5.1b Extended Coverage Packet data traffic channel (EC-PDTCH)

3GPP45.003GSM/EDGE Channel codingRelease 17TS

Tools: ARFCN - Frequency Conversion for 5G NR/LTE/UMTS/GSM

5.1b.1 General

Twelve coding schemes are specified for the Extended Coverage Packet data traffic channels. The coding schemes MCS-1 to MCS-9 are identical to those defined for PDTCH in subclause 5.1 and apply for downlink and uplink coverage class CC1. The coding schemes MCS-1/16, MCS-1/8 and MCS-1/4 for downlink and uplink coverage classes CC4, CC3 and CC2, respectively, are based on MCS-1 defined in subclause 5.1. The coding scheme MCS-1’/48 for uplink coverage class CC5 is specified in subclause 5.1b.4.

The following additions apply:

– Blind physical layer transmissions are employed.

– For downlink, multiple sequences of USF bits are delivered to the encoder. Each precoded USF sequence is mapped onto one instance of the repeated block.

The following restrictions apply:

– Reduced Transmission Time Interval (RTTI) shall not be used on EC-PDTCH.

– A PAN field shall not be included in the message delivered to the encoder.

– For data coding, a reduced set of puncturing schemes is used on an EC-PDTCH, compared to a PDTCH. Only the puncturing schemes listed in table 0 shall be used.

Table 0: Allowed puncturing schemes on EC-PDTCH

MCS	Allowed puncturing schemes in subclause 5.1
MCS-1/16 MCS-1/8 MCS-1/4 MCS-1 MCS-2	P1
MCS-3 MCS-4	P1 or P2
Other MCSs	No restriction

5.1b.2 Downlink MCS-1/M

5.1b.2.1 Block constitution

The message delivered to the encoder has a fixed size of 206 information bits {a(0),a(1),…,a(205)}. In addition, M sequences of USF bits {u(m,0),u(m,1),u(m,2)}, m=0,…,M-1, are delivered to the encoder, where M=16, 8 and 4 for MCS-1/16, MCS-1/8 and MCS-1/4, respectively.

5.1b.2.2 Encoding and blind physical layer transmission

For each of the M blind physical layer transmissions, the block is encoded as specified for MCS-1 in subclause 5.1.5.1. The message {d(0),d(1),…,d(208)} delivered to the encoder (see subclause 5.1.5.1.1) for blind physical layer transmission k is defined as

d(k)=u(m,k) for k=0,1,2;

d(k)=a(k-3) for k=3,…,208.

Denote the resulting bursts from encoding of blind physical layer transmissions m as R(m,B,j) = e(B,j) for m=0,…,M-1, B=0,…,3 and j=0,…,115, where e(B,j) is defined in subclause 5.1.5.1.6.1.

5.1b.2.3 Mapping onto PDCHs

Burst R(m,B,j), j=0,…,115, is mapped onto burst B’ of PDCH number pn, where

– for EC channels with 4 PDCHs assigned

B’ = B + 4(m div 4);

pn = (m mod 4).

– for EC channels with 2 PDCHs assigned

B’ = B + 4(m div 2);

pn = (m mod 2).

NOTE 1: The burst number B’ denotes the relative transmission order of the bursts on a PDCH. The mapping to absolute TDMA frame number is specified in 3GPP TS 45.002 [8] and 3GPP TS 44.060 [6].

NOTE 2: pn denotes the relative PDCH number within the set of PDCHs on which the EC-PDTCH is mapped, where PDCHs are numbered from lower to higher timeslot number. The absolute timeslot number depends on the assignment, see 3GPP TS 44.060 [6].

5.1b.3 Uplink MCS-1/M

5.1b.3.1 Block constitution

The message delivered to the encoder has a fixed size of 209 information bits {a(0),a(1),…,a(208)}.

5.1b.3.2 Encoding and blind physical layer transmission

For each of the M blind physical layer transmissions the block is encoded as specified for MCS-1 in subclause 5.1.5.2. The message {d(0),d(1),…,d(208)} delivered to the encoder (see subclause 5.1.5.1.1) for blind physical layer transmission m is defined as

d(k)=a(k) for k=0,…,208.

