the channel adds noise and multi-path
interference
to the transmitted signal. The redundant information added to the source data
reduces the effective rate at which the source data is transmitted, but
increases
the potential for successful recovery of the source data from a received
signal.
CA 02704485 2010-04-30
WO 2009/064468
PCT/US2008/012795
2
The ATSC A/53 standard development process was focused on HDTV and fixed
reception. The system was designed to maximize video bit rate for the large
high
resolution television screens that were already beginning to enter the market.
Transmissions broadcast under the ATSC A/53 standard, however, present
difficulties
for mobile receivers. Enhancements to the standard are required for robust
reception of
digital television signals by mobile devices.
Recognizing this fact, in 2007, the ATSC announced the launch of a process to
develop a standard that would enable broadcasters to deliver television
content and data
to mobile and handheld devices via their digital broadcast signal. Multiple
proposals
were received in response. The resulting standard, to be called ATSC-M/H, is
intended
to be backwards compatible with ATSC A/53, allowing operation of existing ATSC
services in the same RF channel without an adverse impact on existing
receiving
equipment.
Many systems for transmission to mobile devices, such as some proposed
ATSC-M/171 systems, perform periodic transmission. Such systems can include a
preamble in their transmissions in order to assist with receiver system
operation.
Preambles typically include known information that portions of the receiving
system may
use for training to improve reception, which can be particularly useful in
difficult
environments such as those found in mobile operation. Such systems may further
encode data at differing code rates. The code rate or information rate of a
forward error
correction (FEC) code, for example a convolutional code, states what portion
of the total
amount of information that is non redundant. The code rate is typically a
fractional
number. If the code rate is k/n, for every k bits of useful information, the
coder generates
totally n bits of data, of which n-k are redundant.
The existing ATSC M/H proposal includes the use of separable block codes to
allow for code-enhanced time and frequency diversity. In the example of a 1/2
rate coded
transmission, the mobile data is input into a FEC coder which outputs 2 bytes
for each
input byte. The two bytes represent the original data and redundant data. A
receiver
CA 02704485 2015-09-14
PU070283
3
can receive either the original data or the redundant data at the receiver
threshold of the original data. If both streams are received there is a coding
gain advantage so that the receiver will recover the data at a threshold below
the original data. Mobile and
pedestrian operation of communications
equipment pose some of the greatest challenges, with extreme transmission
channel impairments present due to buildings and moving vehicles as well as
other impairments. A system of providing data in a redundant manner may be
used. It would be desirable to take advantage of the redundant information
through frequency, time and special diversity to improve reception on mobile
devices.
Summary of the Invention
In one aspect of the invention, there is provided a method for encoding data
comprising the steps of: encoding the data to generate a packet and a
redundant packet, each packet comprising the data; and coupling the packet
and the redundant packet to a transmitter. The packet and the redundant
packet may be coupled to different transmitters; the different transmitters
may
be spatially diverse. The packet and
the redundant packet may be
transmitted at different frequencies. The packet and the redundant packet
may be transmitted at different times.
In another aspect of the invention, there is provided a method of receiving a
signal comprising the steps of: receiving a first packet; receiving a second
packet, the second packet being a redundant version of the first packet;
combining the first and second packets; decoding the first and second
packets. The first packet may be received at a first frequency and the second
packet may be received at a second frequency. The first packet may be
received at a first time, and the second packet may be received at a second
time.
In another aspect of the invention, there is provided an apparatus comprising
an encoder for encoding a data to generate a packet and a redundant packet;
and an interface for coupling the packet to a first transmitter and the
redundant packet to a second transmitter. The packet may be transmitted at
CA 02704485 2015-09-14
PU070283
3a
a first frequency and the redundant packet may be transmitted at a second
frequency. The packet may be transmitted at a first time and the redundant
packet may be transmitted at a second time.
In another aspect of the invention, there is provided an apparatus comprising:
an interface for receiving a first packet and a second packet; a decoder for
decoding the first packet and the second packet to generate a first decoded
packet and a second decoded packet; a processor for combining the first
decoded packet and the second decoded packet to generate a video signal.
The second packet may be a redundant version of the first packet. The first
packet may be received on a first frequency and the second packet may be
received at a second frequency. The first packet may be received at a first
time and the second packet may be received at a second time.
Description of the Drawings
FIG. 1 is a block diagram of an embodiment of a terrestrial broadcast
transmitter for mobile/handheld reception of the present disclosure;
FIG. 2 is a block diagram of an embodiment of a portion of an exemplary
mobile/handheld data stream of the present disclosure;
FIG. 3 is a block diagram of an embodiment of an exemplary data frame of the
present disclosure;
FIG. 4 is a block diagram of an embodiment of a terrestrial broadcast receiver
for mobile/handheld reception of the present disclosure;
FIG. 5 is a block diagram of an embodiment of a decoder of the present
disclosure;
FIG. 6 is a block diagram of another embodiment of a decoder of the present
disclosure;
FIG. 7 is a block diagram of a terrestrial broadcast environment according to
the present invention;
FIG. 8, is a block diagram of an embodiment of a portion of a transmitter
according to the present disclosure.
CA 02704485 2015-09-14
PU070283
4
The exemplifications set out herein illustrate preferred embodiments of the
invention, and such exemplifications are not to be construed as limiting the
scope of the invention in any manner.
Description of the Preferred Embodiment
As described herein, the present invention provides a method and
apparatus for transmitting data in a mobile broadcast system utilizing
diversity
and data redundancy, such as a proposed ATSC-M/H system, while allowing
backward compatibility with legacy transmission and reception paths, such as
ATSC A/53. While this invention has been described as having a preferred
design, the present invention can be further modified within the scope of this
disclosure. This application is therefore intended to cover any variations,
uses, or adaptations of the invention using its general principles. Further,
this
application is intended to cover such departures from the present disclosure
as come within known or customary practice in the art to which this invention
pertains. For instance, the described technique could be applicable to
transmission systems designed for other types of data or that use different
coding, error-correction, redundancy, interleaving, or modulation schemes.
Referring now to the drawings, and more particularly to FIG. 1, a block
diagram of an embodiment of a terrestrial broadcast transmitter for
mobile/handheld reception of the present disclosure is shown. Embodiment
100 of FIG. 1 comprises a plurality of signal transmitting means such as an
MPEG Transport stream source 110, an ATSC M/H preprocessing path 115,
and a legacy ATSC A/53 processing path. The elements within the ATSC-
M/H preprocessing 115 comprise a packet interleaver 120, a serial
concatenated block coder 125, a packet deinterleaver 130, an MPEG
transport stream header modifier 135, a preamble packet inserter 140. The
legacy ATSC A/53 processing path 145 comprises a data randomizer 150, a
reed Solomon encoder 155, a byte interleaver 160, a trellis encoder 165, a
sync inserter 170, a pilot inserter 175 and a modulator 180.
CA 02704485 2010-04-30
WO 2009/064468
PCT/US2008/012795
In the ATSC-M/H preprocessing flow, incoming MPEG transport data 112 from an
MPEG transport stream source 110 is received at the packet interleaver 120.
The
packet interleaver 120 rearranges a sequenced number of bytes into a different
5 sequence to improve bit error rate and frame error rate performance. In
this exemplary
embodiment, the packet interleaver 120 takes the bytes from a fixed number of
consecutive packets in a row by row order, and outputs the bytes column by
column. In
this way, all of the first bytes of the packets are grouped together, all of
the second bytes
of the packets are grouped together, and so on until the last bytes of the
packets. Each
source packet is an MPEG transport stream packet with the sync byte removes,
so each
packet length is 187 bytes. The number of packets in each code frame is the
same as
the number of source symbols required for the GF(256) serial concatenated
block code.
