Chapter 9: Upstream Cable Access Modules (UCAMs)
STANDARD Revision 1.0
C4® CMTS Release 8.3 User Guide
© 2016 ARRIS Enterprises LLC. All Rights Reserved.
314
Figure 67:
Relation of FEC Codewords to Data Packet
A slight decrease in SNR can cause a large increase in the Packet Error Rate (PER). There comes a point where simply
adding additional FEC to attempt to correct for the upstream errors is no longer efficient. Once this point is reached, it is
more efficient to use a lower modulation rate with less FEC overhead than to continue to increase the FEC protection.
The figure below, obtained using the ICO tool, illustrates that it can be more efficient to use a lower FEC setting and a
lesser modulation rate. Ideally the channel will operate as far towards the upper left of this chart as possible. The Shannon
curve displays what is theoretically possible. The curves for the different modulation rates show the effect of increasing
the value of T for the FEC for each of the different modulation rates for the specified modulation profile parameters.
Because the lab chassis used to obtain the data in the figure below was provisioned with a very short cable plant, these
numbers approach the theoretical values listed above for SNR. They do not reflect real-world noise levels. In the 64 QAM
modulation profile, if the SNR of the plant is about 20 dB, then a high T value is required to limit the Packet Error Rate
(PER) to 0.1%. The highest bit rate one can achieve is about 8 Mbps. However, if we reduce the modulation to 32 QAM and
decrease the FEC T value, the bit rate goes up to 12 Mbps under the same conditions.
For practical reasons the minimum value of k must be at least 16. Because of this, if there are not 16 data bytes either to
be sent or remaining to be sent, then the modem must pad out this data to be 16 bytes. This especially comes into play for
IUC 1 where the request frame is 6 bytes. Therefore, to have even the minimal values for FEC of T = 1 and k = 16 for IUC 1,