Denote the resulting bursts from encoding of blind physical layer transmission m as R(m,B,j) = e(B,j) for m=0,…,M-1, B=0,…,3 and j=0,…,115, where e(B,j) is defined in subclause 5.1.5.1.6.1.

5.1b.3.3 Mapping onto PDCHs

The mapping is done as specified for MCS-1/M DL in subclause 5.1b.2.3, except when 2 PDCHs are assigned. In this case, burst R(m,B,j), j=0,…,115, is mapped onto burst B’ of PDCH number pn, where:

B’ = B(M div 2) + (m div 2);

pn = (m mod 2).

5.1b.4 Uplink MCS-1’/48

5.1b.4.1 Block constitution

The message delivered to the encoder has a fixed size of 194 information bits {d(0),d(1),…,d(193)}. It is delivered on a burst mode.

5.1b.4.2 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)D23 +…+ d(15)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 30 bits {u"(‑6),…,u"(0),u"(1),…,u"(23)} with six negative indexes:

u"(k-6) = p(k+2) for k = 0,1,…,5

u"(k) = d(k) for k = 0,1,…,15

u"(k) = p(k‑16) for k = 16,17,…,23

c) Convolutional encoder

This block of 30 bits {u"(-6),…,u"(0),u"(1),…,u"(23)} 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 72 coded bits: {C(0),C(1),…,C(71)} 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,…,23

The code is punctured in such a way that the following coded bits:

C(2+12j), C(5+12j), C(8+12j), C(11+12j) for j = 0,1,…,5} are not transmitted

The result is a block of 48 coded bits, {hc(0),hc(1),…,hc(47)}.

5.1b.4.3 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(16)D189 +…+ d(193)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+16) 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 using always P1 as defined below.

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

The result is a block of 372 coded bits, {dc(0),dc(1),…,dc(371)}.

5.1b.4.4 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,…,47

c(k) = dc(k‑48) ) for k = 48,49,…,419

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,…,400

c’(n,k) = c(n,k-4) for k = 402,403,…,423

c’(n,25) = q(10) c’(n,82) = q(11) c’(n,139) = q(12) c’(n,401) = q(13)

c(n,k) are the coded bits and q(10),q(11),q(12),q(13) = 0,0,0,0 are four extra stealing flags

The resulting block of 424 bits is interleaved according to the following rule:

i(B,j) = c’(n,k) for k = 0,1,…,423

n = 0,1,…,N,N+1,…

B = B0 + 4n + (k mod 4)

j = 2((49k) mod 53) + ((k mod 8) div 4)

5.1b.4.5 Mapping on a burst

The mapping is given by the rule:

e(B,j) = i(B,j) and e(B,63+j) = i(B,53+j) for j = 0,1,…,52

and

e(B+m,53+i) = q(i) and e(B+58+i) =q(i+5) for m = 0,1,2,3 and i = 0,1,2,…,4

where

q(0),q(1),..,q(9) = 0,0,0,0,0,0,0,0,0,0.

5.1b.4.6 Blind physical layer transmission

The resulting bursts from encoding of blind physical layer transmission m are denoted as R(m,B,j) = e(B,j) for m=0,…,47, B=0,…,3 and j=0,…,115, where e(B,j) is defined in subclause 5.1b.4.5.

5.1b.4.7 Mapping onto PDCHs

Burst R(m,B,j), j=0,…,115, is mapped onto burst B’ of PDCH number pn, where

– for EC channels with 4 PDCHs assigned

B’ = B + 4*(m div 4);

pn = (m mod 4).

– for EC channels with 2 PDCHs assignedB’ = B(M div 2) + (m div 2);

pn = (m mod 2).

NOTE 1: The burst number B’ denotes the relative transmission order of the bursts on a PDCH. The mapping to absolute TDMA frame number is specified in 3GPP TS 45.002 [8] and 3GPP TS 44.060 [6].