The interleaved data is then coupled to the Galois Field GF(256) serial
concatenated block coder (SCBC) 125. The GF(256) Serial Concatenated Block
Code
(SCBC) 125 Decoder will take on different forms depending on the Rate Mode for
the
current symbols. In general, it consists of constituent decoders that
iteratively decode
soft information in a turbo decoding manner. The SCBC 125 codes the packet
interleaved data in one of a plurality of forms depending on the desired data
rate and the
codeword length. The SCBC 125 consists of one or more constituent GF(256)
codes
cascaded in a serial fashion, linked by GF(256) code optimized block
interleavers to
improve overall code performance. This may be optionally followed by a GF(256)
puncture to achieved desired codeword length.
Specifically, a Galois Field GF(p") is a mathematical set comprising a finite
number of elements pn where the values of p and n are integers. A particular
Galois
Field is defined using a generator polynomial g(x). Each element of the Galois
Field
may be represented by a unique bit pattern having n bits. Furthermore, a
unique
polynomial of degree pn may be associated with each element where each
coefficient of
the polynomial is between 0 and p-1. Further, mathematical operations in the
Galois
Field have important properties. Addition of two elements of the Galois Field
GF(pn) is
defined as an element associated with a polynomial that has coefficients that
are the
CA 02704485 2010-04-30
WO 2009/064468 PCT/US2008/012795
6
modulo-p sum of the coefficients of the polynomials associated with the two
elements
being added. Similarly, multiplication of two elements is defined as the
multiplication of
the polynomials associated with the two elements modulo the generator
polynomial g(x)
associated with the Galois Field. Addition and multiplication operators are
defined over
the Galois Field such that the sum and product of any two elements of the
Galois Field
are elements of the Galois Field. A property of the Reed-Solomon codeword is
that
multiplying each byte of the codeword by an element of the Galois Field
results in
another valid Reed-Solomon codeword. Furthermore, byte-by-byte addition of two
Reed-Solomon codewords produces another Reed-Solomon codeword. The legacy A53
standard defines a 256 element Galois Field GF(28) and the associated
generator
polynomial g(x) for use in the Reed-Solomon algorithm. The properties of the
Galois
Field also create the ability to generate syndromes for the codewords in order
to
determine errors. Another important property of the codeword
In an exemplary embodiment, two codewords or packets are generated by a rate
1/2 byte-code encoder include a duplicate of the originally input codeword and
a new
codeword that provides redundancy to the original codeword. The two codewords
may
also be described as systematic data and non-systematic data. It is important
to note
that the codewords representing systematic and non-systematic data may be
arranged
to form larger data structures. In a preferred embodiment, the codewords may
be
organized to form a rugged data stream of data packets. The rugged data stream
includes systematic packets, which are duplicates of the data packets in a
stream
portion A, and non-systematic packets generated by the processing of a byte-
code
encoder in a stream portion A'. Non-systematic packets also include packets
that may
be derived from other systematic and non-systematic packets of the rugged data
stream.
Further, the packets in the rugged data stream may be further composed of
systematic
bytes and non-systematic bytes. In such embodiments, a systematic byte is a
duplicate
of byte of content data and a non-systematic byte is one that is derived from
other
systematic and non-systematic bytes.
The redundant or non-systematic codeword or packet output by a byte-code
encoder is the result of multiplying each byte of the incoming codeword or
packet by an
CA 02704485 2010-04-30
WO 2009/064468 PCT/US2008/012795
element b of the Galois Field GF(256). In one embodiment, if the MPEG
transmission
source 110 generate a message M, which is comprised of bytes M(1), M(2),
...,M(187),
where M(1) is the first byte of the message, M(2) is the second byte of the
message,
etc., then subsequently, the byte-code encoder 104 produces the codewords A
and A'
from the codeword M, as follows:
A(i) = M(i) 1=1,2, ...,187 (1)
A'(i) = b * M(i) 1=1,2, ...,187 (2)
The value b is a predetermined (non-zero) element of the same Galois Field
GF(256) that may used by the Reed-Solomon encoder 155.
In an illustrative
embodiment, the value of the b element is 2. It should be apparent that using
the same
Galois Field for both the byte-code encoder and the Reed-Solomon encoder
allows
operations between the two encoders based on the properties of the Galois
Field. Byte-
code encoder 125 encodes all of the bytes of the data packet, including the
bytes that
form the header containing the PID, to generate one or more non-systematic
packets of
the rugged data stream. Thus, the PID of each non-systematic packet is byte-
code
encoded and may no longer represent a PID value that is recognizable to a
receiving
device.
It should be apparent that any packets encoded by the embodiment of the
transmitter depicted by encoder 100 may be decoded by an embodiment of a
decoder
used in a legacy receiver that complies with the A53 standard. The decoder in
a legacy
receiver provides packets of the rugged data stream to a data decoder. The
rugged
data stream includes non-systematic packets that are encoded using a byte-code
encoder that will be decoded correctly by a decoder in a legacy receiver, but
will result in
data content that is unrecognizable by the legacy receiver. However, because
such
packets have a PID that is not associated in the Program Map Table (PMT) with
an
existing or legacy data format, the content decoder in a legacy receiver
ignores these
non-systematic packets of the rugged data stream.
CA 02704485 2010-04-30
WO 2009/064468
PCT/US2008/012795
8
Byte-code encoder 125 uses equation (2) above to generate a non-systematic
packet for each systematic packet and provides both packets to the legacy 8-
VSB
encoder for transmission to produce an encoded stream with an effective data
rate of 1/2
(that is, 1 byte in, 2 bytes out). As mentioned earlier, byte code encoder 125
may be
capable using other encoding rates to produce other effective data rates. In
some
embodiments, the byte-code encoder may produce one byte-encoded packet for
every
two source packets, MA and MB, received from the MPEG TS Source 110 to
generate a
rate 2/3 rugged data stream comprising the two systematic packets and one non-
systematic packet calculated as follows:
MAO) = MA(i)*1), + M8(i)*b2 i=1,2,...,187 (3)
where MA and MB are consecutive systematic packets produced by the data
generator 102 and b1 and b2 are predetermined elements of a Galois Field, such
as the
Galois Field used by the Reed-Solomon encoder 155. In an illustrative
embodiment, the
value of the bl and b2 elements is 2. In some embodiments, the values of ID,
and b2 may
not be identical. The byte-code encoder 125 provides the packets MA, MB, and
MAB to
the legacy 8-VSB encoder 130 for further encoding and transmission.
Byte-code encoder 125 may use different coding rates to produce rugged data
streams (i.e., ones having lower data rates) by including additional input
data packets for
generating the redundant packets. Another embodiment of the byte-code encoder
125
produces a rate 4/9 data stream by employing four systematic packets MA, MB,
MC, and
MD from MPEG TS Source 110 and 5 non-systematic packets calculated as follows:
MA8(i) = MA(i)101 + ME3(i)*b2 = i=1,2,...,187 (4)
Mco(i) = Mc(i) *b3 Mo(i)*b4 i=1,2,...,187 (5)
MAC(i) = MA(i)*los + MD(i)136 i=1,2,...,187 (6)
MBD(i) = M8(i)*b7 + MD(i)*b8 i=1,2,...,187 (7)
IVIABcD(i) MAB(irbs MCD(ir010 i=1,2,...,187 (8)
CA 02704485 2010-04-30
WO 2009/064468 PCT/US2008/012795
9
The values 131, b2,..., blo are predetermined elements selected from the
Galois
Field. In an illustrative embodiment, the values for 131, b2,..., bolo are 2.
