5.1 Packet data traffic channel (PDTCH)
3GPP45.003GSM/EDGE Channel codingRelease 17TS
Thirteen coding schemes are specified for the packet data traffic channels. For the coding schemes CS-2 to CS-4 and MCS-1 to MCS-4, the first three bits (USF-bits) of the data block are encoded such that the first twelve coded bits are representing the same bit pattern, irrespective of the coding scheme, depending only on the USF-bits. For these coding schemes, the USF-bits can therefore always be decoded from these twelve bits in the same way. It should be noted that the USF precoding is done in the uplink direction for coding schemes CS-2 – CS-4, despite the fact that uplink RLC data block structure (3GPP TS 44.060) does not define USF-field.
For the nine coding schemes MCS-1 to MCS-9, the block structure differs between uplink and downlink since header sizes before coding are not the same.
In BTTI configuration, the RLC/MAC layer delivers to the encoder one data block every 20 ms. In RTTI configuration, the RLC/MAC layer delivers to the encoder one data block every 10 ms or, if BTTI USF mode is used (see 3GPP TS 45.002), the RLC/MAC layer in the downlink may deliver to the encoder two data blocks every 20 ms.
In the downlink direction, if BTTI USF mode is used (see 3GPP TS 45.002), one value of the USF per PDCH is delivered to the encoder every 20 ms; if RTTI USF mode is used (see 3GPP TS 45.002), one value of the USF per corresponding downlink PDCH-pair is delivered to the encoder every 10 ms.
If BTTI USF mode is used when sending downlink data blocks in RTTI configuration, then the USF need not be delivered to the encoder as the first three bits of a data block. In this case, the first three bits of a data block are set to an unspecified value (see 3GPP TS 44.060).
NOTE: How the USFs are delivered to the encoder in this case is implementation dependent.
If BTTI USF mode is used when sending downlink data blocks in RTTI configuration, then both data blocks sent in a 20 ms block period shall be encoded using coding schemes with the same modulation, unless
– Reduced Latency is supported by the MS to which the USF is intended or
– EGPRS2 is supported by the MS to which the USF is intended or
– The USF to be transmitted is set to an unused value (see 3GPP TS 45.002).
If an MS, that neither supports Reduced Latency nor EGPRS2, receives in a 20ms block period data blocks sent using different modulations, the MS shall ignore the USF.
If BTTI USF mode is used when sending downlink data blocks in RTTI configuration and different modulations are used in the two data blocks sent in a 20 ms block period, the USF will be sent with mixed modulation. In this case, the coding of the USF bits sent in the first 10 ms block period is according to the MCS used in that block period, while the coding of the USF bits sent in the second 10 ms block period is according to the MCS used in that block period. The network shall use only MCSs supported by the MS to which the USF is intended.
5.1.1 Packet data block type 1 (CS-1)
The coding scheme used for packet data block type 1 is the same as for SACCH as specified in section 4.1.
The flags hl(B) and hu(B) set to "1" identify the coding scheme CS-1.
In RTTI configuration with RTTI USF mode, the bursts with B = 0,2 shall be mapped on the PDCH having the lower timeslot number, whereas the bursts with B = 1,3 shall be mapped on the PDCH having the higher timeslot number, see 3GPP TS 45.002.
5.1.2 Packet data block type 2 (CS-2)
5.1.2.1 Block constitution
The message delivered to the encoder has a fixed size of 271 information bits {d(0),d(1),…,d(270)}. It is delivered on a burst mode.
5.1.2.2 Block code
a) USF precoding:
The first three bits d(0),d(1),d(2) are precoded into six bits u’(0),u’(1),…,u’(5) according to the following table:
|
d(0),d(1),d(2) |
u’(0),u’(1),…,u’(5) |
|
000 |
000 000 |
|
001 |
001 011 |
|
010 |
010 110 |
|
011 |
011 101 |
|
100 |
100 101 |
|
101 |
101 110 |
|
110 |
110 011 |
|
111 |
111 000 |
b) Parity bits:
Sixteen parity bits p(0),p(1),…,p(15) are defined in such a way that in GF(2) the binary polynomial:
d(0)D286 +…+ d(270)D16 + p(0)D15 +…+ p(15), when divided by:
D16 + D12 + D5 + 1, yields a remainder equal to:
D15 + D14 + D13 + D12 + D11 + D10 + D9 + D8 + D7 + D6 + D5 + D4 + D3 + D2 + D+1.
c) Tail bits:
Four tail bits equal to 0 are added to the information and parity bits, the result being a block of 294 bits {u(0),u(1),…,u(293)}:
u(k) = u’(k) for k = 0,1,…,5
u(k) = d(k-3) for k = 6,7,…,273
u(k) = p(k‑274) for k = 274,275,…,289
u(k) = 0 for k = 290,291,292,293 (tail bits)
5.1.2.3 Convolutional encoder
This block of 294 bits {u(0),u(1),…,u(293)} is encoded with the ½ rate convolutional code (identical to the one used for TCH/FS) defined by the polynomials:
G0 = 1 + D3 + D4
G1 = 1 + D + D3 + D4
This results in a block of 588 coded bits: {C(0),C(1),…,C(587)} defined by:
C(2k) = u(k) + u(k‑3) + u(k‑4)
C(2k+1) = u(k) + u(k‑1) + u(k‑3) + u(k‑4) for k = 0,1,…,293 ; u(k) = 0 for k < 0
The code is punctured in such a way that the following coded bits:
{C(3+4j) for j = 3,4,…,146 except for j = 9,21,33,45,57,69,81,93,105,117,129,141} are not transmitted
The result is a block of 456 coded bits, {c(0),c(1),…,c(455)}.
5.1.2.4 Interleaving
The interleaving is done as specified for SACCH in section 4.1.4.
5.1.2.5 Mapping on a burst
The mapping is given by the rule:
e(B,j) = i(B,j) and e(B,59+j) = i(B,57+j) for j = 0,1,…,56
and
e(B+m,57) = q(2m) and e(B+m,58) = q(2m+1) for m = 0,1,2,3
where
q(0),q(1),…,q(7) = 1,1,0,0,1,0,0,0 identifies the coding scheme CS-2.
5.1.3 Packet data block type 3 (CS-3)
5.1.3.1 Block constitution
The messages delivered to the encoder has a fixed size of 315 information bits {d(0),d(1),…,d(314)}. It is delivered on a burst mode.
5.1.3.2 Block code
a) USF precoding:
The first three bits d(0),d(1),d(2) are precoded into six bits u’(0),u’(1),…,u’(5) as specified for CS-2 in section 5.1.2.2.a).
b) Parity bits:
Sixteen parity bits p(0),p(1),…,p(15) are defined in such a way that in GF(2) the binary polynomial:
d(0)D330 +…+ d(314)D16 + p(0)D15 +…+ p(15), when divided by:
D16 + D12 + D5 + 1, yields a remainder equal to:
D15 + D14 + D13 + D12 + D11 + D10 + D9 + D8 + D7 + D6 + D5 + D4 + D3 + D2 + D+1.
c) Tail bits:
Four tail bits equal to 0 are added to the information and parity bits, the result being a block of 338 bits {u(0),u(1),…,u(337)}:
u(k) = u’(k) for k = 0,1,…,5
u(k) = d(k-3) for k = 6,7,…,317
u(k) = p(k‑318) for k = 318,319,…,333
u(k) = 0 for k = 334,335,336,337 (tail bits)
5.1.3.3 Convolutional encoder
This block of 338 bits {u(0),u(1),…,u(337)} is encoded with the ½ rate convolutional code (identical to the one used for TCH/FS) defined by the polynomials:
G0 = 1 + D3 + D4
G1 = 1 + D + D3 + D4
This results in a block of 676 coded bits: {C(0),C(1),…,C(675)} defined by:
C(2k) = u(k) + u(k‑3) + u(k‑4)
C(2k+1) = u(k) + u(k‑1) + u(k‑3) + u(k‑4) for k = 0,1,…,337 ; u(k) = 0 for k < 0
The code is punctured in such a way that the following coded bits:
{C(3+6j) and C(5+6j) for j = 2,3,…,111} are not transmitted
The result is a block of 456 coded bits, {c(0),c(1),…,c(455)}.
5.1.3.4 Interleaving
The interleaving is done as specified for SACCH in subclause 4.1.4.
5.1.3.5 Mapping on a burst
The mapping is given by the rule:
e(B,j) = i(B,j) and e(B,59+j) = i(B,57+j) for j = 0,1,…,56
and
e(B+m,57) = q(2m) and e(B+m,58) = q(2m+1) for m = 0,1,2,3
where
q(0),q(1),…,q(7) = 0,0,1,0,0,0,0,1 identifies the coding scheme CS-3.
5.1.4 Packet data block type 4 (CS-4)
5.1.4.1 Block constitution
The message delivered to the encoder has a fixed size of 431 information bits {d(0),d(1),…,d(430)}. It is delivered on a burst mode.
5.1.4.2 Block code
a) USF precoding:
The first three bits d(0),d(1),d(2) are block coded into twelve bits u’(0),u’(1),…,u’(11) according to the following table:
|
d(0),d(1),d(2) |
u’(0),u’(1),…,u’(11) |
|
000 |
000 000 000 000 |
|
001 |
000 011 011 101 |
|
010 |
001 101 110 110 |
|
011 |
001 110 101 011 |
|
100 |
110 100 001 011 |
|
101 |
110 111 010 110 |
|
110 |
111 001 111 101 |
|
111 |
111 010 100 000 |
b) Parity bits:
Sixteen parity bits p(0),p(1),…,p(15) are defined in such a way that in GF(2) the binary polynomial:
d(0)D446 +…+ d(430)D16 + p(0)D15 +…+ p(15), when divided by:
D16 + D12 + D5 + 1, yields a remainder equal to:
D15 + D14 + D13 + D12 + D11 + D10 + D9 + D8 + D7 + D6 + D5 + D4 + D3 + D2 + D+1.
The result is a block of 456 coded bits, {c(0),c(1),…,c(455)}:
c(k) = u’(k) for k = 0,1,…,11
c(k) = d(k-9) for k = 12,13,…,439
c(k) = p(k‑440) for k = 440,441,…,455
5.1.4.3 Convolutional encoder
No convolutional coding is done.
5.1.4.4 Interleaving
The interleaving is done as specified for SACCH in section 4.1.4.
5.1.4.5 Mapping on a burst
The mapping is given by the rule:
e(B,j) = i(B,j) and e(B,59+j) = i(B,57+j) for j = 0,1,…,56
and
e(B+m,57) = q(2m) and e(B+m,58) = q(2m+1) for m = 0,1,2,3
where
q(0),q(1),…,q(7) = 0,0,0,1,0,1,1,0 identifies the coding scheme CS-4.
5.1.4a Packet data block type 5a (MCS-0)
5.1.4a.1 Downlink (MCS-0 DL)
5.1.4a.1.1 Block constitution
The message delivered to the encoder has a fixed size of 207 information bits {d(0),d(1),…,d(206)}. It is delivered on a burst mode.
The message delivered to the encoder will have a fixed size of 210 information bits {d(0),d(1),…,d(209)}, if an eTFI field is included (see 3GPP TS 44.060).
5.1.4a.1.2 USF precoding
Twelve bits u’(0),u’(1),…,u’(11) are generated as described for MCS-1 DL in subclause 5.1.5.1.2.2.
5.1.4a.1.3 Data coding
a) Parity bits:
Eighteen data parity bits p(0),p(1),…,p(17) are defined in such a way that in GF(2) the binary polynomial:
d(31)D193 +…+ d(206)D18 + p(0)D17 +…+ p(17), when divided by:
D18 + D17 + D14 + D13 + D11 + D10 + D8 + D7 + D6 + D3 + D2 + 1, yields a remainder equal to:
D17 + D16 + D15 + D14 + D13 + D12 + D11 + D10 + D9 + D8 + D7 + D6 + D5 + D4 + D3 + D2 + D + 1.
In case an eTFI field (see 3GPP TS 44.060) is included the three bits {d(207),d(208),d(209} are added bit-wise modulo 2 to the last three parity bits {p(15), p(16),p(17)}. This results in eighteen modified parity bits {pt(0),…,pt(17)} defined as:
pt(k) = p(k) for k=0,…,14
pt(k) = d(k+192) + p(k) for k=15,…,17
b) Tail bits:
Six tail bits equal to 0 are added to the information bits, the result being a block of 182 bits
{u(0),u(1),…,u(181)}:
u(k) = d(k+31) for k = 0,1,…,175
u(k) = 0 for k = 176,177,…,181 (tail bits)
c) Convolutional encoder
This block of 182 bits {u(0),u(1),…,u(181)} is encoded with the 1/3 rate convolutional mother code defined by the polynomials:
G4 = 1 + D2 + D3 + D5 + D6
G7 = 1 + D + D2 + D3 + D6
G5 = 1 + D + D4 + D6
This results in a block of 546 coded bits: {C(0),C(1),…,C(545)} defined by:
C(3k) = u(k) + u(k‑2) + u(k‑3) + u(k‑5) + u(k‑6)
C(3k+1) = u(k) + u(k‑1) + u(k‑2) + u(k‑3) + u(k‑6)
C(3k+2) = u(k) + u(k‑1) + u(k-4) + u(k-6) for k = 0,1,…,181; u(k) = 0 for k < 0
The code is punctured in such a way that the following coded bits are not transmitted:
|
{C(2+3j) for j = 0,1,…,181} are not transmitted except {C(k) for k = 35, 104, 173, 242, 308, 377, 446, 515} which are transmitted |
The result is a block of 372 coded bits, {dc(0),dc(1),…,dc(371)}.
5.1.4a.1.4 Header coding
The header bits {d(3),d(4),…,d(30)} shall be coded as for Packet data block type 5 (MCS-1) in subclause 5.1.5.1.3.
When applying the coding in subclause 5.1.5.1.3, the operations specific to eTFI inclusion therein shall be ignored.
Before coding {d(8)..d(23)} is replaced by {p(0),..,p(15) and {d(29),d(30)} is replaced by {p(16),p(17)}, where {p (0),..,p(17)} is defined in 5.1.4a.1.3.
In case an eTFI field is included p(0),…,p(17) is replaced by pt(0),…pt(17) defined in 5.1.4a.1.3.
The result is a block of 68 coded bits, {hc(0),hc(1),…,hc(67)}.
5.1.4a.1.5 Interleaving
The interleaving is done as specified for MCS-1 DL in subclause 5.1.5.1.5.
5.1.4a.1.6 Mapping on a burst
The mapping is done as specified for MCS-1 DL in subclause 5.1.5.1.6.2.
5.1.5 Packet data block type 5 (MCS-1)
5.1.5.1 Downlink (MCS-1 DL)
5.1.5.1.1 Block constitution
The message delivered to the encoder has a fixed size of 209 information bits {d(0),d(1),…,d(208)}. It is delivered on a burst mode.
The message delivered to the encoder will have a fixed size of 212 information bits {d(0),d(1),…,d(211)}, if an eTFI field is included (see 3GPP TS 44.060).
