Page 1184 - 5G Basics - Core Network Aspects

2 Transport aspects

Table 17-12 – m, n and CnD for CBR clients into OPU4

Bit-rate
Nominal bit rate
Client signal tolerance m n CnD
(kbit/s)
(ppm)
100GBASE-R (see 103 125 000 100 640 8 Yes
17.7.5.1)

Table 17-13 – Replacement signal for CBR clients

Client signal Replacement signal Bit-rate tolerance (ppm)

100GBASE-R Continuous 100GBASE-R local fault sequence ordered 100
(see 17.7.5.1) sets with 20 PCS lane alignment markers inserted after
each 16383 x 20 sixty-six-bit blocks

A 100GBASE-R local fault sequence ordered set is a 66B control block (sync header = 10) with a block type
of 0x4B, an "O" code of 0x00, a value of 0x01 to indicate a "local fault" in lane 3, and all of the other octets
(before scrambling) equal to 0x00.

17.7.5.1 100GBASE-R multi-lane processing

The 100GBASE-R client signal (64B/66B encoded, nominal aggregate bit-rate of 103 125 000 kbit/s 100
ppm) is recovered using the process described in Annex E for parallel 64B/66B interfaces. The lane(s) of the
physical interface are bit-disinterleaved, if necessary, into twenty streams of 5 161 250 kbit/s. 66B block
lock and lane alignment marker lock are acquired on each PCS lane, allowing the 66B blocks to be de-
skewed and reordered.

In the mapper, the received Ethernet PCS lane BIP may be compared with the expected Ethernet PCS lane
BIP as a non-intrusive monitor.

The de-mapper will pass through the PCS lane BIP from the ingress as described in Annex E. In addition, the
received Ethernet PCS lane BIP may be compared with the expected Ethernet PCS lane BIP as a non-
intrusive monitor.

For 100GBASE-R client mapping, 1-bit timing information (C1) is not needed.
The de-mapper will recover from the output of the GMP processor 64B/66B block lock per the state
diagram in Figure 82-10 [IEEE 802.3]. The 66B blocks are re-distributed round-robin to PCS lanes. If the
number of PCS lanes is greater than the number of physical lanes of the egress interface, the appropriate
numbers of PCS lanes are bit-multiplexed onto the physical lanes of the egress interface.

17.8 Mapping a 1000BASE-X and FC-1200 signal via timing transparent transcoding into OPUk

17.8.1 Mapping a 1000BASE-X signal into OPU0
Refer to clause 17.7.1 for the mapping of the transcoded 1000BASE-X signal and to clause 17.7.1.1 for the
transcoding of the 1000BASE-X signal.

17.8.2 Mapping an FC-1200 signal into OPU2e
The nominal line rate for FC-1200 is 10 518 750 kbit/s ± 100 ppm, and must therefore be compressed to a
suitable rate to fit into an OPU2e.
The adaptation of the 64B/66B encoded FC-1200 client is done by transcoding a group of eight 66B blocks
into one 513B block (as described in Annex B), assembling eight 513B blocks into one 516-octet superblock
and encapsulating seventeen 516-octet superblocks into an 8800 octet GFP frame as illustrated in Figure
17-17. The GFP frame consists of 2200 rows with 32 bits per row. The first row contains the GFP core
header, the second row the GFP payload header. The next four rows contain 16 bytes reserved for future
international standardization. The next seventeen times 129 rows contain the seventeen superblocks #1 to

1174