Page 1265 - 5G Basics - Core Network Aspects

Transport aspects 2

Figure D.6 – JC1, JC2 and JC3 generation

A parallel logic implementation of the source CRC-8 and CRC-6 are illustrated in Appendix VI.

The GMP sink synchronizes its Cm(t) value to the GMP source through the following process, which is
illustrated in Figure D.7 and the equivalent process in Figure D.8.

When the received JC octets contain II = DI and a valid CRC-8, the GMP sink accepts the received C1-CLas its
Cm(t) value for the next frame. At this point the GMP sink is synchronized to the GMP source. When II ≠ DI
with a valid CRC in the current received frame (frame i), the GMP sink must examine the received JC octets
in the next frame (frame i+1) in order to obtain Cm(t) synchronization. II ≠ DI in frame i indicates that the
source is performing a count increment or decrement operation that will modify the Cm(t) value it sends in
frame i+1. Since this modification to the Cm(t) will affect the count LSBs, the GMP sink uses the LSBs, II, and
DI in frame i to determine its synchronization hunt state when it receives frame i+1. Specifically, in Figure
D.7, the Hunt state (A-F) is determined using C13, C14, II and DI for the 14-bit count and C9, C10, II and DI
for the 10-bit count. Equivalently, in Figure D.8, the Hunt state (A or B) is determined using C14, II, and DI
for the 14-bit count and C10, II, and DI for the 10-bit count. If II = DI with a valid CRC in frame i+1, Cm(t)
synchronization is achieved by directly accepting the received C1-CL as the new Cm(t). If II ≠ DI with a valid
CRC in frame i+1, the sink uses the new LSBs, II and DI values to determine whether the source is
communicating an increment or decrement operation and the magnitude of the increment/decrement

1255