The message delivered to the encoder will have a fixed size of 234 information bits {d(0),d(1),…,d(233)}, if a PAN field is included (see 3GPP TS 44.060).
The message delivered to the encoder will have a fixed size of 237 information bits {d(0),d(1),…,d(236)}, if a PAN field and an eTFI field are included (see 3GPP TS 44.060).
NOTE: The presence of the PAN is indicated by the PANI field in the header (see 3GPP TS 44.060).
5.1.5.1.2 USF precoding
5.1.5.1.2.1 BTTI configuration
The first three bits d(0),d(1),d(2) are block coded into twelve bits u’(0),u’(1),…,u’(11) as for Packet data block type 4 (CS-4) in subclause 5.1.4.2.
5.1.5.1.2.2 RTTI configuration
If the USF is sent in RTTI USF mode (see 3GPP TS 45.002) when data blocks are transmitted in RTTI configuration, then the first three bits d(0),d(1),d(2) are block coded into twelve bits u’(0),u’(1),…,u’(11) as for Packet data block type 4 (CS-4) in subclause 5.1.4.2.
If the USF is sent in BTTI USF mode (see 3GPP TS 45.002) when data blocks are transmitted in RTTI configuration, then the three bits of the USF to be sent on the lower numbered PDCH of a corresponding downlink PDCH-pair are block coded into twelve bits uL(0),uL(1),…,uL(11) as for Packet data block type 4 (CS-4) in subclause 5.1.4.2; the three bits of the USF to be sent on the higher numbered PDCH of a corresponding downlink PDCH-pair are block coded into twelve bits uH(0),uH(1),…,uH(11) as for Packet data block type 4 (CS-4) in subclause 5.1.4.2.
NOTE: If BTTI USF mode is used when sending data blocks in RTTI configuration, then d(0),d(1),d(2) need not contain a USF; in this case, they are ignored by the encoder. How the USFs are delivered to the encoder in this case is implementation dependent.
If the data block is sent in the first 10ms of a 20ms block period, then
u’(i) = uL(i) for i = 0, 4, 8;
u’(i) = uH(i-1) for i = 1, 5, 9;
u’(i) = uL(i‑1) for i = 2, 6, 10;
u’(i) = uH(i‑2) for i = 3, 7, 11.
If the data block is sent in the second 10ms of a 20ms block period, then
u’(i) = uL(i+2) for i = 0, 4, 8;
u’(i) = uH(i+1) for i = 1, 5, 9;
u’(i) = uL(i+1) for i = 2, 6, 10;
u’(i) = uH(i) for i = 3, 7, 11.
NOTE: In case mixed modulation USF is used (see subclause 5.1), the USF bits sent during the other half of the 20 ms block period may be sent with a different modulation. In this case, the half of uL and uH not sent in the present data block will be discarded.
5.1.5.1.3 Header coding
a) Parity bits:
Eight header parity bits p(0),p(1),…,p(7) are defined in such a way that in GF(2) the binary polynomial:
d(3)D35 +…+ d(30)D8 + p(0)D7 +…+ p(7), when divided by:
D8 + D6 + D3 + 1, yields a remainder equal to:
D7 + D6 + D5 + D4 + D3 + D2 + D + 1.
In case an eTFI field (see 3GPP TS 44.060) is included and a PAN field is not the three bits {d(209),d(210),d(211)} are added bit-wise modulo 2 to the last three parity bits {p(5), p(6),p(7)}. This results in the eight modified header parity bits {pt(0),…,pt(7)} defined as:
pt(k) = p(k) for k=0,…,4
pt(k) = d(k+204) + p(k) for k=5,…,7
In case an eTFI field (see 3GPP TS 44.060) and a PAN field are included the three bits {d(234),d(235),d(236)} are added bit-wise modulo 2 to the last three parity bits {p(5), p(6),p(7)}. This results in the eight modified header parity bits {pt(0),…,pt(7)} defined as:
pt(k) = p(k) for k=0,…,4
pt(k) = d(k+229) + p(k) for k=5,…,7
b) Tail biting:
The six last header parity bits are added before information and parity bits, the result being a block of 42 bits {u"(‑6),…,u"(0),u"(1),…,u"(35)} with six negative indexes. In case an eTFI field is not included this operation is defined as:
u"(k-6) = p(k+2) for k = 0,1,…,5
u"(k) = d(k+3) for k = 0,1,…,27
u"(k) = p(k‑28) for k = 28,29,…,35
In case an eTFI field is included p(k) is replaced by pt(k) in the above operation.
c) Convolutional encoder
This block of 42 bits {u"(-6),…,u"(0),u"(1),…,u"(35)} is encoded with the 1/3 rate convolutional mother code defined by the polynomials:
G4 = 1 + D2 + D3 + D5 + D6
G7 = 1 + D + D2 + D3 + D6
G5 = 1 + D + D4 + D6
This results in a block of 108 coded bits: {C(0),C(1),…,C(107)} defined by:
C(3k) = u"(k) + u"(k‑2) + u"(k‑3) + u"(k‑5) + u"(k‑6)
C(3k+1) = u"(k) + u"(k‑1) + u"(k‑2) + u"(k‑3) + u"(k‑6)
C(3k+2) = u"(k) + u"(k‑1) + u"(k-4) + u"(k-6) for k = 0,1,…,35
The code is punctured in such a way that the following coded bits:
{C(2+3j) for j = 0,1,…,35} as well as {C(k) for k = 34,58,82,106} are not transmitted
The result is a block of 68 coded bits, {hc(0),hc(1),…,hc(67)}.
5.1.5.1.4 Data coding
a) Parity bits:
Twelve data parity bits p(0),p(1),…,p(11) are defined in such a way that in GF(2) the binary polynomial:
d(31)D189 +…+ d(208)D12 + p(0)D11 +…+ p(11), when divided by:
D12 + D11 + D10 + D8 + D5 + D4 + 1, yields a remainder equal to:
D11 + D10 + D9 + D8 + D7 + D6 + D5 + D4 + D3 + D2 + D + 1.
b) Tail bits:
Six tail bits equal to 0 are added to the information and parity bits, the result being a block of 196 bits
{u(0),u(1),…,u(195)}:
u(k) = d(k+31) for k = 0,1,…,177
u(k) = p(k‑178) for k = 178,179,…,189
u(k) = 0 for k = 190,191,…,195 (tail bits)
c) Convolutional encoder
This block of 196 bits {u(0),u(1),…,u(195)} is encoded with the 1/3 rate convolutional mother code defined by the polynomials:
G4 = 1 + D2 + D3 + D5 + D6
G7 = 1 + D + D2 + D3 + D6
G5 = 1 + D + D4 + D6
This results in a block of 588 coded bits: {C(0),C(1),…,C(587)} defined by:
C(3k) = u(k) + u(k‑2) + u(k‑3) + u(k‑5) + u(k‑6)
C(3k+1) = u(k) + u(k‑1) + u(k‑2) + u(k‑3) + u(k‑6)
C(3k+2) = u(k) + u(k‑1) + u(k-4) + u(k-6) for k = 0,1,…,195; u(k) = 0 for k < 0
The code is punctured depending on the value of the CPS field as defined in 3GPP TS 44.060. Two puncturing schemes named P1 or P2 are applied in such a way that the following coded bits:
|
P1 |
{C(2+21j), C(5+21j), C(8+21j), C(10+21j), C(11+21j), C(14+21j), C(17+21j), C(20+21j) for j = 0,1,…,27} are not transmitted except {C(k) for k = 73,136,199,262,325,388,451,514} which are transmitted |
|
P2 |
{C(1+21j), C(4+21j), C(7+21j), C(9+21j), C(13+21j), C(15+21j), C(16+21j), C(19+21j) for j = 0,1,…,27} are not transmitted except {C(k) for k = 78,141,204,267,330,393,456,519} which are transmitted |
The result is a block of 372 coded bits, {dc(0),dc(1),…,dc(371)}.
For the FANR procedure, the code is punctured depending on the CPS field and the PANI field as defined in 3GPP TS 44.060. If the PANI field is set to 0, the puncturing is the same as for EGPRS. If the PANI field is set to 1, the puncturing schemes named P1 or P2 are applied in such a way that, in addition to the bits punctured for EGPRS, the following coded bits:
|
P1 |
{C(3+21j), C(12+21j) for j = 0,1,…,27} are not transmitted except {C(k) for k = 33,96,159,222,369,432,495,558} which are transmitted |
|
P2 |
{C(6+21j), C(18+21j) for j = 0,1,…,27} are not transmitted except {C(k) for k = 39,102,165,228,375,438,501,564} which are transmitted |
The result is a block of 324 coded bits {pc(0),pc(1),…,pc(323)}.
5.1.5.1.4a Piggy-backed Ack/Nack coding
The operations in this subclause shall be carried out only if a PAN field is included.
a) Parity bits
Ten PAN parity bits p(0),p(1),…,p(9) are defined in such a way that in GF(2) the binary polynomial:
d(209)D29 +…+ d(228)D10 + p(0)D9 +…+ p(9), when divided by:
D10 + D9 + D5 + D4 + D + 1, yields a remainder equal to:
D9 + D8 + D7 + D6 + D5 + D4 + D3 + D2 + D1 + 1.
The five bits {d(229),…,d(233)} (TFI value or 00000, see 3GPP TS 44.060) are added bit-wise modulo 2 to the 5 last parity bits {p(5),…,p(9)}. This results in the ten modified PAN parity bits {pt(0),…,pt(9)} defined as:
pt(k) = p(k) for k=0,…,4
pt(k) = d(k+224) + p(k) for k=5,…,9
b) Tail biting:
The six last modified PAN parity bits are added before information and modified PAN parity bits, the result being a block of 36 {u’’(-6),…,u’’(0),u’’(1),…,u’’(29)} bits with six negative indexes:
u’’(k-6) = pt(k+4) for k = 0,1,…,5
u’’(k) = d(k+209) for k = 0,1,…,19
u’’(k) = pt(k-20) for k = 20,21,…,29
c) Convolutional encoder
The block of 36 bits {u’’(-6),…,u’’(0),u’’(1),…,u’’(29)} is encoded with the 1/3 rate convolutional mother code defined by the polynomials:
G4 = 1 + D2 + D3 + D5 + D6
G7 = 1 + D + D2 + D3 + D6
G5 = 1 + D + D4 + D6
This results in a block of 90 coded bits {C(0),C(1),…,C(89)} defined by:
C(3k) = u"(k) + u"(k‑2) + u"(k‑3) + u"(k‑5) + u"(k‑6)
C(3k+1) = u"(k) + u"(k‑1) + u"(k‑2) + u"(k‑3) + u"(k‑6)
C(3k+2) = u"(k) + u"(k‑1) + u"(k‑4) + u"(k‑6) for k = 0,1,…,29
The block of 90 coded bits is punctured in such way that the following coded bits:
{ C(15k), C(2+15k), C(4+15k), C(6+15k), C(7+15k), C(10+15k), C(13+15k) for k = 0,1,…,5} are not transmitted.
The result is a block of 48 coded bits {ac(0),ac(1),…,ac(47)}.
The data coded bits {pc(0),pc(1),…,pc(323)} are appended to the PAN coded bits by the following rule:
dc(k) = ac(k) for k = 0,1,…,47
dc(k) = pc(k-48) for k = 48,49,…,371
The result is a block of 372 coded bits {dc(0),dc(1),…,dc(371)}.
5.1.5.1.5 Interleaving
The USF, header and data are put together as one entity as described by the following rule:
c(k) = u’(k) for k = 0,1,…,11
c(k) = hc(k-12) for k = 12,13,…,79
c(k) = dc(k‑80) for k = 80,81,…,451
c’(n,k) = c(n,k) for k = 0,1,…,24
c’(n,k) = c(n,k-1) for k = 26,27,…,81
c’(n,k) = c(n,k-2) for k = 83,84,…,138
c’(n,k) = c(n,k-3) for k = 140,141,…,423
c’(n,k) = c(n,k-4) for k = 425,426,…,455
c’(n,25) = q(8) c’(n,82) = q(9) c’(n,139) = q(10) c’(n,424) = q(11)
c(n,k) are the coded bits and q(8),q(9),…,q(11) = 0,0,0,0 are four extra stealing flags
The resulting block is interleaved according to the following rule:
i(B,j) = c’(n,k) for k = 0,1,…,455
n = 0,1,…,N,N+1,…
B = B0 + 4n + (k mod 4)
j = 2((49k) mod 57) + ((k mod 8) div 4)
5.1.5.1.6 Mapping on a burst
5.1.5.1.6.1 BTTI configuration
The mapping is given by the rule:
e(B,j) = i(B,j) and e(B,59+j) = i(B,57+j) for j = 0,1,…,56
and
e(B+m,57) = q(2m) and e(B+m,58) = q(2m+1) for m = 0,1,2,3
where
q(0),q(1),…,q(7) = 0,0,0,1,0,1,1,0.
Note: For a standard GPRS MS, bits q(0),…,q(7) indicates that the USF is coded as for CS-4.
5.1.5.1.6.2 RTTI configuration
a) Bit swapping
After the interleaving the following bits are swapped:
If the RTTI radio block is sent in the first 10ms of a 20ms block period:
Swap i(B+1,98) with i(B+1,0)
Swap i(B+1,35) with i(B+1,51)
Swap i(B+1,84) with i(B+1,100)
Swap i(B+2,98) with i(B+2,82)
Swap i(B+2,35) with i(B+2,19)
Swap i(B+2,84) with i(B+2,68)
Swap i(B+3,35) with i(B+3,3)
Swap i(B+3,84) with i(B+3,52)
Swap i(B+3,98) with i(B+3,66)
If the RTTI radio block is sent in the second 10ms of a 20ms block period:
Swap i(B,19) with i(B,51)
Swap i(B,68) with i(B,100)
Swap i(B,82) with i(B,0)
Swap i(B+1,19) with i(B+1,35)
Swap i(B+1,68) with i(B+1,84)
Swap i(B+1,82) with i(B+1,98)
Swap i(B+2,19) with i(B+2,3)
Swap i(B+2,68) with i(B+2,52)
Swap i(B+2,82) with i(B+2,66)
b) Mapping on bursts
The mapping is given by the rule:
e(B,j) = i(B,j) and e(B,59+j) = i(B,57+j) for j = 0,1,…,56
and
e(B+m,57) = q(2m) and e(B+m,58) = q(2m+1) for m = 0,1,2,3
where q(0),q(1),…,q(7) are set according to the following table, depending on the USF mode (see 3GPP TS 45.002):
|
in the first 10ms of a 20ms block period |
in the second 10ms of a 20ms block period |
|
|
USF sent in BTTI USF mode |
0,0,0,0,0,1,0,1 |
0,1,0,1,1,0,1,0 |
|
USF sent in RTTI USF mode |
0,0,0,1,0,1,1,0 |
|
c) Mapping on PDCHs
The bursts with B = 0,2 shall be mapped on the PDCH having the lower timeslot number, whereas the bursts with B = 1,3 shall be mapped on the PDCH having the higher timeslot number, see 3GPP TS 45.002.