In addition, as
shown in equation (8), the packet MASCO is a redundant packet generated from
other
redundant packets only, specifically packets MAg and McD. It should be
apparent that the
redundant packet M
¨ABCD may be alternately generated using the elements of the
redundant packets MAC and Mgc. In some embodiments of the MPEG transmission
source generator 110, elimination of one or more non-systematic packets may be
performed in an operation known as puncturing. For instance, a punctured rate
4/8 may
be produced by not generating one of the packets that would have employed only
redundant packets (i.e., MABCD in this case) because this packet contains the
smallest
amount of intrinsic data. Any packet or codeword may be removed. However,
removal
of a packet or codeword containing the smallest amount of intrinsic data may
be optimal.
Code puncturing may be used to change the number of transmitted packets in
order to
meet certain limitations on number of packets or codewords transmitted.
Further, byte-code encoder 125 may also produce a rugged data stream that has
a data rate of 8/27 by employing 8 data packets MA, MB, === MH to produce 19
non-
systematic packets, as follows:
MAa(i) = MA(i)*bi + Ms(1)*b2 i=1,2, ...,187 (9)
McD(i) = Mc(i) *b3 MD(i)114 i=1,2, ...,187 (10)
MAC(i) = MA(i)*b5 + Mc(i)*be i=1,2, ...,187 (11)
MsD(i) = MBOrb7 + Mo(i)*b5 1=1,2, ...,187 (12)
RA It
= MAB(irb9 rocDorbio i=1,2,
MEF(i) = ME(1)*1311 + MF(i)*b12 1=1,2, ...,187 (14)
MGH(l) = MG(i) *b13 MH(i)*b14i=1,2, ...,187 (15)
MEG(i) = MEOrbis MG(i)11161=1,2, ...,187 (16)
MFH(I) = MFOrb17 MHOrbit3 i=1,2, ...,187 (17)
MEFGH(l) = MEFOrb19 MGHOrb20 i=1 ,2,
MAE(i) = MA(i)*b21 ME(i)*b22 1=1,2, ...,187 (19)
MsF(i) = Ms(i)*b23 + MFOrb24 i=1,2, ...,187 (20)
McG(i) = Mc(i)*b25 MG(i)*b26 i=1,2, ...,187 (21)
CA 02704485 2010-04-30
WO 2009/064468 PCT/US2008/012795
MoH(I) = Mo(I)*b27 MH(i)*b28 1=1 , 2, ... , 1 87 (22)
MACEGO) MAC(irb29 MEGO)1030 1=1,2,
MBDFH(1) = MeD(1)*b31 MFH(1)*b32 1=1,2, , 1 87 (24)
MABEF(1) = MAB(1)*b33 MEF(1)*b34 1=1,2, , 1 87 (25)
5 McDGH(1) = McD(1)*b35 MGH(1)*b36 1=1 , 2, ... , 187
(26)
MABCDEFGH(i) = MABCD(i)*b37 MEFGHOrb38 i=1 ,2, , 1 87 (27)
Additionally, a punctured code with data rate of 8/26 may be generated by the
byte-code encoder 125 by not generating the smallest intrinsic data value
packet
10 MABCDEFGH, or another packet generated from only redundant packets.
As described above, a byte-code encoder may be configured to produce certain
encoding code rates based on the number of codewords or packets used and the
number of codewords or packets formed through a single encoding process. In
addition,
more oomplicated code rates may be constructed using particular arrangements
of the
previously described code rate encoders as building blocks or constituent code
rate
encoders. Further, additional processing blocks may be included to form
a
concatenated byte-code encoder. For example, a concatenated byte-code encoder
may
use additional interleaving blocks between constituent byte-code encoders in
addition to
redundancy to improve the ruggedness of the data stream produced. Various
embodiments of redundant and code enhanced staggercasting transmission methods
will be described below.
After encoding, the data is coupled to a packet deinterleaver 130. The packet
deinterleaver 130 takes the bytes from the resulting SCBC codewords for the
original
group of packets in a column-by-column order, and outputs the bytes in a row
by row
order. The original packets are reconstituted and new packets are created from
the
parity bytes of the SCBC codewords. Each packet corresponds to a common
GF(256)
symbol location in all the created SCBC codewords. The number of packets
created in
each code frame is nSCBC, where the first kSCBC packets are the original data
packets
and the last (nSCBC - kSCBC) packets are parity packets.
CA 02704485 2010-04-30
WO 2009/064468
PCT/US2008/012795
11
The data is then coupled to the MPEG TS header modifier 135 where the MPEG
headers are modified. The MPEG TS header modifier may modify the packet
Identifier
(PID) of the MPEG transport stream headers to indicate the code rate used by
the error
correction scheme. The code rates is expressed as a fraction of the original
number of
data bytes over the total number of data bytes used. For example, in a 12/52
rate mode,
which supplements 12 data bytes with 40 parity bytes, each group of 12 bytes
uses one
R=1/2 Encoder, and two R=12/26 Encoders, with each 12/26 Encoder using two
R=2/3
Encoders and one 27/26 puncture, results in a 12/52 rate mode The R=27/26
puncture
is performed in such away that the last byte of the 27 bytes is dropped. Two
Data
Blocks are used to transmit 12 MPEG TS packets under the 12/52 Rate Mode. The
12/26 rate mode supplements 12 data bytes with 14 parity bytes, each group of
12 data
bytes uses two R=2/3 Encoders, and one R=27/26 puncture, results in a 12/26
rate
mode. The R=27/26 puncture shall be performed in such a way that the last byte
of the
27 bytes is dropped. One Data Block is used to transmit 12 MPEG TS packets
under
the 12/26 Rate Mode.The17/26 rate mode supplements 17 data bytes with 9 parity
bytes, each group of 17 data bytes group uses one R=2/3 Encoder to supplement
16
data bytes with 8 parity bytes, and one R=1/2 Encoder to supplement 1 data
byte with 1
parity byte, results in a 17/26 rate mode. One Data Block is used to transmit
17 MPEG
TS packets under the 17/26 Rate Mode. The 24/208 rate mode supplements 24 data
bytes with 184 parity bytes, each group of 24 data bytes uses 24 R=1/4
Encoders, and
eight 12/26 Encodes, results in a 24/208 rate mode. The R=27/26 puncture shall
be
performed in such a way that the last byte of the 27 bytes is dropped. Eight
Data Blocks
are used to transmit 24 MPEG TS packets under the 24/208 Rate Mode.
Each packet utilizing the MPEG protocol typically contains a packet
identification portion
or PID. The current system allows for over 8600 possible unique identification
elements,
and at present, only 50 are used. The PID is typically one or more bytes of
information
used for identifying the type of data in the packet. At present many of the
PID portions of
the bits remain reserved and unused. These PIDs can be used to identify a
specific
error correction code rate that will be imposed on the packet. Certain rules
based on
MPEG protocol should be maintained in order to assure the PID is properly
identified by
any receiving system. A three-byte header 440 contains a 13-bit packet
identifier (PID)
CA 02704485 2010-04-30
WO 2009/064468
PCT/US2008/012795
12
identifying the packet as part of a mobile/handheld transmission. The headers
440 of
MPEG packets from the ATSC-M/H stream are modified after packet-deinterleaving
to
contain packet identifiers (PIDs) that are not recognized by legacy ATSC A/53
receivers.
Thus, a legacy receiver should ignore the ATSC-M/H specific data, providing
backward
compatibility.
This data is then coupled to the Preamble packet inserter 140, where preamble
packets consisting of consecutive MPEG packets are formed into a preamble
block. The
MPEG packets are formed with a valid MPEG header with data bytes generated
from a
PN generator (not shown). The number of data bytes generated from the PN
generator
varies with the code rate used, for example, 184 data bytes are generated in
12/52 rate
mode to result in a total of 2208 bytes of PN data. According to an exemplary
embodiment, the PN generator is a 16-bit shift register with 9 feedback taps.
