Page 1276 - 5G Basics - Core Network Aspects
P. 1276
2 Transport aspects
Annex F
Improved robustness for mapping of 40GBASE-R
into OPU3 using 1027B code blocks
(This annex forms an integral part of this Recommendation.)
F.1 Introduction
When a parallel 40GBASE-R signal is transcoded as per Annexes B and E and directly mapped into OPU3
without GFP framing, another method is needed to locate the start of 513B blocks and to provide
protection to prevent bit errors creating an unacceptable increase in mean time to false packet acceptance
(MTTFPA).
F.2 513B code block framing and flag bit protection
The mapping of 513B code blocks into OPU3 requires a mechanism for locating the start of the code blocks.
A mechanism is also needed to protect the flag bit, whose corruption could cause data to be erroneously
interpreted as control and vice versa.
Both of these requirements can be addressed by providing parity across the flag bits of two 513B blocks
produced from the transcoding of Annex B.
Figure F.1 illustrates the flag parity bit across two 513B blocks. This creates a minimum two-bit Hamming
distance between valid combinations of flag bits.
F 512 F 512
512 bit payload (8 66b block)
Flag parity bit 1,024 bit payload (2 8 66b block)
P F F 512 512
1 0 0 Data Data
0 0 1 Data Control
0 1 0 Control Data
1 1 1 Control Control
Scramble G.709-Y.1331(12)_FF.1
Figure F.1 – Flag parity bit on two 513B blocks (1027B code)
The flag parity bit creates a sequence that can be used for framing to locate the 513B blocks in a stream of
bits. The state diagram of Figure 49-12 of [IEEE 802.3] is applied to locate a 3-bit pattern appearing once
per 1027 bits (rather than a 2-bit pattern appearing once per 66 bits) where four out of eight 3-bit
sequences (rather than two out of four two-bit values as used in IEEE 802.3) match the pattern. The
additional step required is to scramble the non-flag bits so that the legal sequences of these bits are not
systematically mimicked in the data itself. The scrambler to be used for this purpose is the Ethernet self-
58
39
synchronous scrambler using the polynomial G(x) = 1 + x + x .
1266