5.1.5.2 Uplink (MCS-1 UL)
5.1.5.2.1 Block constitution
The block constitution is the same as for MCS-1 DL specified in subclause 5.1.5.1.1, with the exception that an eTFI field can only be included in combination with a PAN field, see sub-clause 5.1.5.2.3a.
5.1.5.2.2 Header coding
a) Parity bits:
Eight header parity bits p(0),p(1),…,p(7) are defined in such a way that in GF(2) the binary polynomial:
d(0)D38 +…+ d(30)D8 + p(0)D7 +…+ p(7), when divided by:
D8 + D6 + D3 + 1, yields a remainder equal to:
D7 + D6 + D5 + D4 + D3 + D2 + D+1.
b) Tail biting:
The six last header parity bits are added before information and parity bits, the result being a block of 45 bits {u"(‑6),…,u"(0),u"(1),…,u"(38)} with six negative indexes:
u"(k-6) = p(k+2) for k = 0,1,…,5
u"(k) = d(k) for k = 0,1,…,30
u"(k) = p(k‑31) for k = 31,32,…,38
c) Convolutional encoder
This block of 45 bits {u"(-6),…,u"(0),u"(1),…,u"(38)} is encoded with the 1/3 rate convolutional mother code defined by the polynomials:
G4 = 1 + D2 + D3 + D5 + D6
G7 = 1 + D + D2 + D3 + D6
G5 = 1 + D + D4 + D6
This results in a block of 117 coded bits: {C(0),C(1),…,C(116)} defined by:
C(3k) = u"(k) + u"(k‑2) + u"(k‑3) + u"(k‑5) + u"(k‑6)
C(3k+1) = u"(k) + u"(k‑1) + u"(k‑2) + u"(k‑3) + u"(k‑6)
C(3k+2) = u"(k) + u"(k‑1) + u"(k-4) + u"(k-6) for k = 0,1,…,38
The code is punctured in such a way that the following coded bits:
{C(5+12j), C(8+12j), C(11+12j), for j = 0,1,…,8} as well as {C(k) for k = 26,38,50,62,74,86,98,110,113,116} are not transmitted
The result is a block of 80 coded bits, {hc(0),hc(1),…,hc(79)}.
5.1.5.2.3 Data coding
The data coding is the same as for downlink as specified in subclause 5.1.5.1.4.
5.1.5.2.3a Piggy-backed Ack/Nack coding
If a PAN field is included and an eTFI field is not included, the PAN coding is the same as for the downlink as specified in subclause 5.1.5.1.4a.
If a PAN field and an eTFI field are included, the PAN coding is the same as for the downlink as specified in subclause 5.1.5.1.4a, with the addition that the five bits {d(229),d(230),d(231),d(232),d(233)} (TFI value or 00000, see 3GPP TS 44.060) and the three bits {d(234),d(235),d(236)} (eTFI value or 000, see 3GPP TS 44.060) are added bit-wise modulo 2 to the PAN parity bits {p(2),p(3),p(4),p(5),p(6),p(7),p(8),p(9)} resulting in the ten modified PAN parity bits defined as:
pt(k) = p(k) for k=0,…,1
pt(k) = d(k+232) + p(k) for k=2,…,4
pt(k) = d(k+224) + p(k) for k=5,…,9
5.1.5.2.4 Interleaving
The header and data are put together as one entity as described by the following rule:
c(k) = hc(k) for k = 0,1,…,79
c(k) = dc(k‑80) ) for k = 80,81,…,451
c’(n,k) = c(n,k) for k = 0,1,…,24
c’(n,k) = c(n,k-1) for k = 26,27,…,81
c’(n,k) = c(n,k-2) for k = 83,84,…,138
c’(n,k) = c(n,k-3) for k = 140,141,…,423
c’(n,k) = c(n,k-4) for k = 425,426,…,455
c’(n,25) = q(8) c’(n,82) = q(9) c’(n,139) = q(10) c’(n,424) = q(11)
c(n,k) are the coded bits and q(8),q(9),…,q(11) = 0,0,0,0 are four extra stealing flags
The resulting block is interleaved according to the following rule:
i(B,j) = c’(n,k) for k = 0,1,…,455
n = 0,1,…,N,N+1,…
B = B0 + 4n + (k mod 4)
j = 2((49k) mod 57) + ((k mod 8) div 4)
5.1.5.2.5 Mapping on a burst
In BTTI configuration, the mapping is the same as for MCS-1 DL as specified in subclause 5.1.5.1.6.1.
NOTE: This mapping is also applied in RTTI configuration.
In RTTI configuration, the bursts with B = 0,2 shall be mapped on the PDCH having the lower timeslot number, whereas the bursts with B = 1,3 shall be mapped on the PDCH having the higher timeslot number, see 3GPP TS 45.002.
5.1.6 Packet data block type 6 (MCS-2)
5.1.6.1 Downlink (MCS-2 DL)
5.1.6.1.1 Block constitution
The message delivered to the encoder has a fixed size of 257 information bits {d(0),d(1),…,d(256)}. It is delivered on a burst mode.
The message delivered to the encoder will have a fixed size of 260 information bits {d(0),d(1),…,d(259)}, if an eTFI field is included (see 3GPP TS 44.060).
The message delivered to the encoder will have a fixed size of 282 information bits {d(0),d(1),…,d(281)} if a PAN field is included.
The message delivered to the encoder will have a fixed size of 285 information bits {d(0),d(1),…,d(284)}, if a PAN field and an eTFI field are included (see 3GPP TS 44.060).
NOTE: The presence of the PAN is indicated by the PANI field in the header (see 3GPP TS 44.060).
5.1.6.1.2 USF precoding
5.1.6.1.2.1 BTTI configuration
The first three bits d(0),d(1),d(2) are block coded into twelve bits u’(0),u’(1),…,u’(11) as for Packet data block type 4 (CS-4) in subclause 5.1.4.2.
5.1.6.1.2.2 RTTI configuration
Twelve bits u’(0),u’(1),…,u’(11) are generated as described for MCS-1 DL in subclause 5.1.5.1.2.2.
5.1.6.1.3 Header coding
A block of 68 coded bits {hc(0),hc(1),…,hc(67)} is derived from {d(3),d(4),…,d(30)} as described for MCS-1 DL in subclause 5.1.5.1.3, with the exception that eTFI field bits are now located in {d(257),d(258),d(259)} if included when PAN is not included, and located in {d(282),d(283),d(284)} if included in combination with a PAN field.
5.1.6.1.4 Data coding
a) Parity bits:
Twelve data parity bits p(0),p(1),…,p(11) are defined in such a way that in GF(2) the binary polynomial:
d(31)D237 +…+ d(256)D12 + p(0)D11 +…+ p(11), when divided by:
D12 + D11 + D10 + D8 + D5 + D4 + 1, yields a remainder equal to:
D11 + D10 + D9 + D8 + D7 + D6 + D5 + D4 + D3 + D2 + D + 1.
b) Tail bits:
Six tail bits equal to 0 are added to the information and parity bits, the result being a block of 244 bits {u(0),u(1),…,u(243)}:
u(k) = d(k+31) for k = 0,1,…,225
u(k) = p(k‑226) for k = 226,227,…,237
u(k) = 0 for k = 238,239,…,243 (tail bits)
c) Convolutional encoder
This block of 244 bits {u(0),u(1),…,u(243)} is encoded with the 1/3 rate convolutional mother code defined by the polynomials:
G4 = 1 + D2 + D3 + D5 + D6
G7 = 1 + D + D2 + D3 + D6
G5 = 1 + D + D4 + D6
This results in a block of 732 coded bits: {C(0),C(1),…,C(731)} defined by:
C(3k) = u(k) + u(k‑2) + u(k‑3) + u(k‑5) + u(k‑6)
C(3k+1) = u(k) + u(k‑1) + u(k‑2) + u(k‑3) + u(k‑6)
C(3k+2) = u(k) + u(k‑1) + u(k-4) + u(k-6) for k = 0,1,…,243; u(k) = 0 for k < 0
The code is punctured depending on the value of the CPS field as defined in 3GPP TS 44.060. Two puncturing schemes named P1 or P2 are applied in such a way that the following coded bits:
|
P1 |
{C(6j), C(1+6j), C(5+6j) for j = 0,1,…,121} and {C(k) for k = 57,171,285,399,513,627} are transmitted |
|
P2 |
{C(2+6j), C(3+6j), C(4+6j) for j = 0,1,…,121} and {C(k) for k = 108,222,336,450,564,678} are transmitted |
The result is a block of 372 coded bits, {dc(0),dc(1),…,dc(371)}.
For the FANR procedure, the code is punctured depending on the CPS field and the PANI field as defined in 3GPP TS 44.060. If the PANI field is set to 0, the puncturing is the same as for EGPRS. If the PANI field is set to 1, the puncturing schemes named P1 or P2 are applied in such a way that, in addition to the bits punctured for EGPRS, the following coded bits:
|
P1 |
{C(18+30j), C(30+30j) for j = 0,1,…,23} are not transmitted |
|
P2 |
{C(9+30j), C(27+30j) for j = 0,1,…,23} are not transmitted |
The result is a block of 324 coded bits {pc(0),pc(1),…,pc(323)}.
5.1.6.1.4a Piggy-backed Ack/Nack coding
The operations in this subclause shall be carried out only if a PAN field is included.
A block of 48 coded bits {ac(0),ac(1),…,ac(47)} is derived from {d(257),d(258),…,d(281)} as described for MCS-1 DL in subclause 5.1.5.1.4a, with bits {d(209),d(210),…,d(233)} replaced by bits {d(257),d(258),…,d(281)}.
The data coded bits {pc(0),pc(1),…,pc(323)} are appended to the PAN coded bits as described for MCS-1 DL in subclause 5.1.5.1.4a. The result is a block of 372 coded bits {dc(0),dc(1),…,dc(371)}.
5.1.6.1.5 Interleaving
The interleaving is done as specified for MCS-1 DL in subclause 5.1.5.1.5.
5.1.6.1.6 Mapping on a burst
The mapping is done as specified for MCS-1 DL in subclause 5.1.5.1.6.
5.1.6.2 Uplink (MCS-2 UL)
5.1.6.2.1 Block constitution
The block constitution is the same as for MCS-2 DL specified in subclause 5.1.6.1.1, with the exception that an eTFI field can only be included in combination with a PAN field, see sub-clause 5.1.6.2.3a.
5.1.6.2.2 Header coding
A block of 80 coded bits {hc(0),hc(1),…,hc(79)} is derived from {d(0),d(1),…,d(30)} as described for MCS-1 UL in subclause 5.1.5.2.2.
5.1.6.2.3 Data coding
The data coding is the same as for downlink as specified in subclause 5.1.6.1.4.
5.1.6.2.3a Piggy-backed Ack/Nack coding
If a PAN field is included and an eTFI field is not included, its coding is the same as for downlink as specified in subclause 5.1.6.1.4a.
If a PAN field and an eTFI field are included, the coding of these fields are the same as for the downlink as specified in subclause 5.1.6.1.4a, with the addition that the TFI and eTFI values being added bit-wise modulo 2 to the last eight PAN parity bits as specified in subclause 5.1.5.2.3a.
5.1.6.2.4 Interleaving
The interleaving is the same as for MCS-1 UL as specified in subclause 5.1.5.2.4.
5.1.6.2.5 Mapping on a burst
The mapping is the same as for MCS-1 UL as specified in subclause 5.1.5. 2.5.
5.1.7 Packet data block type 7 (MCS-3)
5.1.7.1 Downlink (MCS-3 DL)
5.1.7.1.1 Block constitution
The message delivered to the encoder has a fixed size of 329 information bits {d(0),d(1),…,d(328)}. It is delivered on a burst mode.
The message delivered to the encoder will have a fixed size of 332 information bits {d(0),d(1),…,d(331)}, if an eTFI field is included (see 3GPP TS 44.060).
The message delivered to the encoder will have a fixed size of 354 information bits {d(0),d(1),…,d(353)} if a PAN field is included.
The message delivered to the encoder will have a fixed size of 357 information bits {d(0),d(1),…,d(356)}, if a PAN field and an eTFI field are included (see 3GPP TS 44.060).
NOTE: The presence of the PAN is indicated by the PANI field in the header (see 3GPP TS 44.060).
5.1.7.1.2 USF precoding
5.1.7.1.2.1 BTTI configuration
The first three bits d(0),d(1),d(2) are block coded into twelve bits u’(0),u’(1),…,u’(11) as for Packet data block type 4 (CS-4) in subclause 5.1.4.2.
5.1.7.1.2.2 RTTI configuration
Twelve bits u’(0),u’(1),…,u’(11) are generated as described for MCS-1 DL in subclause 5.1.5.1.2.2.
5.1.7.1.3 Header coding
A block of 68 coded bits {hc(0),hc(1),…,hc(67)} is derived from {d(3),d(4),…,d(30)} as described for MCS-1 DL in subclause 5.1.5.1.3, with the exception that eTFI field bits are now located in {d(329),d(330),d(331)} if included when PAN is not included, and located in {d(354),d(355),d(356)} if included in combination with a PAN field.
5.1.7.1.4 Data coding
a) Parity bits:
Twelve data parity bits p(0),p(1),…,p(11) are defined in such a way that in GF(2) the binary polynomial:
d(31)D309 +…+ d(328)D12 + p(0)D11 +…+ p(11), when divided by:
D12 + D11 + D10 + D8 + D5 + D4 + 1, yields a remainder equal to:
D11 + D10 + D9 + D8 + D7 + D6 + D5 + D4 + D3 + D2 + D + 1.
b) Tail bits:
Six tail bits equal to 0 are added to the information and parity bits, the result being a block of 316 bits
{u(0),u(1),…,u(315)}:
u(k) = d(k+31) for k = 0,1,…,297
u(k) = p(k‑298) for k = 298,299,…,309
u(k) = 0 for k = 310,311,…,315 (tail bits)
c) Convolutional encoder
This block of 316 bits {u(0),u(1),…,u(315)} is encoded with the 1/3 rate convolutional mother code defined by the polynomials:
G4 = 1 + D2 + D3 + D5 + D6
G7 = 1 + D + D2 + D3 + D6
G5 = 1 + D + D4 + D6
This results in a block of 948 coded bits: {C(0),C(1),…,C(947)} defined by:
C(3k) = u(k) + u(k‑2) + u(k‑3) + u(k‑5) + u(k‑6)
C(3k+1) = u(k) + u(k‑1) + u(k‑2) + u(k‑3) + u(k‑6)
C(3k+2) = u(k) + u(k‑1) + u(k-4) + u(k-6) for k = 0,1,…,315; u(k) = 0 for k < 0
The code is punctured depending on the value of the CPS field as defined in 3GPP TS 44.060. Three puncturing schemes named P1, P2 or P3 are applied in such a way that the following coded bits:
|
P1 |
{C(18j), C(1+18j), C(3+18j), C(6+18j), C(10+18j), C(14+18j), C(17+18j) for j = 0,1,…,51} |
|
P2 |
{C(2+18j), C(5+18j), C(6+18j), C(7+18j), C(9+18j), C(12+18j), C(16+18j) for j = 0,1,…,51} |
|
P3 |
{C(18j), C(4+18j), C(8+18j), C(11+18j), C(12+18j), C(13+18j), C(15+18j) for j = 0,1,…,51} |
The result is a block of 372 coded bits, {dc(0),dc(1),…,dc(371)}.