8 of the
shift register outputs are selected as the output byte. ATSC M/H packets are
placed in
between Preamble blocks in Data Blocks. Every Data Block contains 26 ATSC M/H
encoded packets that have the same coding or 26 ATSC A/53 encoded packets.
Once
the preamble packets have been inserted 140, the ATSC M/H stream has been
formed.
The ATSC-M/H data stream is then processed by the legacy ATSC A/53 path
145, including data randomizer 150, Reed-Solomon encoder 155, byte interleaver
160,
12-1 trellis encoder 165, sync insertion 170, pilot insertion 175, and
modulation 180. In
the data randomizer 150, each byte value is changed according to known pattern
of
pseudo-random number generation. This process is reversed in the receiver in
order to
recover the proper data values. With the exception of the segment and field
syncs, it is
desirable for the 8-VSB bit stream to have a completely random, noise-like
nature to
afford the transmitted signal frequency response must have a flat noise-like
spectrum in
order to use the allotted channel space with maximum efficiency.
The data is then coupled to the Reed-Solomon encoder 155, where Reed-
Solomon (RS) coding provides additional error correction potential at the
receiver
through the addition of additional data to the transmitted stream. In an
exemplary
embodiment, the RS code used in the VSB transmission system is a t =10
(207,187)
CA 02704485 2010-04-30
WO 2009/064468
PCT/US2008/012795
13
code. The RS data block size is 187 bytes, with 20 RS parity bytes added for
error
correction. A total RS block size of 207 bytes is transmitted per RS code
word. In
creating bytes from the serial bit stream, the MSB shall be the first serial
bit and the 20
RS parity bytes are sent at the end of the data block or RS code word.
The byte interleaver 160 then processes the output of the Reed-Solomon
encoder 155. Interleaving is a common technique for dealing with burst errors
that can
occur during transmission. Without interleaving, a burst error could have a
large impact
on one particular segment of the data, thereby rendering that segment
uncorrectable. If
the data is interleaved prior to transmission, however, the effect of a burst
error can be
effectively spread across multiple data segments. Rather than large errors
being
introduced in one localized segment that cannot be corrected, smaller errors
may be
introduced in multiple segments that are each separately within the correction
capabilities of forward error correction, parity bit, or other data integrity
schemes. For
instance, a common (255, 223) Reed-Solomon code will allow correction of up to
16
symbol errors in each code word. If the Reed-Solomon coded data is interleaved
before
transmission, a long error burst is more likely to be spread across multiple
codewords
after deinterleaving, reducing the chances that more than the correctable 16
symbol
errors are present in any particular codeword.
The interleaver employed in a VSB transmission system is a 52 data segment
(intersegment) convolutional byte interleaver. Interleaving is provided to a
depth of
about 1/6 of a data field (4 ms deep). Only data bytes are interleaved. The
interleaver
is synchronized to the first data byte of the data field. lntrasegment
interleaving is also
performed for the benefit of the trellis coding process.
The signal is then coupled to the Trellis encoder 165. Trellis coding is
another
form of Forward Error Correction. Unlike Reed-Solomon coding, which treats the
entire
MPEG-2 packet simultaneously as a block, trellis coding is an evolving code
that tracks
the progressing stream of bits as it develops through time. Accordingly, Reed-
Solomon
coding is known as a form of block code, while trellis coding is a
convolutional code.
CA 02704485 2010-04-30
WO 2009/064468
PCT/US2008/012795
14
In ATSC trellis coding, each 8-bit byte is split up into a stream of four, 2-
bit words.
In the trellis coder, each 2-bit word that arrives is compared to the past
history of
previous 2-bit words. A 3-bit binary code is mathematically generated to
describe the
transition from the previous 2-bit word to the current one. These 3-bit codes
are
substituted for the original 2-bit words and transmitted over-the-air as the
eight level
symbols of 8-VSB (3 bits = 8 combinations or levels). For every two bits that
go into the
trellis coder, three bits come out. For this reason, the trellis coder in the
8-VSB system is
said to be a 2/3 rate coder. The signaling waveform used with the trellis code
is an 8-
level (3 bit) one-dimensional constellation. The transmitted signal is
referred to as 8
VSB. A 4-state trellis encoder shall be used.
In an exemplary embodiment, trellis code intrasegment interleaving is used.
This
uses twelve identical trellis encoders and precoders operating on interleaved
data
symbols. The code interleaving is accomplished by encoding symbols (0, 12, 24
36 ...)
as one group, symbols (1, 13, 25, 37, ...) as a second group, symbols (2, 14,
26, 38, ...)
as a third group, and so on for a total of 12 groups.
Once the data has been trellis encoded, it is coupled to the sync inserter
170.
The sync inserter 170 is a multiplexer which inserts the various
synchronization signals
(Data Segment Sync and Data Field Sync). A two-level (binary) 4-symbol Data
Segment
Sync is inserted into the 8-level digital data stream at the beginning of each
Data
Segment. The MPEG sync byte is replaced by Data Segment Sync. In an exemplary
embodiment using ATSC transmission standards, a complete segment shall consist
of
832 symbols: 4 symbols for Data Segment Sync, and 828 data plus parity
symbols. The
same sync pattern occurs regularly at 77.3 s intervals, and is the only signal
repeating at
this rate. Unlike the data, the four symbols for Data Segment Sync are not
Reed-
Solomon or trellis encoded, nor are they interleaved. The ATSC segment sync is
a
repetitive four symbol (one byte) pulse that is added to the front of the data
segment and
replaces the missing first byte (packet sync byte) of the original MPEG-2 data
packet.
Correlation circuits in the 8-VSB receiver home in on the repetitive nature of
the segment
sync, which is easily contrasted against the background of completely random
data. The
recovered sync signal is used to generate the receiver clock and recover the
data.
CA 02704485 2010-04-30
WO 2009/064468
PCT/US2008/012795
Segment syncs are easily recoverable by the receiver because of their
repetitive nature
and extended duration. Accurate clock recovery can be had at noise and
interference
levels well above those where accurate data recovery is impossible allowing
for quick
data recovery during channel changes and other transient conditions.
5
After sync insertion, the signal is coupled to the pilot insertion where a
small DC shift is
applied to the 8-VSB baseband signal causing a small residual carrier to
appear at the
zero frequency point of the resulting modulated spectrum. This ATSC pilot
signal gives
the RF PLL circuits in the 8-VSB receiver a signal to lock onto that is
independent of the
10 data being transmitted. The frequency of the pilot is the same as the
suppressed-carrier
frequency. This may be generated by a small (digital) DC level (1.25) added to
every
symbol (data and sync) of the digital baseband data plus sync signal (+I, +3,
+5,+7). The
power of the pilot is typically 11.3 dB below the average data signal power.
15 After the pilot signal is inserted, the data is coupled to the modulator
180. The
modulator amplitude modulates the 8 VSB baseband signal on an intermediate
frequency (IF) carrier. With traditional amplitude modulation, we generate a
double
sideband RF spectrum about our carrier frequency, with each RF sideband being
the
mirror image of the other. This represents redundant information and one
sideband can
be discarded without any net information loss. In 8 VSB modulation, the VSB
modulator
receives the 10.76 Msymbols/s, 8-level trellis encoded composite data signal
(pilot and
sync added). The ATV system performance is based on a linear phase raised
cosine
Nyquist filter response in the concatenated transmitter and receiver, as shown
in Figure
12. The system filter response is essentially flat across the entire band,
except for the
transition regions at each end of the band. Nominally, the roll-off in the
transmitter shall
have the response of a linear phase root raised cosine filter.