For the FANR procedure, the code is punctured depending on the CPS field and the PANI field as defined in 3GPP TS 44.060. If the PANI field is set to 0, the puncturing is the same as for EGPRS. If the PANI field is set to 1, the puncturing schemes named P1 or P2 are applied in such a way that, in addition to the bits punctured for EGPRS, the following coded bits:
|
P1 |
{C(18j) for j = 0,1,…,51} are not transmitted except {C(k) for k = 108,342,576,810} which are transmitted |
|
P2 |
{C(6+18j) for j = 0,1,…,51} are not transmitted except {C(k) for k = 186,294,528,762} which are transmitted |
|
P3 |
{C(12+18j) for j = 0,1,…,51} are not transmitted except {C(k) for k = 66,390,642,876} which are transmitted |
The result is a block of 324 coded bits {pc(0),pc(1),…,pc(323)}.
5.1.7.1.4a Piggy-backed Ack/Nack coding
The operations in this subclause shall be carried out only if a PAN field is included.
A block of 48 coded bits {ac(0),ac(1),…,ac(47)} is derived from {d(329),d(330),…,d(353)} as described for MCS-1 DL in subclause 5.1.5.1.4a, with bits {d(209),d(210),…,d(233)} replaced by bits {d(329),d(330),…,d(353)}.
The data coded bits {pc(0),pc(1),…,pc(323)} are appended to the PAN coded bits as described for MCS-1 DL in subclause 5.1.5.1.4a. The result is a block of 372 coded bits {dc(0),dc(1),…,dc(371)}.
5.1.7.1.5 Interleaving
The interleaving is done as specified for MCS-1 DL in subclause 5.1.5.1.5.
5.1.7.1.6 Mapping on a burst
The mapping is done as specified for MCS-1 DL in subclause 5.1.5.1.6.
5.1.7.2 Uplink (MCS-3 UL)
5.1.7.2.1 Block constitution
The block constitution is the same as for MCS-3 DL specified in subclause 5.1.7.1.1, with the exception that an eTFI field can only be included in combination with a PAN field, see sub-clause 5.1.7.2.3a.
5.1.7.2.2 Header coding
A block of 80 coded bits {hc(0),hc(1),…,hc(79)} is derived from {d(0),d(1),…,d(30)} as described for MCS-1 UL in subclause 5.1.5.2.2.
5.1.7.2.3 Data coding
The data coding is the same as for downlink as specified in subclause 5.1.7.1.4.
5.1.7.2.3a Piggy-backed Ack/Nack coding
If a PAN field is included and an eTFI field is not included, its coding is the same as for downlink as specified in subclause 5.1.7.1.4a.
If a PAN field and an eTFI field are included, the coding of theses fields are the same as for the downlink as specified in subclause 5.1.7.1.4a, with the addition that the TFI and eTFI values being added bit-wise modulo 2 to the last eight PAN parity bits as specified in subclause 5.1.5.2.3a.
5.1.7.2.4 Interleaving
The interleaving is the same as for MCS-1 UL as specified in subclause 5.1.5.2.4.
5.1.7.2.5 Mapping on a burst
The mapping is the same as for MCS-1 UL as specified in subclause 5.1.5. 2.5.
5.1.8 Packet data block type 8 (MCS-4)
5.1.8.1 Downlink (MCS-4 DL)
5.1.8.1.1 Block constitution
The message delivered to the encoder has a fixed size of 385 information bits {d(0),d(1),…,d(384)}. It is delivered on a burst mode.
The message delivered to the encoder will have a fixed size of 388 information bits {d(0),d(1),…,d(387)}, if an eTFI field is included (see 3GPP TS 44.060).
5.1.8.1.2 USF precoding
5.1.8.1.2.1 BTTI configuration
The first three bits d(0),d(1),d(2) are block coded into twelve bits u’(0),u’(1),…,u’(11) as for Packet data block type 4 (CS-4) in subclause 5.1.4.2.
5.1.8.1.2.2 RTTI configuration
Twelve bits u’(0),u’(1),…,u’(11) are generated as described for MCS-1 DL in subclause 5.1.5.1.2.2.
5.1.8.1.3 Header coding
A block of 68 coded bits {hc(0),hc(1),…,hc(67)} is derived from {d(3),d(4),…,d(30)} as described for MCS-1 DL in subclause 5.1.5.1.3, with the exception that eTFI field bits are now located in {d(385),d(386),d(387)} if included.
5.1.8.1.4 Data coding
a) Parity bits:
Twelve data parity bits p(0),p(1),…,p(11) are defined in such a way that in GF(2) the binary polynomial:
d(31)D365 +…+ d(384)D12 + p(0)D11 +…+ p(11), when divided by:
D12 + D11 + D10 + D8 + D5 + D4 + 1, yields a remainder equal to:
D11 + D10 + D9 + D8 + D7 + D6 + D5 + D4 + D3 + D2 + D + 1.
b) Tail bits:
Six tail bits equal to 0 are added to the information and parity bits, the result being a block of 372 bits {u(0),u(1),…,u(371)}:
u(k) = d(k+31) for k = 0,1,…,353
u(k) = p(k‑354) for k = 354,355,…,365
u(k) = 0 for k = 366,367,…,371 (tail bits)
c) Convolutional encoder
This block of 372 bits {u(0),u(1),…,u(371)} is encoded with the 1/3 rate convolutional mother code defined by the polynomials:
G4 = 1 + D2 + D3 + D5 + D6
G7 = 1 + D + D2 + D3 + D6
G5 = 1 + D + D4 + D6
This results in a block of 1116 coded bits: {C(0),C(1),…,C(1115)} defined by:
C(3k) = u(k) + u(k‑2) + u(k‑3) + u(k‑5) + u(k‑6)
C(3k+1) = u(k) + u(k‑1) + u(k‑2) + u(k‑3) + u(k‑6)
C(3k+2) = u(k) + u(k‑1) + u(k-4) + u(k-6) for k = 0,1,…, 371; u(k) = 0 for k < 0
The code is punctured depending on the value of the CPS field as defined in 3GPP TS 44.060. Three puncturing schemes named P1, P2 or P3 are applied in such a way that the following coded bits:
|
P1 |
{C(3j) for j = 0,1,…,371} are transmitted |
|
P2 |
{C(1+3j) for j = 0,1,…,371} are transmitted |
|
P3 |
{C(2+3j) for j = 0,1,…,371} are transmitted |
The result is a block of 372 coded bits, {dc(0),dc(1),…,dc(371)}.
5.1.8.1.5 Interleaving
The interleaving is done as specified for MCS-1 DL in subclause 5.1.5.1.5.
5.1.8.1.6 Mapping on a burst
The mapping is done as specified for MCS-1 DL in subclause 5.1.5.1.6.
5.1.8.2 Uplink (MCS-4 UL)
5.1.8.2.1 Block constitution
The message delivered to the encoder has a fixed size of 385 information bits {d(0),d(1),…,d(384)}. It is delivered on a burst mode.
5.1.8.2.2 Header coding
A block of 80 coded bits {hc(0),hc(1),…,hc(79)} is derived from {d(0),d(1),…,d(30)} as described for MCS-1 UL in subclause 5.1.5.2.2.
5.1.8.2.3 Data coding
The data coding is the same as for downlink as specified in subclause 5.1.8.1.4.
5.1.8.2.4 Interleaving
The interleaving is the same as for MCS-1 UL as specified in subclause 5.1.5.2.4.
5.1.8.2.5 Mapping on a burst
The mapping is the same as for MCS-1 UL as specified in subclause 5.1.5. 2.5.
5.1.9 Packet data block type 9 (MCS-5)
5.1.9.1 Downlink (MCS-5 DL)
5.1.9.1.1 Block constitution
The message delivered to the encoder has a fixed size of 478 information bits {d(0),d(1),…,d(477)}. It is delivered on a burst mode.
The message delivered to the encoder will have a fixed size of 481 information bits {d(0),d(1),…,d(480)}, if an eTFI field is included (see 3GPP TS 44.060).
The message delivered to the encoder will have a fixed size of 503 information bits {d(0),d(1),…,d(502)} if a PAN field is included.
The message delivered to the encoder will have a fixed size of 506 information bits {d(0),d(1),…,d(505)}, if a PAN field and an eTFI field are included (see 3GPP TS 44.060).
NOTE: The presence of the PAN is indicated by the PANI field in the header (see 3GPP TS 44.060).
5.1.9.1.2 USF precoding
5.1.9.1.2.1 BTTI configuration
The first three bits d(0),d(1),d(2) are block coded into 36 bits u’(0),u’(1),…,u’(35) according to the following table:
|
d(0),d(1),d(2) |
u’(0),u’(1),…,u’(35) |
|||
|
burst 0 |
burst 1 |
burst 2 |
burst 3 |
|
|
000 |
0 0 0 0 0 0 0 0 0 |
0 0 0 0 0 0 0 0 0 |
0 0 0 0 0 0 0 0 0 |
0 0 0 0 0 0 0 0 0 |
|
001 |
1 1 1 1 1 0 0 0 0 |
1 1 1 1 0 0 0 0 0 |
1 1 1 1 1 1 0 0 0 |
1 1 1 1 1 0 0 0 1 |
|
010 |
1 1 1 0 0 1 1 1 0 |
1 1 1 0 1 1 1 0 0 |
1 1 0 0 0 0 1 1 0 |
1 1 0 0 0 1 1 0 0 |
|
011 |
1 0 0 1 1 1 1 0 0 |
1 1 0 0 0 0 0 1 1 |
1 0 1 1 1 0 1 1 1 |
0 0 1 0 0 1 1 1 1 |
|
100 |
0 0 0 1 1 0 0 1 1 |
0 0 1 0 1 1 0 1 0 |
1 0 0 0 0 1 1 0 1 |
1 1 1 1 1 1 1 1 0 |
|
101 |
1 1 0 1 0 1 0 1 1 |
0 0 0 1 1 0 1 0 1 |
0 1 1 1 0 1 0 1 1 |
1 0 0 1 0 1 0 1 1 |
|
110 |
0 0 1 0 0 1 1 0 1 |
1 0 1 1 1 1 1 1 1 |
0 1 1 0 1 0 0 0 1 |
0 0 1 1 1 0 1 0 0 |
|
111 |
0 1 1 0 1 0 1 1 1 |
0 1 0 1 0 1 1 1 1 |
0 0 0 1 1 1 1 1 0 |
0 1 0 0 1 0 0 1 1 |
5.1.9.1.2.2 RTTI configurations
If the USF is sent in RTTI USF mode (see 3GPP TS 45.002) when data blocks are transmitted in RTTI configuration, then the first three bits d(0),d(1),d(2) are block coded into 36 bits u’(0),u’(1),…,u’(35) as described in subclause 5.1.9.1.2.1.
If the USF is sent in BTTI USF mode (see 3GPP TS 45.002) when data blocks are transmitted in RTTI configuration, then the three bits of the USF to be sent on the lower numbered PDCH of a corresponding downlink PDCH-pair are block coded into 36 bits uL(0),uL(1),…,uL(35) as described in subclause 5.1.9.1.2.1; the three bits of the USF to be sent on the higher numbered PDCH of a corresponding downlink PDCH-pair are block coded into 36 bits uH(0),uH(1),…,uH(35) as described in subclause 5.1.9.1.2.1.
NOTE: If BTTI USF mode is used when sending data blocks in RTTI configuration, then d(0),d(1),d(2) need not contain a USF; in this case, they are ignored by the encoder. How the USFs are delivered to the encoder in this case is implementation dependent.
If the data block is sent in the first 10ms of a 20ms block period, then:
u’(j)=uL(j), j=0…8
u’(j)=uH(j-9), j=9…17
u’(j)=uL(j-9), j=18…26
u’(j)=uH(j-18) j=27…35
If the data block is sent in the second 10ms of a 20ms block period, then:
u’(j)=uL(j+18), j=0…8
u’(j)=uH(j+9), j=9…17
u’(j)=uL(j+9), j=18…26
u’(j)=uH(j) j=27…35
NOTE: In case mixed modulation USF is used (see subclause 5.1), the USF bits sent during the other half of the 20 ms block period may be sent with a different modulation. In this case, the half of uL and uH not sent in the present data block will be discarded.
5.1.9.1.3 Header coding
a) Parity bits:
Eight header parity bits p(0),p(1),…,p(7) are defined in such a way that in GF(2) the binary polynomial:
d(3)D32 +…+ d(27)D8 + p(0)D7 +…+ p(7), when divided by:
D8 + D6 + D3 + 1, yields a remainder equal to:
D7 + D6 + D5 + D4 + D3 + D2 + D + 1.
In case an eTFI field (see 3GPP TS 44.060) is included and a PAN field is not included the three bits {d(478),d(479),d(480)} are added bit-wise modulo 2 to the last three parity bits {p(5), p(6),p(7)}. In case an eTFI field and a PAN field are included the three bits {d(503),d(504),d(505)} are added bit-wise modulo 2 to the last three parity bits {p(5), p(6),p(7)}. This result in eight modified header parity bits {pt(0),…,pt(7)} as specified in subclause 5.1.5.1.3.
b) Tail biting:
The six last header parity bits are added before information and parity bits, the result being a block of 39 bits {u"(‑6),…,u"(0),u"(1),…,u"(32)} with six negative indexes. In case an eTFI field is not included this operation is defined as:
u"(k-6) = p(k+2) for k = 0,1,…,5
u"(k) = d(k+3) for k = 0,1,…,24
u"(k) = p(k‑25) for k = 25,26,…,32
In case an eTFI field is included p(k) is replaced by pt(k) in the above operation.
c) Convolutional encoder
This block of 39 bits {u"(-6),…,u"(0),u"(1),…,u"(32)} is encoded with the 1/3 rate convolutional mother code defined by the polynomials:
G4 = 1 + D2 + D3 + D5 + D6
G7 = 1 + D + D2 + D3 + D6
G5 = 1 + D + D4 + D6
This results in a block of 99 coded bits: {C(0),C(1),…,C(98)} defined by:
C(3k) = u"(k) + u"(k‑2) + u"(k‑3) + u"(k‑5) + u"(k‑6)
C(3k+1) = u"(k) + u"(k‑1) + u"(k‑2) + u"(k‑3) + u"(k‑6)
C(3k+2) = u"(k) + u"(k‑1) + u"(k‑4) + u"(k‑6) for k = 0,1,…,32
A spare bit is added at the end of this block:
hc(k) = C(k) for k = 0,1,…,98
hc(99) = C(98)
The result is a block of 100 coded bits, {hc(0),hc(1),…,hc(99)}.
5.1.9.1.4 Data coding
a) Parity bits:
Twelve data parity bits p(0),p(1),…,p(11) are defined in such a way that in GF(2) the binary polynomial:
d(28)D461 +…+ d(477)D12 + p(0)D11 +…+ p(11), when divided by:
D12 + D11 + D10 + D8 + D5 + D4 + 1, yields a remainder equal to:
D11 + D10 + D9 + D8 + D7 + D6 + D5 + D4 + D3 + D2 + D + 1.
b) Tail bits:
Six tail bits equal to 0 are added to the information and parity bits, the result being a block of 468 bits {u(0),u(1),…,u(467)}:
u(k) = d(k+28) for k = 0,1,…,449
u(k) = p(k‑450) for k = 450,451,…,461
u(k) = 0 for k = 462,463,…,467 (tail bits)
c) Convolutional encoder
This block of 468 bits {u(0),u(1),…,u(467)} is encoded with the 1/3 rate convolutional mother code defined by the polynomials:
G4 = 1 + D2 + D3 + D5 + D6
G7 = 1 + D + D2 + D3 + D6
G5 = 1 + D + D4 + D6
This results in a block of 1404 coded bits: {C(0),C(1),…,C(1403)} defined by:
C(3k) = u(k) + u(k‑2) + u(k‑3) + u(k‑5) + u(k‑6)
C(3k+1) = u(k) + u(k‑1) + u(k‑2) + u(k‑3) + u(k‑6)
C(3k+2) = u(k) + u(k‑1) + u(k-4) + u(k-6) for k = 0,1,…,467; u(k) = 0 for k < 0
The code is punctured depending on the value of the CPS field as defined in 3GPP TS 44.060. Two puncturing schemes named P1 or P2 are applied in such a way that the following coded bits:
|
P1 |
{C(2+9j) for j = 0,1,…,153} as well as {C(1388+3j) for j = 0,1,…,5}are not transmitted |
|
P2 |
{C(1+9j) for j = 0,1,…,153} as well as {C(1387+3j) for j = 0,1,…,5}are not transmitted |
The result is a block of 1248 coded bits, {dc(0),dc(1),…,dc(1247)}.
For the FANR procedure, the code is punctured depending on the CPS field and the PANI field as defined in 3GPP TS 44.060. If the PANI field is set to 0, the puncturing is the same as for EGPRS. If the PANI field is set to 1, the puncturing schemes named P1 or P2 are applied in such a way that, in addition to the bits punctured for EGPRS, the following coded bits:
|
P1 |
{C(4+18j) for j = 0,1,…,76} are not transmitted except {C(k) for k = 526} which is transmitted |
|
P2 |
{C(14+18j) for j = 0,1,…,76} are not transmitted except {C(k) for k = 626} which is transmitted |
The result is a block of 1172 coded bits {pc(0),pc(1),…,pc(1171)}.
5.1.9.1.4a Piggy-backed Ack/Nack coding
The operations in this subclause shall be carried out only if a PAN field is included.
a) Parity bits
Ten PAN parity bits p(0), p(1),…,p(9) are defined in such a way that in GF(2) the binary polynomial:
d(478)D29 +…+ d(497)D10 + p(0)D9 +…+ p(9), when divided by:
D10 + D9 + D5 + D4 + D + 1, yields a remainder equal to:
D9 + D8 + D7 + D6 + D5 + D4 + D3 + D2 + D1 + 1.
The five bits {d(498),…,d(502)} (TFI value or 00000, see 3GPP TS 44.060) are added bit-wise modulo 2 to the 5 last parity bits {p(5),…,p(9)}. This results in the ten modified PAN parity bits {pt(0),…,pt(9)} defined as:
pt(k) = p(k) for k=0,…,4
pt(k) = d(k+493) + p(k) for k=5,…,9
b) Tail biting:
The six last modified PAN parity bits are added before information and modified PAN parity bits, the result being a block of 36 {u’’(-6),…,u’’(0),u’’(1),…,u’’(29)} bits with six negative indexes:
u’’(k-6) = pt(k+4) for k = 0,1,…,5
u’’(k) = d(k+478) for k = 0,1,…,19
u’’(k) = pt(k-20) for k = 20,21,…,29
c) Convolutional encoder
The block of 36 bits {u’’(-6),…,u’’(0),u’’(1),…,u’’(29)} is encoded with the 1/3 rate convolutional mother code defined by the polynomials:
G4 = 1 + D2 + D3 + D5 + D6
G7 = 1 + D + D2 + D3 + D6
G5 = 1 + D + D4 + D6
This results in a block of 90 coded bits {C(0),C(1),…,C(89)} defined by:
C(3k) = u"(k) + u"(k-2) + u"(k-3) + u"(k-5) + u"(k-6)
C(3k+1) = u"(k) + u"(k-1) + u"(k-2) + u"(k-3) + u"(k-6)
C(3k+2) = u"(k) + u"(k-1) + u"(k-4) + u"(k-6) for k = 0,1,…,29
The block of 90 coded bits is punctured in such way that the following coded bits:
{ C(5+6k), C(50+6k) for k = 0,1,…,6} are not transmitted.
The result is a block of 76 coded bits {ac(0),ac(1),…,ac(75)}.
The data coded bits {pc(0),pc(1),…,pc(1171)} are appended to the PAN coded bits by the following rule:
dc(k) = ac(k) for k = 0,1,…,75
dc(k) = pc(k-76) for k = 76,49,…,1247
The result is a block of 1248 coded bits {dc(0),dc(1),…,dc(1247)}.
5.1.9.1.5 Interleaving
a) Header
The 100 coded bits of the header, {hc(0),hc(1),…,hc(99)}, are interleaved according to the following rule:
hi(j) = hc(k) for k = 0,1,…,99
j = 25(k mod 4) + ((17k) mod 25)
b) Data
There is no closed expression describing the interleaver, but it has been derived taking the following approach:
1. A block interleaver with a 1392 bit block size is defined:
The kth input data bit is mapped to the jth bit of the Bth burst, where
k = 0,…,1391
B = mod(k,4)
d = mod(k,464)
j = 3*(2mod(25d,58) + div(mod(d,8),4) + 2(-1)Bdiv(d,232)) + mod(k,3)
2. The data bit positions being mapped onto header positions in the interleaved block are removed (the header positions are j = 156,157,…,191 when the header is placed next to the training sequence. This leaves 1248 bits in the mapping.
3. The bits are renumbered to fill out the gaps both in j and k, without changing the relative order
The resulting interleaver transform the block of 1248 coded bits, {dc(0),dc(1),…,dc(1247)} into a block of 1248 interleaved bits, {di(0),di(1),…,di(1247)}.
di(j’) = dc(k’) for k’ = 0,1,…,1247
(An explicit relation between j’ and k’ is given in table 15)
5.1.9.1.6 Mapping on a burst
a) Straightforward Mapping
The mapping is given by the rule:
For B=0,1,2,3, let
e(B,j) = di(312B+j) for j = 0,1,…,155
e(B,j) = hi(25B+j-156) for j = 156,157,…,167
e(B,j) = u’(9B+j-168) for j = 168,169,…,173
e(B,j) = q(2B+j-174) for j = 174,175
e(B,j) = u’(9B+j-170) for j = 176,177,178
e(B,j) = hi(25B+j-167) for j = 179,180,…,191
e(B,j) = di(312B+j-36) for j = 192,193,…,347
where
q(0),q(1),…,q(7) = 0,0,0,0,0,0,0,0 identifies the coding scheme MCS-5 or MCS-6.
b) Bit swapping
After this mapping the following bits are swapped:
For B = 0,1,2,3,
Swap e(B,142) with e(B,155)
Swap e(B,144) with e(B,158)
Swap e(B,145) with e(B,161)
Swap e(B,147) with e(B,164)
Swap e(B,148) with e(B,167)
Swap e(B,150) with e(B,170)
Swap e(B,151) with e(B,173)
Swap e(B,176) with e(B,195)
Swap e(B,179) with e(B,196)
Swap e(B,182) with e(B,198)
Swap e(B,185) with e(B,199)
Swap e(B,188) with e(B,201)
Swap e(B,191) with e(B,202)
Swap e(B,194) with e(B,204).
In RTTI configuration, the bursts with B = 0,2 shall be mapped on the PDCH having the lower timeslot number, whereas the bursts with B = 1,3 shall be mapped on the PDCH having the higher timeslot number, see 3GPP TS 45.002.
c) PAN bit swapping
In case a PAN is included in the radio block, the following additional bits are swapped after the bit swapping in b):
For B = 0
Swap e(B,50) with e(B,49)
Swap e(B,113) with e(B,112)
Swap e(B,167) with e(B,154)
Swap e(B,221) with e(B,220)
Swap e(B,278) with e(B,277)
Swap e(B,341) with e(B,340)
For B = 1
Swap e(B,8) with e(B,7)
Swap e(B,59) with e(B,58)
Swap e(B,71) with e(B,70)
Swap e(B,116) with e(B,115)
Swap e(B,173) with e(B,154)
Swap e(B,182) with e(B,193)
Swap e(B,236) with e(B,235)
Swap e(B,299) with e(B,298)
For B = 2
Swap e(B,17) with e(B,16)
Swap e(B,74) with e(B,73)
Swap e(B,137) with e(B,136)
Swap e(B,257) with e(B,256)
Swap e(B,302) with e(B,301)
Swap e(B,314) with e(B,313)
For B = 3
Swap e(B,32) with e(B,31)
Swap e(B,95) with e(B,94)
Swap e(B,152) with e(B,154)
Swap e(B,215) with e(B,214)
Swap e(B,260) with e(B,259)
Swap e(B,272) with e(B,271)
Swap e(B,323) with e(B,322)
5.1.9.2 Uplink (MCS-5 UL)
5.1.9.2.1 Block constitution
The message delivered to the encoder has a fixed size of 487 information bits {d(0),d(1),…,d(486)}. It is delivered on a burst mode.
The message delivered to the encoder will have a fixed size of 512 information bits {d(0),d(1),…,d(511)} if a PAN field is included (see 3GPP TS 44.060).
The message delivered to the encoder will have a fixed size of 515 information bits {d(0),d(1),…,d(514)}, if a PAN field and an eTFI field are included (see 3GPP TS 44.060).
NOTE: The presence of the PAN is indicated by the PANI field in the header (see 3GPP TS 44.060).
5.1.9.2.2 Header coding
a) Parity bits:
Eight header parity bits p(0),p(1),…,p(7) are defined in such a way that in GF(2) the binary polynomial:
d(0)D44 +…+ d(36)D8 + p(0)D7 +…+ p(7), when divided by:
D8 + D6 + D3 + 1, yields a remainder equal to:
D7 + D6 + D5 + D4 + D3 + D2 + D+1.
b) Tail biting:
The six last header parity bits are added before information and parity bits, the result being a block of 51 bits {u"(‑6),…,u"(0),u"(1),…,u"(44)} with six negative indexes:
u"(k-6) = p(k+2) for k = 0,1,…,5
u"(k) = d(k) for k = 0,1,…,36
u"(k) = p(k‑37) for k = 37,38,…,44
c) Convolutional encoder
This block of 51 bits {u"(-6),…,u"(0),u"(1),…,u"(44)} is encoded with the 1/3 rate convolutional mother code defined by the polynomials:
G4 = 1 + D2 + D3 + D5 + D6
G7 = 1 + D + D2 + D3 + D6
G5 = 1 + D + D4 + D6
This results in a block of 135 coded bits: {C(0),C(1),…,C(134)} defined by:
C(3k) = u"(k) + u"(k‑2) + u"(k‑3) + u"(k‑5) + u"(k‑6)
C(3k+1) = u"(k) + u"(k‑1) + u"(k‑2) + u"(k‑3) + u"(k‑6)
C(3k+2) = u"(k) + u"(k‑1) + u"(k-4) + u"(k-6) for k = 0,1,…,44
The code is punctured in such a way that the following coded bits:
hc(k) = C(k) for k = 0,1,…,134
hc(135) = C(134)
The result is a block of 136 coded bits, {hc(0),hc(1),…,hc(135)}.
5.1.9.2.3 Data coding
The data coding is the same as for downlink as specified in subclause 5.1.9.1.4 where bits {d(28),d(29),…,d(477)} are replaced by bits {d(37),d(38),…,d(486)}.
5.1.9.2.3a Piggy-backed Ack/Nack coding
If a PAN field is included and an eTFI field is not included, the coding of the PAN field is the same as for the downlink as specified in subclause 5.1.9.1.4a where bits {d(478), d(479),…,d(502)} are replaced by bits {d(487), d(488),…,d(511)}.
If a PAN field and an eTFI field are included, the PAN coding is the same as for the downlink as specified in sub-clause 5.1.9.1.4a, with the exception that the five bits {d(507),d(508),d(509),d(510),d(511)} (TFI value or 00000, see 3GPP TS 44.060) and the three bits {d(512),d(513),d(514)} (eTFI value or 000, see 3GPP TS 44.060) are added bit-wise modulo 2 to the PAN parity bits {p(2),p(3),p(4),p(5),p(6),p(7),p(8),p(9)} resulting in the ten modified PAN parity bits defined as:
pt(k) = p(k) for k=0,…,1
pt(k) = d(k+510) + p(k) for k=2,…,4
pt(k) = d(k+502) + p(k) for k=5,…,9
The data coded bits {pc(0),pc(1),…,pc(1171)} are appended to the PAN coded bits as described for the downlink in subclause 5.1.9.1.4a. The result is a block of 1248 coded bits {dc(0),dc(1),…,dc(1247)}.
5.1.9.2.4 Interleaving
a) Header
The 136 coded bits of the header, {hc(0),hc(1),…,hc(135)}, are interleaved according to the following rule:
hi(j) = hc(k) for k = 0,1,…,135
j = 34(k mod 4) + 2((11k) mod 17) + [(k mod 8)/4]
b) Data
The data interleaving is the same as for MCS-5 DL as specified in subclause 5.1.9.1.5.
5.1.9.2.5 Mapping on a burst
a) Straightforward Mapping
The mapping is given by the rule:
For B=0,1,2,3, let
e(B,j) = di(312B+j) for j = 0,1,…,155
e(B,j) = hi(34B+j-156) for j = 156,157,…,173
e(B,j) = q(2B+j-174) for j = 174,175
e(B,j) = hi(34B+j-158) for j = 176,177,…,191
e(B,j) = di(312B+j-36) for j = 192,193,…,347
where
q(0),q(1),…,q(7) = 0,0,0,0,0,0,0,0 identifies the coding scheme MCS-5 or MCS-6.
b) Bit swapping
The bit swapping is the same as for MCS-5 DL as specified in subclause 5.1.9.1.6 b).
In RTTI configuration, the bursts with B = 0,2 shall be mapped on the PDCH having the lower timeslot number, whereas the bursts with B = 1,3 shall be mapped on the PDCH having the higher timeslot number, see 3GPP TS 45.002.
c) PAN bit swapping
In case a PAN is included in the radio block, additional bits are swapped as specified in subclause 5.1.9.1.6 c).
5.1.10 Packet data block type 10 (MCS-6)
5.1.10.1 Downlink (MCS-6 DL)
5.1.10.1.1 Block constitution
The message delivered to the encoder has a fixed size of 622 information bits {d(0),d(1),…,d(621)}. It is delivered on a burst mode.