The transmission system includes operation for mobile and portable devices in
a
burst mode of transmission. Several key advantages of operating in burst mode,
are
described throughout the above document and include ability to be received by
a new
class of devices while still maintaining backward compatibility. These new
classes of
devices require .a lower level of video resolution than is found in the
existing broadcast
CA 02704485 2010-04-30
WO 2009/064468
PCT/US2008/012795
16
standard, and can therefore also allow higher coding and compression, as well
as other
features including working in the presence of higher noise levels. An
additional
advantage of burst mode types of operation is focused on the potential device
power
savings by focusing use of the device only when signals intended for the
device or to be
received.
Burst mode operations such as those described may take advantage of time
periods during which high data transmission of a signal is not required in
order to
maintain full performance of a legacy system and receiver. Burst mode
operation may
be based on processing signals based on a so-called new information processing
rate,
which may change depending on the current broadcast signal characteristics.
Backward compatibility with the legacy system is maintained by focusing the
burst mode operations at a data packed level by introducing information for
new program
identifiers. The new program identifiers allow the new class of equipment to
recognize
the data, without affecting the operation of existing equipment. Further
legacy support
exists by including an Overlay structure in order to maintain legacy signal
transmission
operation during certain burst mode profiles.
Taking advantage of the full redundancy of the systematic data and non-
systematic data and the ability of the signal to be recovered from either of
the totally
separable codewords, adding frequency or special diversity as well as time
diversity can
increase the probability of reception by the mobile device. This will be
discussed further
in the discussions of Fig. 7.
Referring to FIG. 2, a block diagram of an embodiment of a portion of an
exemplary mobile/handheld data stream 200 of the present disclosure is shown.
26
ATSC M/H coded packets are grouped into 1 Data Block. In legacy ATSC
transmission
every Data Block typically has the same coding, although this is not
physically required.
Preamble blocks are two blocks long and have 52 ATSC M/H coded packets. The
very =
first MPEG packet following the Preamble block is a control packet that
contains system
information. Following randomization and forward error correction processing,
the data
CA 02704485 2010-04-30
WO 2009/064468
PCT/US2008/012795
17
packets are formatted into Data Frames for transmission and Data Segment Sync
and
Data Field Sync are added.
The ATSC-M/H data stream 200 is made up of bursts having a Preamble block
210 followed by a predetermined number of Data Blocks 230 appropriate for the
selected data rate mode. According to the exemplary embodiment, each Data
Block 230
consists of 26 MPEG packets. Each Data Frame consists of two Data Fields, each
containing 313 Data Segments. The first Data Segment of each Data Field is a
unique
synchronizing signal (Data Field Sync) and includes the training sequence used
by the
equalizer in the receiver. The remaining 312 Data Segments each carry the
equivalent
of the data from one 188-byte transport packet plus its associated FEC
overhead. The
actual data in each Data Segment comes from several transport packets because
of
data interleaving. Each Data Segment consists of 832 symbols. The first 4
symbols are
transmitted in binary form and provide segment synchronization. This Data
Segment
Sync signal also represents the sync byte of the 188-byte MPEG-compatible
transport
packet. The remaining 828 symbols of each Data Segment carry data equivalent
to the
remaining 187 bytes of a transport packet and its associated FEC overhead.
These 828
symbols are transmitted as 8-level signals and therefore carry three bits per
symbol.
Thus, 828 x 3 = 2484 bits of data are carried in each Data Segment, which is
the
requirement to send a protected transport packet:
The ATSC M/H data stream consists of a sequence of blocks, each block
consisting of 26 packets of either the legacy VSB N53 system or ATSC M/H
system.
The ATSC M/H data stream is made up of bursts of blocks that each burst has a
Preamble block followed by Nb Data Blocks, where Nb is a system variable
parameter
and a function of the overall ATSC M/H data rate to be transmitted. Each Data
Block is
encoded at one of the defined ATSC M/H rate modes. This rate mode is applied
to the
entire Data Block. For each burst of blocks, the Data Blocks are delivered
such that the
highest coded FEC rates (i.e. the lowest fractional numbers) in the burst of
blocks will be
delivered earliest and the lowest coded FEC rates (i.e. the highest fractional
numbers)
will be delivered the latest such that starting from a Preamble block, any
following Data
Blocks will have equal or less robustness than the current Data Block. ATSC
A/53 8VSB
CA 02704485 2010-04-30
WO 2009/064468
PCT/US2008/012795
18
coded legacy Data Blocks of 26 packets can be placed at one or more block for
legacy
overlay operation.
Turning now to FIG. 3, a data frame 300 is shown according to the present
invention is
shown. The data frame 300 shown is organized for transmission where each Data
Frame consists of two Data Fields, each containing 313 Data Segments. The
first Data
Segment of each Data Field is a unique synchronizing signal (Data Field Sync)
and
includes the training sequence used by the equalizer in the receiver. The
remaining 312
Data Segments each carry the equivalent of the data from one 188-byte
transport packet
plus its associated FEC overhead. The actual data in each Data Segment comes
from
several transport packets because of data interleaving. Each Data Segment
consists of
832 symbols. The first 4 symbols are transmitted in binary form and provide
segment
synchronization. This Data Segment Sync signal also represents the sync byte
of the 188-
byte MPEG-compatible transport packet. The remaining 828 symbols of each Data
Segment carry data equivalent to the remaining 187 bytes of a transport packet
and its
associated FEC overhead. These 828 symbols are transmitted as 8-level signals
and
therefore carry three bits per symbol. Thus, 828 x 3 = 2484 bits of data are
carried in each
Data Segment, which exactly matches the requirement to send a protected
transport
packet:
187 data bytes + 20 RS parity bytes = 207 bytes
207 bytes x 8 bits/byte = 1656 bits
2/3 rate trellis coding requires 3/2 x 1656 bits = 2484 bits.
The exact symbol rate is given by equation 1 below:
(1) Sr (MHz) = 4.5/286 x 684 = 10.76... MHz
The frequency of a Data Segment is given in equation 2 below:
(2) fseg= Sr /832 = 12.94... X 103 Data Segments/s.
The Data Frame rate is given by equation (3) below:
(3) fframe = fseg/626 = 20.66 ... frames/s.
The symbol rate Sr and the transport rate Tr shall be locked to each other in
frequency.
CA 02704485 2010-04-30
WO 2009/064468
PCT/US2008/012795
19
The 8-level symbols combined with the binary Data Segment Sync and Data Field
Sync signals is used to suppressed-carrier modulate a single carrier. Before
transmission,
however, most of the lower sideband shall be removed. The resulting spectrum
is flat,
except for the band edges where a nominal square root raised cosine response
results in
620 kHz transition regions. At the suppressed-carrier frequency, 310 kHz from
the lower
band edge, a small pilot is added to the signal as described previously.
Turning now to FIG. 4, an embodiment of a terrestrial broadcast receiver 400
for
mobile/handheld reception of the present disclosure is shown. The receiver 400
comprises a signal receiving element 410, a first tuner 420, a second tuner
425, a first
pre-equalizer demodulator 430, a second pre-equalizer demodulator 430, a
equalizer
controller 440, an equalizer 450, a post-equalizer correction processor 460, a
transport
decoder 470 and a tuner controller 480.
The signal receiving element 410 is operative to receive signals including
audio,
video, and/or data signals (e.g., television signals, etc.) from one or more
signal sources,
such as a terrestrial broadcast system and/or other type of signal broadcast
system.
According to an exemplary embodiment, signal receiving element 410 is embodied
as an
antenna such as a log periodic antenna, but may also be embodied as any type
of signal
receiving element. The antenna 410, of this exemplary embodiment, is operative
to
receive ATSC M/H terrestrially transmitted audio, video, and data signals over
a
frequency bandwidth. ATSC signals are generally transmitted over the frequency
range
of 54 to 870MHz, with a bandwidth of approximately 6 MHz per channel. Sub
channels
may be time multiplexed. The signal is coupled from the antenna vie a
transmission line
such as a coaxial cable or printed circuit board trace.