The message delivered to the encoder will have a fixed size of 625 information bits {d(0),d(1),…,d(624)}, if an eTFI field is included (see 3GPP TS 44.060).
The message delivered to the encoder will have a fixed size of 647 information bits {d(0),d(1),…,d(646)} if a PAN field is included.
The message delivered to the encoder will have a fixed size of 650 information bits {d(0),d(1),…,d(649)}, if a PAN field and an eTFI field are included (see 3GPP TS 44.060).
NOTE: The presence of the PAN is indicated by the PANI field in the header (see 3GPP TS 44.060).
5.1.10.1.2 USF precoding
5.1.10.1.2.1 BTTI configuration
A block of 36 bits {u’(0),u’(1),…,u’(35)} is derived from {d(0),d(1),d(2)} as described for MCS-5 DL in subclause 5.1.9.1.2. 1.
5.1.10.1.2.2 RTTI configuration
A block of 36 bits {u’(0),u’(1),…,u’(35)} is generated as described for MCS-5 DL in subclause 5.1.9.1.2.2.
5.1.10.1.3 Header coding
A block of 100 coded bits {hc(0),hc(1),…,hc(99)} is derived from {d(3),d(4),…,d(27)} as described for MCS-5 DL in subclause 5.1.9.1.3, with the exception that eTFI field bits are now located in {d(622),d(623),d(624)} if included when PAN is not included, and located in {d(647),d(648),d(649)} if included in combination with a PAN field.
5.1.10.1.4 Data coding
a) Parity bits:
Twelve data parity bits p(0),p(1),…,p(11) are defined in such a way that in GF(2) the binary polynomial:
d(28)D605 +…+ d(621)D12 + p(0)D11 +…+ p(11), when divided by:
D12 + D11 + D10 + D8 + D5 + D4 + 1, yields a remainder equal to:
D11 + D10 + D9 + D8 + D7 + D6 + D5 + D4 + D3 + D2 + D + 1.
b) Tail bits:
Six tail bits equal to 0 are added to the information and parity bits, the result being a block of 612 bits {u(0),u(1),…,u(611)}:
u(k) = d(k+28) for k = 0,1,…,593
u(k) = p(k‑594) for k = 594,595,…,605
u(k) = 0 for k = 606,607,…,611 (tail bits)
c) Convolutional encoder
This block of 612 bits {u(0),u(1),…,u(611)} is encoded with the 1/3 rate convolutional mother code defined by the polynomials:
G4 = 1 + D2 + D3 + D5 + D6
G7 = 1 + D + D2 + D3 + D6
G5 = 1 + D + D4 + D6
This results in a block of 1836 coded bits: {C(0),C(1),…,C(1835)} defined by:
C(3k) = u(k) + u(k‑2) + u(k‑3) + u(k‑5) + u(k‑6)
C(3k+1) = u(k) + u(k‑1) + u(k‑2) + u(k‑3) + u(k‑6)
C(3k+2) = u(k) + u(k‑1) + u(k-4) + u(k-6) for k = 0,1,…,611; u(k) = 0 for k < 0
The code is punctured depending on the value of the CPS field as defined in 3GPP TS 44.060. Two puncturing schemes named P1 or P2 are applied in such a way that the following coded bits:
|
P1 |
{C(2+3j) for j = 0,1,…,611} are not transmitted except {C(k) for k = 32,98,164,230,296,428,494,560, 626,692,824,890,956,1022,1088,1220,1286,1352,1418,1484,1616,1682,1748,1814} which are transmitted |
|
P2 |
{C(1+3j) for j = 0,1,…,611} are not transmitted except {C(k) for k = 16,82,148,214,280,412,478,544, 610,676,808,874,940,1006,1072,1204,1270,1336,1402,1468,1600,1666,1732,1798} which are transmitted |
The result is a block of 1248 coded bits, {dc(0),dc(1),…,dc(1247)}.
For the FANR procedure, the code is punctured depending on the CPS field and the PANI field as defined in 3GPP TS 44.060. If the PANI field is set to 0, the puncturing is the same as for EGPRS. If the PANI field is set to 1, the puncturing schemes named P1 or P2 are applied in such a way that, in addition to the bits punctured for EGPRS, the following coded bits:
|
P1 |
{C(6+24j) for j = 0,1,…,75} are not transmitted |
|
P2 |
{C(18+24j) for j = 0,1,…,75} are not transmitted |
The result is a block of 1172 coded bits {pc(0),pc(1),…,pc(1171)}.
5.1.10.1.4a Piggy-backed Ack/Nack coding
The operations in this subclause shall be carried out only if a PAN field is included.
A block of 76 coded bits {ac(0),ac(1),…,ac(75)} is derived from {(622),d(623),…,d(646)} as described for MCS-5 DL in subclause 5.1.9.1.4a, with bits {d(478),d(479),…,d(502)} replaced by bits {d(622),d(623),…,d(646)}.
The data coded bits {pc(0),pc(1),…,pc(1171)} are appended to the PAN coded bits as described for MCS-5 DL in subclause 5.1.9.1.4a. The result is a block of 1248 coded bits {dc(0),dc(1),…,dc(1247)}.
5.1.10.1.5 Interleaving
The interleaving is done as specified for MCS-5 DL in subclause 5.1.9.1.5.
5.1.10.1.6 Mapping on a burst
The mapping is done as specified for MCS-5 DL in subclause 5.1.9.1.6.
5.1.10.2 Uplink (MCS-6 UL)
5.1.10.2.1 Block constitution
The message delivered to the encoder has a fixed size of 631 information bits {d(0),d(1),…,d(630)}. It is delivered on a burst mode.
The message delivered to the encoder will have a fixed size of 656 information bits {d(0),d(1),…,d(655)} if a PAN field is included (see 3GPP TS 44.060).
The message delivered to the encoder will have a fixed size of 659 information bits {d(0),d(1),…,d(658)}, if a PAN field and an eTFI field are included (see 3GPP TS 44.060).
NOTE: The presence of the PAN is indicated by the PANI field in the header (see 3GPP TS 44.060).
5.1.10.2.2 Header coding
A block of 136 coded bits {hc(0),hc(1),…,hc(135)} is derived from {d(0),d(1),…,d(36)} as described for MCS-5 UL in subclause 5.1.9.2.2.
5.1.10.2.3 Data coding
The data coding is the same as for downlink as specified in subclause 5.1.10.1.4 where bits {d(28),d(29),…,d(621)} are replaced by bits {d(37),d(38),…,d(630)}.
5.1.10.2.3a Piggy-backed Ack/Nack coding
If a PAN field is included and an eTFI field is not included, the coding of the PAN field is the same as for the MCS-5 DL as specified in subclause 5.1.9.1.4a where bits {d(478), d(479),…,d(502)} are replaced by bits {d(631),d(632),…,d(655)}.
If a PAN field and an eTFI field are included, the PAN coding is the same as for the downlink as specified in sub-clause 5.1.10.1.4a, with the exception that the five bits {d(651),d(652),d(653),d(654),d(655)} (TFI value or 00000, see 3GPP TS 44.060) and the three bits {d(656),d(657),d(658)} (eTFI value or 000, see 3GPP TS 44.060) are added bit-wise modulo 2 to the PAN parity bits {p(2),p(3),p(4),p(5),p(6),p(7),p(8),p(9)} resulting in the ten modified PAN parity bits defined as:
pt(k) = p(k) for k=0,…,1
pt(k) = d(k+654) + p(k) for k=2,…,4
pt(k) = d(k+646) + p(k) for k=5,…,9
The data coded bits {pc(0),pc(1),…,pc(1171)} are appended to the PAN coded bits as described for MCS-5 DL in subclause 5.1.9.1.4a. The result is a block of 1248 coded bits {dc(0),dc(1),…,dc(1247)}.
5.1.10.2.4 Interleaving
The interleaving is the same as for MCS-5 UL as specified in subclause 5.1.9.2.4.
5.1.10.2.5 Mapping on a burst
The mapping is the same as for MCS-5 UL as specified in subclause 5.1.9.2.5.
5.1.11 Packet data block type 11 (MCS-7)
5.1.11.1 Downlink (MCS-7 DL)
5.1.11.1.1 Block constitution
The message delivered to the encoder has a fixed size of 940 information bits {d(0),d(1),…,d(939)}. It is delivered on a burst mode.
The message delivered to the encoder will have a fixed size of 943 information bits {d(0),d(1),…,d(942)}, if an eTFI field is included (see 3GPP TS 44.060).
The message delivered to the encoder will have a fixed size of 965 information bits {d(0),d(1),…,d(964)} if a PAN field is included.
The message delivered to the encoder will have a fixed size of 968 information bits {d(0),d(1),…,d(967)}, if a PAN field and an eTFI field are included (see 3GPP TS 44.060).
NOTE: The presence of the PAN is indicated by the PANI field in the header (see 3GPP TS 44.060).
5.1.11.1.2 USF precoding
5.1.11.1.2.1 BTTI configuration
A block of 36 bits {u’(0),u’(1),…,u’(35)} is derived from {d(0),d(1),d(2)} as described for MCS-5 DL in subclause 5.1.9.1.2.1.
5.1.11.1.2.2 RTTI configuration
A block of 36 bits {u’(0),u’(1),…,u’(35)} is generated as described for MCS-5 DL in subclause 5.1.9.1.2.2.
5.1.11.1.3 Header coding
a) Parity bits:
Eight header parity bits p(0),p(1),…,p(7) are defined in such a way that in GF(2) the binary polynomial:
d(3)D44 +…+ d(39)D8 + p(0)D7 +…+ p(7), when divided by:
D8 + D6 + D3 + 1, yields a remainder equal to:
D7 + D6 + D5 + D4 + D3 + D2 + D + 1.
In case an eTFI field (see 3GPP TS 44.060) is included and a PAN field is not included the three bits {d(940),d(941),d(942)} are added bit-wise modulo 2 to the last three parity bits {p(5), p(6),p(7)}. In case an eTFI field and a PAN field are included the three bits {d(965),d(966),d967)} are added bit-wise modulo 2 to the last three parity bits {p(5), p(6),p(7)}. This result in eight modified header parity bits {pt(0),…,pt(7)} as specified in subclause 5.1.5.1.3.
b) Tail biting:
The six last header parity bits are added before information and parity bits, the result being a block of 51 bits {u"(‑6),…,u"(0),u"(1),…,u"(44)} with six negative indexes. In case an eTFI field is not included this operation is defined as:
u"(k-6) = p(k+2) for k = 0,1,…,5
u"(k) = d(k+3) for k = 0,1,…,36
u"(k) = p(k‑37) for k = 37,38,…,44
In case an eTFI field is included p(k) is replaced by pt(k) in the above operation.
c) Convolutional encoder
This block of 51 bits {u"(-6),…,u"(0),u"(1),…,u"(44)} is encoded with the 1/3 rate convolutional mother code defined by the polynomials:
G4 = 1 + D2 + D3 + D5 + D6
G7 = 1 + D + D2 + D3 + D6
G5 = 1 + D + D4 + D6
This results in a block of 135 coded bits: {C(0),C(1),…,C(134)} defined by:
C(3k) = u"(k) + u"(k‑2) + u"(k‑3) + u"(k‑5) + u"(k‑6)
C(3k+1) = u"(k) + u"(k‑1) + u"(k‑2) + u"(k‑3) + u"(k‑6)
C(3k+2) = u"(k) + u"(k‑1) + u"(k-4) + u"(k-6) for k = 0,1,…,44
The code is punctured in such a way that the following coded bits:
{C(k) for k = 14,23,33,50,59,69,86,95,105,122,131} are not transmitted
The result is a block of 124 coded bits, {hc(0),hc(1),…,hc(123)}.
5.1.11.1.4 Data coding
I) First half:
a) Parity bits:
Twelve data parity bits p(0),p(1),…,p(11) are defined in such a way that in GF(2) the binary polynomial:
d(40)D461 +…+ d(489)D12 + p(0)D11 +…+ p(11), when divided by:
D12 + D11 + D10 + D8 + D5 + D4 + 1, yields a remainder equal to:
D11 + D10 + D9 + D8 + D7 + D6 + D5 + D4 + D3 + D2 + D + 1.
b) Tail bits:
Six tail bits equal to 0 are added to the information and parity bits, the result being a block of 468 bits {u(0),u(1),…,u(467)}:
u(k) = d(k+40) for k = 0,1,…,449
u(k) = p(k‑450) for k = 450,451,…,461
u(k) = 0 for k = 462,463,…,467 (tail bits)
c) Convolutional encoder
This block of 468 bits {u(0),u(1),…,u(467)} is encoded with the 1/3 rate convolutional mother code defined by the polynomials:
G4 = 1 + D2 + D3 + D5 + D6
G7 = 1 + D + D2 + D3 + D6
G5 = 1 + D + D4 + D6
This results in a block of 1404 coded bits: {C(0),C(1),…,C(1403)} defined by:
C(3k) = u(k) + u(k‑2) + u(k‑3) + u(k‑5) + u(k‑6)
C(3k+1) = u(k) + u(k‑1) + u(k‑2) + u(k‑3) + u(k‑6)
C(3k+2) = u(k) + u(k‑1) + u(k-4) + u(k-6) for k = 0,1,…,467; u(k) = 0 for k < 0
The code is punctured depending on the value of the CPS field as defined in 3GPP TS 44.060. Three puncturing schemes named P1, P2 or P3 are applied in such a way that the following coded bits:
|
P1 |
{C(18j), C(1+18j), C(4+18j), C(8+18j), C(11+18j), C(12+18j), C(13+18j), C(15+18j) |
|
P2 |
{C(2+18j), C(3+18j), C(5+18j), C(6+18j), C(10+18j), C(14+18j), C(16+18j), C(17+18j) |
|
P3 |
{C(2+18j), C(5+18j), C(6+18j), C(7+18j), C(9+18j), C(12+18j), C(13+18j), C(16+18j) |
The result is a block of 612 coded bits, {c1(0),c1(1),…,c1(611)}.
For the FANR procedure, the code is punctured depending on the CPS field and the PANI field as defined in 3GPP TS 44.060. If the PANI field is set to 0, the puncturing is the same as for EGPRS. If the PANI field is set to 1, the puncturing schemes named P1 or P2 are applied in such a way that, in addition to the bits punctured for EGPRS, the following coded bits:
|
P1 |
{C(13+36j) for j = 0,1,…,38} are not transmitted except {C(k) for k = 13,49,1381} which are transmitted |
|
P2 |
{C(5+36j) for j = 0,1,…,38} are not transmitted except {C(k) for k = 185,545,1085} which are transmitted |
|
P3 |
{C(6+36j) for j = 0,1,…,38} are not transmitted except {C(k) for k = 294,654,1194} which are transmitted |
The result is a block of 576 coded bits {pc1(0),pc1(1),…,pc1(575)}.
II) Second half:
The same data coding as for first half is proceeded with bits {d(40),d(41),…,d(489)} replaced by bits {d(490),d(491),…,d(939)}. The result is a block of 612 coded bits, {c2(0),c2(1),…,c2(611)}.