The first and second tuners 420, 425 are operative to perform a signal tuning
function responsive to a control signal from the tuner controller 480.
According to an
exemplary embodiment, the each tuner 420, 425 receives a different time, or
frequency
diverse RF signal from either one or a plurality of antennas 410, and performs
the signal
CA 02704485 2010-04-30
WO 2009/064468
PCT/US2008/012795
tuning function by filtering and frequency down converting (i.e., single or
multiple stage
down conversion) the RE signal to thereby generate an intermediate frequency
(IF)
signal. The RE and IF signals may include audio, video and/or data content
(e.g.,
television signals, etc.), and may be of an analog signal standard (e.g.,
NTSC, PAL,
5 SECAM, etc.) and/or a digital signal standard (e.g., ATSC, QAM, QPSK,
etc.). Each
tuner 420 is operative to convert the received ATSC M/H signal from the
carrier
frequency to an intermediate frequency. For example, the tuner may convert a
57 MHz
signal received at the antenna 410 to a 43 MHz IF signal. The Pre-Equalizer
Demodulator 430 is operative to demodulate the IF signal from the Tuner 420,
to a
10 baseband digital stream. Demodulator 435 is operative to demodulate the
IF signal from
tuner 425. The baseband digital streams are then coupled to the equalizer.
The tuner controller 480 is operative receive instructions from the transport
decoder 470 in response to the signal level and frequency of the tuned channel
or a
15 desired tuned channel. The tuner controller 480 generates a control
signal in response
to these received instructions to control the tuner 420, 425 operation.
The equalizer controller 440 is operative to generate an error term in
response to
the decoded data received from the demulators 430, 435. This provides the
ability for a
20 data directed equalizer. The equalizer controller 440 estimates the
error between the
received data and the decoded data and generates an error term. The error term
is fed
to the equalizer 450 to be minimized.
The equalizer 450 is operative to receive the tuned and demodulated MPEG
stream from the pre-equalizer demodulators 430, 435 and calculates equalizer
coefficients which are applied to an equalization filter within the equalizer
to produce an
error free signal. The equalizer 450 is operative to compensate for
transmission errors,
such as attenuation and intersymbol interference. The equalizer comprises a
matched
filter which performs roll off filtering which is operative to cancel the
intersymbol
interference. During the equalizer training period, a previously chosen
training signal is
transmitted through the channel and a properly delayed version of this signal,
that is
prestored in the receiver, is used as a reference signal. The training signal
is usually a
CA 02704485 2010-04-30
WO 2009/064468
PCT/US2008/012795
21
pseudo-noise sequence long enough to allow the equalizer to compensate for the
channel distortions. The equalizer according to an exemplary embodiment of the
present invention is operative to store a plurality of pseudo-noise sequences,
wherein
each pseudo-random sequence corresponds to a code rate. When the equalizer 450
receives the pseudorandom sequence training signal, the equalizer compares a
portion
of the received sequence with the plurality of stored sequences. When a match
is made,
the code rate associated with the received sequence is used by the decoder to
decode
the data received after the training sequence.
The post-equalizer correction processor 460 and transport decoder 470 are
operative to perform error correction and to decode the MPEG data stream.
These
elements are shown and discussed in detail in FIGs. 5 and 6.
The receiver may be configured to operate with a single tuner and demodulator
by time sharing the tuner to receive different frequencies and different
times.
Alternately, the tuner may be configured with a bandwidth wide enough to
receive two
signals simultaneously such that both signals may be tuned to different IF
frequencies
and each of these IF frequencies can be simultaneously processed or time
multiplexed
processed by a demodulator. In a single tuner, using either time or frequency
diversity
the packet combination is not done in the equalizer, but is instead done in
the code as
the equalizer must follow the transmitted signals. This gives three
possibilities of
receiving the packet correctly, the first packet correctly, the second packet
correctly, or
the combination after the byte decoder correctly. When coding is being used to
combine
the packets as opposed to receiving a single packet, it decreases the minimum
amount
of signal to noise ratio required to receive a virtually error free signal.
For example, in a
1/2 code rate, the minimum threshold is decreased from 15dB for a single
packet with no
coding to 7dB for 2 packets with coding and 3.5 dB for 4 packets with coding.
Turning now to FIG. 5, a block diagram of an embodiment of a decoder 500 used
in a receiver system is shown. Decoder 500 includes circuitry that is adapted
to use
redundant packets, such as the non-systematic packets in a data stream as
described
above, to aid in decoding data received by the receiver. Decoder 500 is also
generally
CA 02704485 2010-04-30
WO 2009/064468
PCT/US2008/012795
22
capable of decoding data that has been encoded using the legacy or existing
A53
standard.
In decoder 500, following initial tuning, demodulation, and processing by
other
circuits (FIG. 4) a trellis decoder 502 receives the incoming signal. The
trellis decoder
502 is connected to a convolutional de-interleaver 504. The output of the
convolutional
de-interleaver 504 is connected to a byte-code decoder 506. The byte-code
decoder
506 has an output that is connected to a Reed-Solomon decoder 508. The output
of the
Reed-Solomon decoder 508 is connected to a de-randomizer 510. The de-
randominizer
510 output is connected to a data decoder 512. The data decoder 512 provides
an
output signal for use in the remaining portion of the receiver system such as
video
display or audio reproduction.
In accordance with the existing or legacy A53 standard, the trellis decoder
502
includes a signal de-multiplexer, twelve 2/3-rate trellis decoders and a
signal multiplexer.
The de-multiplexer distributes the digital samples among the twelve 2/3-rate
trellis
decoders and the multiplexer multiplexes the estimates generated byte each of
the
twelve 2/3-rate trellis decoders. A de-interleaver 504, such as a
convolutional
interleaver, de-interleaves the stream of trellis-decoded bit estimates,
producing
sequences or packets arranged to include 207 bytes. The packet arrangement is
performed in conjunction with the determination and identification of the
location of the
synchronization signals, not shown. A Reed-Solomon error correction circuit
508
considers each sequence of 207 bytes produced by the de-interleaver 504 as one
or
more codewords and determines if any bytes in the codewords or packets were
corrupted due to an error during transmission. The determination is often
performed by
calculating and evaluating a set of syndromes or error patterns for the
codewords. If
corruption is detected, the Reed-Solomon error correction circuit 508 attempts
to recover
the corrupted bytes using the information encoded in the parity bytes. The
resulting
error-corrected data stream is then de-randomized by a de-randomizer 510 and
thereafter provided to a data decoder 512 that decodes the data stream in
accordance
with the type of content being transmitted. Typically, the combination of the
trellis
decoder 502, the de-interleaver 504, the Reed-Solomon decoder 508, and the de-
CA 02704485 2010-04-30
WO 2009/064468
PCT/US2008/012795
23
randomizer 510 are identified as an 8-VSB decoder within a receiver. It is
important to
note that, in general, the typical receiver for receiving signals compliant
with the legacy
A53 standard performs the receiving process in the reverse order of the
transmitting
process.
The received data, in the form of bytes of data in data packets, is decoded by
trellis decoder 502 and de-interleaved by de-interleaver 504. The data packets
may
include 207 bytes of data and further may be grouped in groups or 24, 26, or
52 packets.
The trellis decoder 502 and de-interleaver 504 are capable of processing
incoming
legacy format data as well as byte-code encoded data. Based on a predetermined
packet transmission sequence that is also known by the receiver, the byte-code
decoder
506 determines if the packet is a packet included in a byte-code encoded or
robust data
stream. If the received packet is not from the byte-code encoded data stream
then the
received packet is provided to the Reed-Solomon decoder 508 without any
further
processing in byte-code decoder 506. Byte code decoder 506 may also include a
de-
randomizer that removes the known sequence of constants multiplied by or added
to the
data stream during encoding. It is important to note that a rugged data stream
includes
both systematic packets and bytes that are identical to the original data and
non-
systematic packets and bytes that contain redundant data.
If the byte-code decoder 506 determines that the received is a byte-code
encoded packet belonging to robust or rugged data stream, the packet may be
decoded
along with other packets comprising the same data stream. In one embodiment,
byte-
code encoded packets of the same data stream are decoded by multiplying each
byte
within the packet by the inverse of the value of the element that was used to
develop the
byte-coded packet. The decoded values of the bytes of the non-systematic
packet are
compared to the values of the bytes of the systematic packet and the values of
any bytes
in the two packets that are not identical may be erased (i.e., set to zero) in
the
systematic packet or may be replaced by the information in the non-systematic
packet.
The systematic packet with error bytes erased may thereafter be decoded using
Reed-
Solomon decoding performed in Reed-Solomon decoder 508. Further description of
other embodiments of byte-code decoders will be discussed below.
CA 02704485 2010-04-30
WO 2009/064468
PCT/US2008/012795
24
Byte code decoder 506 may also be adapted to operate as a block coder for
decoding signals encoded as shown in FIG. 1. For instance, byte code decoder
506
may include a packet interleaver similar to packet interleaver 120 and a
packet
deinterleaver similar to packet deinterleaver 130. Additionally, the byte code
encoder
function may be adapted to decode a GF(256) Serial Concatenated Block Coded
(SCBC) signal. The byte code decoder 506 may further include an identifier
block used
for identifying data encoded for mobile or ATSC M/H reception and/or
identification of a-
priori training packets. Additionally, the identifier block may include a
packet identifier
block to determine, for example, if the headers in the incoming packets
include a PID
used for mobile reception.
It is important to note that in a preferred encoder byte-code encoding
precedes
the Reed-Solomon encoding of data packets. However, in decoder 500 shown here,
the
incoming data is byte-code decoded before being the Reed-Solomon decoded. The
re-
ordering is possible because both the byte-code operation and Reed-Solomon
code
operation are linear over the Galois Field(256) used in the A53 standard, and
linear
operators are commutative in a Galois Field. . It is advantageous to do block
decoding
first before the Reed Solomon because there are soft decoding algorithms which
make it
practical to have an iterative decoding algorithm. The importance of the re-
ordering is
important because the byte-code encoding provides a soft decoding algorithm,
which
then makes possible iterative decoding or turbo decoding, which has higher
reliability for
recovering errors in the received signal. As a result, performing byte-code
decoding
prior to Reed-Solomon decoding results in improved receiver performance as
measured
in terms of bit-error rate and signal to noise ratio
Turning now to FIG. 6, a block diagram of another embodiment of a decoder 600
used in a receiver is shown. Decoder 600 includes additional circuitry and
processing
for receiving and decoding signals that have been adversely affected by
transmission of
the signal over a transmission medium such as electromagnetic waves over the
air.
Decoder 600 is capable of decoding both a rugged data stream as well as a
legacy data
stream.
CA 02704485 2010-04-30
WO 2009/064468
PCT/US2008/012795
In decoder 600, the incoming signal, following initial processing, is provided
to
equalizer 606. Equalizer 606 is connected to trellis decoder 610, which
provides two
outputs. A first output from trellis decoder 610 provides feedback and is
connected back
5 as a feedback input to equalizer 606. The second output from trellis
decoder 610 is
connected to a convolutional de-interleaver 614. The convolutional de-
interleaver 614 is
connected to a byte-code decoder 616, which also provides two outputs. A first
output
from byte-code decoder 616 is connected back as a feedback input to trellis
decoder
610 through a convolutional interleaver 618. The second output from byte-code
decoder
10 616 is connected to a Reed-Solomon decoder 620. The output of the Reed-
Solomon
decoder 620 is connected to de-randomizer 624. The output of the de-randomizer
624
is connected to a data decoder 626. Reed-Solomon decoder 620, de-randomizer
624,
and data decoder 626 are connected, and functionally operate, in a manner
similar to
Reed-Solomon, de-randomizer, and data decoder blocks described in FIG. 5 and
will not
15 be further described here.
An input signal from the front end processing (e.g. antenna, tuner,
demodulator,
ND converter) of the receiver (not shown) is provided to equalizer 606.
Equalizer 606
processes the received signal to completely or partially remove the
transmission channel
20 effect in an attempt to recover the received signal. The various removal
or equalization
methods are well known to those skilled in the art and will not be discussed
here.
Equalizer 506 may include multiple sections of processing circuitry including
a feed-
forward equalizer (FFE) section and a decision-feedback-equalizer (DFE)
section.
25 The equalized signal is provided to trellis decoder 610. The trellis
decoder 610
produces, as one output, a set of decision values that are provided to the DFE
section of
equalizer 606. The trellis decoder 610 may also generate intermediate decision
values
that are also provided to the DFE section of equalizer 606. The DFE section
uses the
decision values along with intermediate decision values from the trellis
decoder 610 to
adjust values of filter taps in equalizer 606. The adjusted filter tap values
cancel
interference and signal reflections that are present in the received signal.
The iterative
process allows equalizer 606, with the assistance of feedback from trellis
decoder 610,
CA 02704485 2010-04-30
WO 2009/064468
PCT/US2008/012795
26
to dynamically adjust to a potential changing signal transmission environment
conditions
over time. It is important to note that the iterative process may occur at a
rate similar to
incoming data rate of the signal, such as 19 Mb/s for a digital television
broadcast signal.
The iterative process also may occur at a rate higher than the incoming data
rate.
The trellis decoder 610 also provides a trellis decoded data stream to
convolutional de-interleaver 614. Convolutional de-interleaver 614 operates
similar to
the de-interleaver described in FIG. 5 generates de-interleaved bytes
organized within
data packets. The data packets are provided to byte-code decoder 5616. As
described
above, packets that are not a part of a rugged data stream are simply passed
through
the byte-code decoder 616 to Reed-Solomon decoder 620. If the byte-code
decoder
616 identifies a group of the packets as part of a rugged data stream, the
byte-code
decoder 616 uses the redundant information in the non-systematic packets to
initially
decode the bytes in the packets as described above.
Byte-code decoder 616 and the trellis decoder 610 operate in an iterative
manner, referred to as a turbo-decoder, to decode the rugged data stream.
Specifically,
the trellis decoder 610 provides, after de-interleaving by convolutional de-
interleaver
614, a first soft decision vector to the byte-code decoder 616 for each byte
of the
packets that are included in the rugged data stream. Typically, the trellis
decoder 610
produces the soft decision as a vector of probability values. In some
embodiments, each
probability value in the vector is associated with a value that the byte
associated with the
vector may have. In other embodiments, the vector of probability values is
generated for
every half-nibble (i.e., two bits) that is contained in the systematic packet
because the
2/3-rate trellis decoder estimates two-bit symbols. In some embodiments the
trellis
decoder 610 combines four soft decisions associated with four half-nibbles of
a byte to
produce one soft-decision that is a vector of the probabilities of values that
the byte may
have. In such embodiments, the soft-decisions corresponding to the byte is
provided to
the byte-code decoder 616. In other embodiments, the byte-code decoder
separates a
soft-decision regarding a byte of the systematic packet into four soft-
decision vectors,
wherein each of the four soft-decisions is associated with a half-nibble of
the byte.
CA 02704485 2010-04-30
WO 2009/064468
PCT/US2008/012795
27
The byte-code decoder 616 uses the soft decision vector associated with the
bytes comprising packets of the rugged data stream to produce a first estimate
of the
bytes that comprise the packets. The byte-code decoder 616 uses both the
systematic
and the non-systematic packets to generate a second soft decision vector for
each byte
of packets comprising the rugged stream and provides the second soft-decision
vector to
the trellis decoder 610, after re-interleaving by convolutional interleaver
618. The trellis
decoder 610 thereafter uses the second soft-decision vector to produce a
further
iteration of the first decision vector, which is provided to the byte-code
decoder 616.