If the PANI field is set to 1, additional bits are punctured as for the first half. The result is a block of 576 coded bits {pc2(0),pc2(1),…,pc2(575)}.
5.1.11.1.4a Piggy-backed Ack/Nack coding
The operations in this subclause shall be carried out only if a PAN field is included.
a) Parity bits
Ten PAN parity bits p(0), p(1),…,p(9) are defined in such a way that in GF(2) the binary polynomial:
d(940)D29 +…+ d(959)D10 + p(0)D9 +…+ p(9), when divided by:
D10 + D9 + D5 + D4 + D + 1, yields a remainder equal to:
D9 + D8 + D7 + D6 + D5 + D4 + D3 + D2 + D1 + 1.
The five bits {d(960),…,d(964)} (TFI value or 00000, see 3GPP TS 44.060) are added bit-wise modulo 2 to the 5 last parity bits {p(5),…,p(9)}. This results in the ten modified PAN parity bits {pt(0),…,pt(9)} defined as:
pt(k) = p(k) for k=0,…,4
pt(k) = d(k+955) + p(k) for k=5,…,9
b) Tail biting:
The six last modified PAN parity bits are added before information and modified PAN parity bits, the result being a block of 36 {u’’(-6),…,u’’(0),u’’(1),…,u’’(29)} bits with six negative indexes:
u’’(k-6) = pt(k+4) for k = 0,1,…,5
u’’(k) = d(k+940) for k = 0,1,…,19
u’’(k) = pt(k-20) for k = 20,21,…,29
c) Convolutional encoder
The block of 36 bits {u’’(-6),…,u’’(0),u’’(1),…,u’’(29)} is encoded with the 1/3 rate convolutional mother code defined by the polynomials:
G4 = 1 + D2 + D3 + D5 + D6
G7 = 1 + D + D2 + D3 + D6
G5 = 1 + D + D4 + D6
This results in a block of 90 coded bits {C(0),C(1),…,C(89)} defined by:
C(3k) = u"(k) + u"(k-2) + u"(k-3) + u"(k-5) + u"(k-6)
C(3k+1) = u"(k) + u"(k-1) + u"(k-2) + u"(k-3) + u"(k-6)
C(3k+2) = u"(k) + u"(k-1) + u"(k-4) + u"(k-6) for k = 0,1,…,29
The block of 90 coded bits is punctured in such way that the following coded bits:
{ C(2+15k), C(8+15k), C(14+15k) for k = 0,1,…,5} are not transmitted.
The result is a block of 72 coded bits {ac(0),ac(1),…,ac(71)}.
The data coded bits {pc1(0),pc1(1),…,pc1(575)} and {pc2(0),pc2(1),…,pc2(575)} are appended to the PAN coded bits by the following rule:
c1(k) = ac(k) for k = 0,1,…,71
c1(k) = pc1(k-72) for k = 72,73,…,611
c2(k) = pc1(k+540) for k = 0,1,…,35
c2(k) = pc2(k-36) for k = 36,37,…,611
The result is two blocks of 612 coded bits {c1(0),c1(1),…,c1(611)} and {c2(0),c2(1),…,c2(611)}.
5.1.11.1.5 Interleaving
a) Header
The 124 coded bits of the header, {hc(0),hc(1),…,hc(123)}, are interleaved according to the following rule:
hi(j) = hc(k) for k = 0,1,…,123
j = 31(k mod 4) + ((17k) mod 31)
b) Data
Data are put together as one entity as described by the following rule:
dc(k) = c1(k) for k = 0,1,…,611
dc(k) = c2(k-612) for k = 612,613,…,1223
The resulting block is interleaved according to the following rule:
di(j) = dc(k) for k = 0,1,…,1223
j = 306(k mod 4) + 3((44k) mod 102 + (k div 4) mod 2) + (k + 2 – (k div 408)) mod 3
5.1.11.1.6 Mapping on a burst
a) Straightforward Mapping
The mapping is given by the rule:
For B=0,1,2,3, let
e(B,j) = di(306B+j) for j = 0,1,…,152
e(B,j) = hi(31B+j-153) for j = 153,154,…,167
e(B,j) = u’(9B+j-168) for j = 168,169,…,173
e(B,j) = q(2B+j-174) for j = 174,175
e(B,j) = u’(9B+j-170) for j = 176,177,178
e(B,j) = hi(31B+j-164) for j = 179,180,…,194
e(B,j) = di(306B+j-42) for j = 195,196,…,347
where
q(0),q(1),…,q(7) = 1,1,1,0,0,1,1,1 identifies the coding scheme MCS-7, MCS-8 or MCS-9.
b) Bit swapping
The bit swapping is the same as for MCS-5 DL as specified in subclause 5.1.9.1.6 b).
In RTTI configuration, the bursts with B = 0,2 shall be mapped on the PDCH having the lower timeslot number, whereas the bursts with B = 1,3 shall be mapped on the PDCH having the higher timeslot number, see 3GPP TS 45.002.
c) PAN bit swapping
In case a PAN is included in the radio block, the following additional bits are swapped after the bit swapping in b):
For B = 0
Swap e(B,2) with e(B,1)
Swap e(B,59) with e(B,58)
Swap e(B,110) with e(B,109)
Swap e(B,209) with e(B,208)
Swap e(B,260) with e(B,259)
Swap e(B,317) with e(B,316)
For B = 1
Swap e(B,23) with e(B,22)
Swap e(B,74) with e(B,73)
Swap e(B,131) with e(B,130)
Swap e(B,314) with e(B,313)
Swap e(B,224) with e(B,223)
Swap e(B,281) with e(B,280)
Swap e(B,191) with e(B,205)
For B = 2
Swap e(B,38) with e(B,37)
Swap e(B,95) with e(B,94)
Swap e(B,146) with e(B,141)
Swap e(B,227) with e(B,226)
Swap e(B,278) with e(B,277)
Swap e(B,335) with e(B,334)
Swap e(B,176) with e(B,205)
For B = 3
Swap e(B,2) with e(B,1)
Swap e(B,59) with e(B,58)
Swap e(B,92) with e(B,91)
Swap e(B,149) with e(B,141)
Swap e(B,242) with e(B,241)
Swap e(B,299) with e(B,298)
5.1.11.2 Uplink (MCS-7 UL)
5.1.11.2.1 Block constitution
The message delivered to the encoder has a fixed size of 946 information bits {d(0),d(1),…,d(945)}. It is delivered on a burst mode.
The message delivered to the encoder will have a fixed size of 971 information bits {d(0),d(1),…,d(970)} if a PAN field is included (see 3GPP TS 44.060).
The message delivered to the encoder will have a fixed size of 974 information bits {d(0),d(1),…,d(973)}, if a PAN field and an eTFI field are included (see 3GPP TS 44.060).
NOTE: The presence of the PAN is indicated by the PANI field in the header (see 3GPP TS 44.060).
5.1.11.2.2 Header coding
a) Parity bits:
Eight header parity bits p(0),p(1),…,p(7) are defined in such a way that in GF(2) the binary polynomial:
d(0)D53 +…+ d(45)D8 + p(0)D7 +…+ p(7), when divided by:
D8 + D6 + D3 + 1, yields a remainder equal to:
D7 + D6 + D5 + D4 + D3 + D2 + D+1.
b) Tail biting:
The six last header parity bits are added before information and parity bits, the result being a block of 60 bits {u"(‑6),…,u"(0),u"(1),…,u"(53)} with six negative indexes:
u"(k-6) = p(k+2) for k = 0,1,…,5
u"(k) = d(k) for k = 0,1,…,45
u"(k) = p(k‑46) for k = 46,47,…,53
c) Convolutional encoder
This block of 60 bits {u"(-6),…,u"(0),u"(1),…,u"(53)} is encoded with the 1/3 rate convolutional mother code defined by the polynomials:
G4 = 1 + D2 + D3 + D5 + D6
G7 = 1 + D + D2 + D3 + D6
G5 = 1 + D + D4 + D6
This results in a block of 162 coded bits: {C(0),C(1),…,C(161)} defined by:
C(3k) = u"(k) + u"(k‑2) + u"(k‑3) + u"(k‑5) + u"(k‑6)
C(3k+1) = u"(k) + u"(k‑1) + u"(k‑2) + u"(k‑3) + u"(k‑6)
C(3k+2) = u"(k) + u"(k‑1) + u"(k-4) + u"(k-6) for k = 0,1,…,53
The code is punctured in such a way that the following coded bits:
{C(k) for k = 35,131} are not transmitted
The result is a block of 160 coded bits, {hc(0),hc(1),…,hc(159)}.
5.1.11.2.3 Data coding
The data coding is the same as for downlink as specified in subclause 5.1.11.1.4 where bits {d(40),d(41),…,d(939)} are replaced by bits {d(46),d(47),…,d(945)}.
5.1.11.2.3a Piggy-backed Ack/Nack coding
If a PAN field is included and an eTFI field is not included, the coding of the PAN field is the same as for the downlink as specified in subclause 5.1.11.1.4a where bits {d(940), d(941),…,d(964)} are replaced by bits {d(946),d(947),…,d(970)}.
If a PAN field and an eTFI field are included, the PAN coding is the same as for the downlink as specified in sub-clause 5.1.11.1.4a, with the exception that the five bits {d(966),d(967),d(968),d(969),d(970)} (TFI value or 00000, see 3GPP TS 44.060) and the three bits {d(971),d(972),d(973)} (eTFI value or 000, see 3GPP TS 44.060) are added bit-wise modulo 2 to the PAN parity bits {p(2),p(3),p(4),p(5),p(6),p(7),p(8),p(9)} resulting in the ten modified PAN parity bits defined as:
pt(k) = p(k) for k=0,…,1
pt(k) = d(k+969) + p(k) for k=2,…,4
pt(k) = d(k+961) + p(k) for k=5,…,9
The data coded bits {pc1(0),pc1(1),…,pc1(575)} and {pc2(0),pc2(1),…,pc2(575)} are appended to the PAN coded bits as described for the downlink in subclause 5.1.11.1.4a. The result is two blocks of 612 coded bits {c1(0),c1(1),…,c1(611)} and {c2(0),c2(1),…,c2(611)}.
5.1.11.2.4 Interleaving
a) Header
The 160 coded bits of the header, {hc(0),hc(1),…,hc(159)}, are interleaved according to the following rule:
hi(j) = hc(k) for k = 0,1,…,159
j = 40(k mod 4) + 2((13(k div 8)) mod 20) + ((k mod 8) div 4)
b) Data
The data interleaving is the same as for MCS-7 DL as specified in subclause 5.1.11.1.5.
5.1.11.2.5 Mapping on a burst
a) Straightforward Mapping
The mapping is given by the rule:
For B=0,1,2,3, let
e(B,j) = di(306B+j) for j = 0,1,…,152
e(B,j) = hi(40B+j-153) for j = 153,154,…,173
e(B,j) = q(2B+j-174) for j = 174,175
e(B,j) = hi(40B+j-155) for j = 176,177,…,194
e(B,j) = di(306B+j-42) for j = 195,196,…,347
where
q(0),q(1),…,q(7) = 1,1,1,0,0,1,1,1 identifies the coding scheme MCS-7, MCS-8 or MCS-9.
b) Bit swapping
The bit swapping is the same as for MCS-5 DL as specified in subclause 5.1.9.1.6 b).
In RTTI configuration, the bursts with B = 0,2 shall be mapped on the PDCH having the lower timeslot number, whereas the bursts with B = 1,3 shall be mapped on the PDCH having the higher timeslot number, see 3GPP TS 45.002.
c) PAN bit swapping
In case a PAN is included in the radio block, additional bits are swapped as specified in subclause 5.1.11.1.6 c).
5.1.12 Packet data block type 12 (MCS-8)
5.1.12.1 Downlink (MCS-8 DL)
5.1.12.1.1 Block constitution
The message delivered to the encoder has a fixed size of 1132 information bits {d(0),d(1),…,d(1131)}. It is delivered on a burst mode.
The message delivered to the encoder will have a fixed size of 1135 information bits {d(0),d(1),…,d(1134)}, if an eTFI field is included (see 3GPP TS 44.060).
The message delivered to the encoder will have a fixed size of 1157 information bits {d(0),d(1),…,d(1156)} if a PAN field is included.
The message delivered to the encoder will have a fixed size of 1160 information bits {d(0),d(1),…,d(1159)}, if a PAN field and an eTFI field are included (see 3GPP TS 44.060).
NOTE: The presence of the PAN is indicated by the PANI field in the header (see 3GPP TS 44.060).
5.1.12.1.2 USF precoding
5.1.12.1.2.1 BTTI configuration
A block of 36 bits {u’(0),u’(1),…,u’(35)} is derived from {d(0),d(1),d(2)} as described for MCS-5 DL in subclause 5.1.9.1.2.1.
5.1.12.1.2.2 RTTI configuration
A block of 36 bits {u’(0),u’(1),…,u’(35)} is generated as described for MCS-5 DL in subclause 5.1.9.1.2.2.
5.1.12.1.3 Header coding
A block of 124 coded bits {hc(0),hc(1),…,hc(123)} is derived from {d(3),d(4),…,d(39)} as described for MCS-7 DL in subclause 5.1.11.1.3, with the exception that eTFI field bits are now located in {d(1132),d(1133),d(1134)} if included and PAN is not included, and located in {d(1157),d(1158),d(1159)} if included in combination with a PAN field.
5.1.12.1.4 Data coding
I) First half:
a) Parity bits:
Twelve data parity bits p(0),p(1),…,p(11) are defined in such a way that in GF(2) the binary polynomial:
d(40)D557 +…+ d(585)D12 + p(0)D11 +…+ p(11), when divided by:
D12 + D11 + D10 + D8 + D5 + D4 + 1, yields a remainder equal to:
D11 + D10 + D9 + D8 + D7 + D6 + D5 + D4 + D3 + D2 + D + 1.
b) Tail bits:
Six tail bits equal to 0 are added to the information and parity bits, the result being a block of 564 bits {u(0),u(1),…,u(563)}:
u(k) = d(k+40) for k = 0,1,…,545
u(k) = p(k‑546) for k = 546,547,…,557
u(k) = 0 for k = 558,559,…,563 (tail bits)
c) Convolutional encoder
This block of 564 bits {u(0),u(1),…,u(563)} is encoded with the 1/3 rate convolutional mother code defined by the polynomials:
G4 = 1 + D2 + D3 + D5 + D6
G7 = 1 + D + D2 + D3 + D6
G5 = 1 + D + D4 + D6
This results in a block of 1692 coded bits: {C(0),C(1),…,C(1691)} defined by:
C(3k) = u(k) + u(k‑2) + u(k‑3) + u(k‑5) + u(k‑6)
C(3k+1) = u(k) + u(k‑1) + u(k‑2) + u(k‑3) + u(k‑6)
C(3k+2) = u(k) + u(k‑1) + u(k-4) + u(k-6) for k = 0,1,…,563; u(k) = 0 for k < 0
The code is punctured depending on the value of the CPS field as defined in 3GPP TS 44.060. Three puncturing schemes named P1, P2 or P3 are applied in such a way that the following coded bits:
|
P1 |
{C(36j), C(2+36j), C(5+36j), C(6+36j), C(10+36j), C(13+36j), C(16+36j), C(20+36j), C(23+36j), C(24+36j), C(27+36j), C(31+36j), C(35+36j), for j = 0,1,…,46} as well as {C(845)} are transmitted |
|
P2 |
{C(1+36j), C(4+36j), C(8+36j), C(11+36j), C(12+36j), C(15+36j), C(17+36j), C(19+36j), C(22+36j), C(25+36j), C(28+36j), C(30+36j), C(33+36j), for j = 0,1,…,46} as well as {C(582)} are transmitted |
|
P3 |
{C(2+36j), C(3+36j), C(7+36j), C(9+36j), C(14+36j), C(17+36j), C(18+36j), C(21+36j), C(26+36j), C(27+36j), C(29+36j), C(32+36j), C(34+36j), for j = 0,1,…,46} as well as {C(1156)} are transmitted |
The result is a block of 612 coded bits, {c1(0),c1(1),…,c1(611)}.
For the FANR procedure, the code is punctured depending on the CPS field and the PANI field as defined in 3GPP TS 44.060. If the PANI field is set to 0, the puncturing is the same as for EGPRS. If the PANI field is set to 1, the puncturing schemes named P1 or P2 are applied in such a way that, in addition to the bits punctured for EGPRS, the following coded bits:
|
P1 |
{C(2+36j) for j = 0,1,…,46} are not transmitted except {C(k) for k = 38,182,326,470,614,758,902,1046,1190,1334,1478} which are transmitted |
|
P2 |
{C(17+36j) for j = 0,1,…,46} are not transmitted except {C(k) for k = 89,233,377,521,665,809,953,1097,1241,1385,1529} which are transmitted |
|
P3 |
{C(27+36j) for j = 0,1,…,46} are not transmitted except {C(k) for k = 135,279,423,567,711,855,999,1143,1287,1431,1575} which are transmitted |
The result is a block of 576 coded bits {pc1(0),pc1(1),…,pc1(575)}.
II) Second half:
The same data coding as for first half is proceeded with bits {d(40),d(41),…,d(585)} replaced by bits {d(586),d(587),…,d(1131)}. The result is a block of 612 coded bits, {c2(0),c2(1),…,c2(611)}.
If the PANI field is set to 1, additional bits are punctured as for the first half. The result is a block of 576 coded bits {pc2(0),pc2(1),…,pc2(575)}.
5.1.12.1.4a Piggy-backed Ack/Nack coding
The operations in this subclause shall be carried out only if a PAN field is included.
a) Parity bits
Ten PAN parity bits p(0), p(1),…,p(9) are defined in such a way that in GF(2) the binary polynomial:
d(1132)D29 +…+ d(1151)D10 + p(0)D9 +…+ p(9), when divided by:
D10 + D9 + D5 + D4 + D + 1, yields a remainder equal to:
D9 + D8 + D7 + D6 + D5 + D4 + D3 + D2 + D1 + 1.
The five bits {d(1152),…,d(1156)} (TFI value or 00000, see 3GPP TS 44.060) are added bit-wise modulo 2 to the 5 last parity bits {p(5),…,p(9)}. This results in the ten modified PAN parity bits {pt(0),…,pt(9)} defined as:
pt(k) = p(k) for k=0,…,4
pt(k) = d(k+1147) + p(k) for k=5,…,9
b) Tail biting:
The six last modified PAN parity bits are added before information and modified PAN parity bits, the result being a block of 36 {u’’(-6),…,u’’(0),u’’(1),…,u’’(29)} bits with six negative indexes:
u’’(k-6) = pt(k+4) for k = 0,1,…,5
u’’(k) = d(k+1132) for k = 0,1,…,19
u’’(k) = pt(k-20) for k = 20,21,…,29
c) Convolutional encoder
The block of 36 bits {u’’(-6),…,u’’(0),u’’(1),…,u’’(29)} is encoded with the 1/3 rate convolutional mother code defined by the polynomials:
G4 = 1 + D2 + D3 + D5 + D6
G7 = 1 + D + D2 + D3 + D6
G5 = 1 + D + D4 + D6
This results in a block of 90 coded bits {C(0),C(1),…,C(89)} defined by:
C(3k) = u"(k) + u"(k-2) + u"(k-3) + u"(k-5) + u"(k-6)
C(3k+1) = u"(k) + u"(k-1) + u"(k-2) + u"(k-3) + u"(k-6)
C(3k+2) = u"(k) + u"(k-1) + u"(k-4) + u"(k-6) for k = 0,1,…,29
The block of 90 coded bits is punctured in such way that the following coded bits:
{ C(2+15k), C(8+15k), C(14+15k) for k = 0,1,…,5} are not transmitted.
The result is a block of 72 coded bits {ac(0),ac(1),…,ac(71)}.
The data coded bits {pc1(0),pc1(1),…,pc1(575)} and {pc2(0),pc2(1),…,pc2(575)} are appended to the PAN coded bits by the following rule:
c1(k) = ac(2k) for k = 0,1,…,35
c1(k) = pc1(k-36) for k = 36,37,…,611
c2(k) = ac(2k+1) for k = 0,1,…,35
c2(k) = pc2(k-36) for k = 36,37,…,611
The result is two blocks of 612 coded bits {c1(0),c1(1),…,c1(611)} and {c2(0),c2(1),…,c2(611)}.
5.1.12.1.5 Interleaving
a) Header
The header interleaving is the same as for MCS-7 DL as specified in subclause 5.1.11.1.5.
b) Data
Data are put together as one entity as described by the following rule:
dc(k) = c1(k) for k = 0,1,…,611
dc(k) = c2(k-612) for k = 612,613,…,1223
The resulting block is interleaved according to the following rule:
di(j) = dc(k) for k = 0,1,…,1223
j = 306(2(k div 612) + (k mod 2)) + 3((74k) mod 102 + (k div 2) mod 2) + (k + 2 – (k div 204)) mod 3
5.1.12.1.6 Mapping on a burst
a) Straightforward Mapping
The mapping is the same as for MCS-7 DL as specified in subclause 5.1.11.1.6 a).
b) Bit swapping
The bit swapping is the same as for MCS-7 DL as specified in subclause 5.1.11.1.6 b).
c) PAN bit swapping
In case a PAN is included in the radio block, the following additional bits are swapped after the bit swapping in b):
For B = 0,2
Swap e(B,2) with e(B,1)
Swap e(B,23) with e(B,22)
Swap e(B,113) with e(B,112)
Swap e(B,128) with e(B,127)
Swap e(B,155) with e(B,141)
Swap e(B,185) with e(B,205)
Swap e(B,260) with e(B,259)
Swap e(B,281) with e(B,280)
For B = 1,3
Swap e(B,59) with e(B,58)
Swap e(B,74) with e(B,73)
Swap e(B,176) with e(B,207)
Swap e(B,206) with e(B,205)
Swap e(B,227) with e(B,226)
Swap e(B,317) with e(B,316)
Swap e(B,332) with e(B,331)
5.1.12.2 Uplink (MCS-8 UL)
5.1.12.2.1 Block constitution
The message delivered to the encoder has a fixed size of 1138 information bits {d(0),d(1),…,d(1137)}. It is delivered on a burst mode.
The message delivered to the encoder will have a fixed size of 1163 information bits {d(0),d(1),…,d(1162)} if a PAN field is included (see 3GPP TS 44.060).
The message delivered to the encoder will have a fixed size of 1166 information bits {d(0),d(1),…,d(1165)}, if a PAN field and an eTFI field are included (see 3GPP TS 44.060).
NOTE: The presence of the PAN is indicated by the PANI field in the header (see 3GPP TS 44.060).
5.1.12.2.2 Header coding
A block of 160 coded bits {hc(0),hc(1),…,hc(159)} is derived from {d(0),d(1),…,d(45)} as described for MCS-7 UL in subclause 5.1.11.2.2.
5.1.12.2.3 Data coding
The data coding is the same as for downlink as specified in subclause 5.1.12.1.4 where bits {d(40),d(41),…,d(1131)} are replaced by bits {d(46),d(47),…,d(1137)}.
5.1.12.2.3a Piggy-backed Ack/Nack coding
If a PAN field is included and an eTFI field is not included, the coding of the PAN field is the same as for the MCS-8 DL as specified in subclause 5.1.12.1.4a where bits {d(1132), d(1133),…,d(1156)} are replaced by bits {d(1138),d(1139),…,d(1162)}.
If a PAN field and an eTFI field are included, the PAN coding is the same as for the downlink as specified in sub-clause 5.1.12.1.4a, with the exception that the five bits {d(1158),d(1159),d(1160),d(1161),d(1162)} (TFI value or 00000, see 3GPP TS 44.060) and the three bits {d(1163),d(1164),d(1165)} (eTFI value or 000, see 3GPP TS 44.060) are added bit-wise modulo 2 to the PAN parity bits {p(2),p(3),p(4),p(5),p(6),p(7),p(8),p(9)} resulting in the ten modified PAN parity bits defined as:
pt(k) = p(k) for k=0,…,1
pt(k) = d(k+1161) + p(k) for k=2,…,4
pt(k) = d(k+1153) + p(k) for k=5,…,9
The data coded bits {pc1(0),pc1(1),…,pc1(575)} and {pc2(0),pc2(1),…,pc2(575)} are appended to the PAN coded bits as described for MCS-8 DL in subclause 5.1.12.1.4a. The result is two blocks of 612 coded bits {c1(0),c1(1),…,c1(611)} and {c2(0),c2(1),…,c2(611)}.
5.1.12.2.4 Interleaving
a) Header
The header interleaving is the same as for MCS-7 UL as specified in subclause 5.1.11.2.4.
b) Data
The data interleaving is the same as for MCS-8 DL as specified in subclause 5.1.12.1.5.
5.1.12.2.5 Mapping on a burst
a) Straightforward mapping
The mapping is the same as for MCS-7 UL as specified in subclause 5.1.11.2.5 a).
b) Bit swapping
The bit swapping is the same as for MCS-7 UL as specified in subclause 5.1.11.2.5 b).
c) PAN bit swapping
In case a PAN is included in the radio block, additional bits are swapped as specified in subclause 5.1.12.1.6 c).
5.1.13 Packet data block type 13 (MCS-9)
5.1.13.1 Downlink (MCS-9 DL)
5.1.13.1.1 Block constitution
The message delivered to the encoder has a fixed size of 1228 information bits {d(0),d(1),…,d(1227)}. It is delivered on a burst mode.
The message delivered to the encoder will have a fixed size of 1231 information bits {d(0),d(1),…,d(1230)}, if an eTFI field is included (see 3GPP TS 44.060).
5.1.13.1.2 USF precoding
5.1.13.1.2.1 BTTI configuration
A block of 36 bits {u’(0),u’(1),…,u’(35)} is derived from {d(0),d(1),d(2)} as described for MCS-5 DL in subclause 5.1.9.1.2.1.
5.1.13.1.2.2 RTTI configuration
A block of 36 bits {u’(0),u’(1),…,u’(35)} is generated as described for MCS-5 DL in subclause 5.1.9.1.2.2.
5.1.13.1.3 Header coding
A block of 124 coded bits {hc(0),hc(1),…,hc(123)} is derived from {d(3),d(4),…,d(39)} as described for MCS-7 DL in subclause 5.1.11.1.3, with the exception that eTFI field bits are now located in {d(1228),d(1229),d(1230)} if included.
5.1.13.1.4 Data coding
I) First half:
a) Parity bits:
Twelve data parity bits p(0),p(1),…,p(11) are defined in such a way that in GF(2) the binary polynomial:
d(40)D605 +…+ d(633)D12 + p(0)D11 +…+ p(11), when divided by:
D12 + D11 + D10 + D8 + D5 + D4 + 1, yields a remainder equal to:
D11 + D10 + D9 + D8 + D7 + D6 + D5 + D4 + D3 + D2 + D + 1.
b) Tail bits:
Six tail bits equal to 0 are added to the information and parity bits, the result being a block of 612 bits {u(0),u(1),…,u(611)}:
u(k) = d(k+40) for k = 0,1,…,593
u(k) = p(k‑594) for k = 594,595,…,605
u(k) = 0 for k = 606,607,…,611 (tail bits)
c) Convolutional encoder
This block of 612 bits {u(0),u(1),…,u(611)} is encoded with the 1/3 rate convolutional mother code defined by the polynomials:
G4 = 1 + D2 + D3 + D5 + D6
G7 = 1 + D + D2 + D3 + D6
G5 = 1 + D + D4 + D6
This results in a block of 1836 coded bits: {C(0),C(1),…,C(1835)} defined by:
C(3k) = u(k) + u(k‑2) + u(k‑3) + u(k‑5) + u(k‑6)
C(3k+1) = u(k) + u(k‑1) + u(k‑2) + u(k‑3) + u(k‑6)
C(3k+2) = u(k) + u(k‑1) + u(k-4) + u(k-6) for k = 0,1,…,611; u(k) = 0 for k < 0
The code is punctured depending on the value of the CPS field as defined in 3GPP TS 44.060. Three puncturing schemes named P1, P2 or P3 are applied in such a way that the following coded bits:
|
P1 |
{C(3j) for j = 0,1,…,611} are transmitted |
|
P2 |
{C(1+3j) for j = 0,1,…,611} are transmitted |
|
P3 |
{C(2+3j) for j = 0,1,…,611} are transmitted |
The result is a block of 612 coded bits, {c1(0),c1(1),…,c1(611)}.
II) Second half:
The same data coding as for first half is proceeded with bits {d(40),d(41),…,d(633)} replaced by bits {d(634),d(635),…,d(1227)}. The result is a block of 612 coded bits, {c2(0),c2(1),…,c2(611)}.
5.1.13.1.5 Interleaving
The interleaving is the same as for MCS-8 DL as specified in subclause 5.1.12.1.5.
5.1.13.1.6 Mapping on a burst
The mapping is the same as for MCS-7 DL as specified in subclause 5.1.11.1.6.
5.1.13.2 Uplink (MCS-9 UL)
5.1.13.2.1 Block constitution
The message delivered to the encoder has a fixed size of 1234 information bits {d(0),d(1),…,d(1233)}. It is delivered on a burst mode.
5.1.13.2.2 Header coding
A block of 160 coded bits {hc(0),hc(1),…,hc(159)} is derived from {d(0),d(1),…,d(45)} as described for MCS-7 UL in subclause 5.1.11.2.2.
5.1.13.2.3 Data coding
The data coding is the same as for downlink as specified in subclause 5.1.13.1.4 where bits {d(40),d(41),…,d(1227)} are replaced by bits {d(46),d(47),…,d(1233)}.
5.1.13.2.4 Interleaving
The interleaving is the same as for MCS-8 UL as specified in subclause 5.1.12.2.4.
5.1.13.2.5 Mapping on a burst
The mapping is the same as for MCS-7 UL as specified in subclause 5.1.11.2.5.