The trellis decoder 610 and the byte-code decoder 616 iterate in this fashion
until the
soft-decision vector produced by the trellis decoder and byte-code decoder
converge or
a predetermined number of iterations are undertaken. Thereafter, the byte-code
decoder
616 uses the probability values in the soft-decision vector for each byte of
the systematic
packets to generate a hard decision for each byte of the systematic packets.
The hard
decision values (i.e., decoded bytes) are output from the byte-code encoder
616 to
Reed-Solomon decoder 620. The trellis decoder 610 may be implemented using a
Maximum a Posteriori (MAP) decoder and may operate on either byte or half-
nibble
(symbol) soft decisions.
It is important to note that turbo-decoding typically utilizes iteration rates
related
to passing decision data between blocks that are higher than the incoming data
rates.
The number of possible iterations is limited to the ratio of the data rate and
the iteration
rate. As a result and to the extent practical, a higher iteration rate in the
turbo-decoder
generally improves the error correction results. In one embodiment, an
iteration rate that
is 8 times the incoming data rate may be used.
A soft input soft output byte-code decoder such as described in FIG. 6 may
include vector decoding functions. Vector decoding involves grouping bytes of
the data
including systematic and non-systematic bytes. For example, for a rate 1/2
byte code
encoded stream, 1 systematic and 1 non-systematic byte will be grouped. The
two
bytes have over 64,000 possible values. The vector decoder determines or
estimates a
probability for each of the possible values of the two bytes and creates a
probability
map. A soft decision is made based on a weighting the probabilities of some or
all of the
CA 02704485 2010-04-30
WO 2009/064468
PCT/US2008/012795
28
possibilities and the Euclidean distance to a possible codeword. A hard
decision may be
made when the error of the Euclidean distance falls below a threshold.
Byte-code decoders, as described in FIG.s 5 and 6 may decode a rugged data
stream that has been encoded by the byte-code encoders described earlier,
including
encoding by simple byte-code encoders or concatenated byte-code encoders. The
byte-
code decoders in FIG.s 5 and 6 describe decoding a rugged data stream encoded
by a
simple or constituent byte-code encoder involving only a single encoding step.
Concatenated byte-code decoding includes decoding the incoming codewords or
bytes
in more than one decoding step in addition to intermediate processing such as
de-
interleaving, de-puncturing, and re-insertion.
Referring to Fig. 7, an exemplary broadcast environment 700 according to the
present invention is shown. A first transmitter 720, a second transmitter 720
and a
mobile receiver 730 are shown. The first transmitter 710 is located at a
distance dl from
the mobile receiver 730 and the second transmitter 720 is located at a
distance d2 from
the mobile receiver 730.
Taking advantage of the separable and redundant codewords, a first codeword
can be transmitted from the first transmitter 710 and a second codeword can be
transmitted from the second transmitter 720. This reduces the occurrence of
total signal
loss by varying the propagation paths and angles. These variations reduce the
probability of total signal nulls or completely destructive multipath. Signal
reception can
further be improved by transmitting each codeword at a different frequency
and/or at a
different time.
This embodiment of spatial and frequency diversity may utilize the inherent
"white space" between the existing broadcast channels in a coverage area,
while not
increasing the equalizer complexity of the current receiver. Such a proposed
embodiment is also particularly well suited, though not limited to, burst mode
transmission as is currently under consideration for advance broadcast
transmission
systems. In burst mode transmission, a single tuner in the receiver can still
receive a
CA 02704485 2010-04-30
WO 2009/064468
PCT/US2008/012795
29
complete transmission by receiving non-simultaneous content at multiple
frequencies. A
complete reception may be achieved by receiving only one of two or more bursts
supplied from more than one source, including the main and secondary
transmitters.
Signal synchronization may be maintained through a number of known techniques
including techniques already in use for SFN.
Referring to Fig. 8, an exemplary embodiment of a portion of a transmitter 800
according to the present invention is shown. The figure below illustrates an
arrangement
'
for a specific implementation of a code enhanced deeply interleaved
staggercasting
structure in the physical layer of a signal transmitting system.
Implementation in a
receiver would result in similar structures rearranged for decoding and
demodulation as
opposed to encoding and modulation. In operation, the processing involves
identification and generation of information for redundant or staggercast
operation. This
content is received via the channel output 805. Next, the content is provided
to a coder
containing two or more parallel encoding branches suitable for nominally
generating a
standard coded burst mode signal. Next, each coding branch processes its
supplied
portion of the signal. Next, one branch is delayed by a delay RAM 815 by a
predetermined amount. The delay amount may represent a number cycles of a
signal
such as field synchronization signal, and further may delay to an equivalent
time
representation of a future or subsequent burst transmission time. Each signal
on each
branched is encoded in their respective input stages 810, 820. The input
stages may
comprise legacy deinterlacing and packet deinterlacing. For time diversity
signals, the
undelayed coding branch signal may be combined 825 with a previously coded and
delayed branch signal containing a portion of previously processed data
content and the
combination provided to the remaining portions of the transmitter. The
combined data is
then block decoded 830 and separated 840 into different output stages. Each
output
stage 845, 850 may combine legacy interleaving and legacy encoding. Processes
for
the receiver may be substantially similar and primarily reversed from the
processes for a
transmitter.
For spatially diverse or frequency diverse transmission, content is coupled
from
the channel input, before or after the delay RAM 815 and optionally coupled to
another
CA 02704485 2010-04-30
WO 2009/064468
PCT/US2008/012795
delay RAM 855. The content may then be coupled to further channels for
transmission
comprising operations as shown with respect to Fig. 1. One advantage of
spatial and
frequency diverse transmitters is using cooperative transmission. In this
example, two
television broadcasters, or one broadcaster with two transmitters or
frequencies, can
5 transmit the first broadcast packet on one transmitter and the
cooperating broadcast
redundant packet. Thus, each broadcaster transmits 2 packets, one burst and
one
redundant burst, but gains a diversity advantage by having their redundant
packet
transmitted on another frequency and/or another transmitter simply by
transmitting the
redundant burst of the cooperating broadcaster. Cooperative transmission gives
the
10 benefit of complete frequency diversity and possibly spatially diversity
without increasing
each cooperating transmitters data output or bandwidth.
Further, in addition to the inherent benefit associated with significant time
interleaving of the data content and the relationship to operation in burst
mode
15 transmissions, each branch of modified code enhanced data may be
transmitted on
separate frequencies. In this manner, frequency diversity, in addition to time
diversity,
may be achieved. For example, a first burst containing a portion of the
staggercast
content, after code enhancing, may be provided or transmitted on a particular
broadcast
channel by a first broadcaster. A second burst containing a remaining portion
of the
20 staggercast content, after code enhancing, may be provided or
transmitted at a later
point in time and on a second broadcast channel, possibly by a second
broadcaster.
Each broadcast channel represents a different frequency spectrum of operation.
The
resultant operation further guarantees recovery of the original data content
by
introducing frequency diversity operation to the already inherent time
diversity system.
While the present invention has been described in terms of a specific
embodiment, it will be appreciated that modifications may be made which will
fall within
the scope of the invention. For example, various processing steps may be
implemented
separately or combined, and may be implemented in general purpose or dedicated
data
processing hardware. Furthermore, various encoding or compression methods may
be
employed for video, audio, image, text, or other types of data. Also, the
packet sizes,
CA 02704485 2010-04-30
WO 2009/064468
PCT/US2008/012795
31
rate modes, block coding, and other information processing parameters may be
varied in
different embodiments of the invention.
