Mocomtech CDM-570 - Manual (Page 283)

CDM-570/570L Satellite Modem with Optional IP Module Revision 4
IP Module Ethernet Interface Overview MN/CDM570L.IOM
Header Compression Configuration – Header Compression is completely independent
from QoS, and there is no configuration required except enabling the Header
Compression feature on both the sending and receiving Comtech EF Data IP modem.
Packets with a Header Compression supported header will automatically be identified for
compression. The only configuration consideration is the Header Compression Refresh
Rate. This is how many compressed header packets will be sent before a single full
header packet is sent. Some compressed header traffic could be lost during deteriorated
satellite link conditions. Sending a full header packet will allow the return of the traffic
stream. The Refresh Rate can be increased for poor satellite link conditions or decreased
to further reduce overhead.
Header Compression Statistics - These statistics will display the total bytes of the pre-
compressed and post-compressed traffic and effective compression ratio.
15.4.3 Payload Compression
Traffic optimization through Payload Compression is another optional feature of the
CDM-570/570L /IP Module modem.
• FAST feature to upgrade
• Uses AHA chip
• Compression algorithm applied to all data (HDLC header excluded).
• Compression statistics are fed back to QoS in order to maximize WAN utilization
while maintaining priority, jitter and latency.
• 1024 simultaneous compression sessions to maximize compression across multiple
distinct traffic flows.
• Compression algorithm not applied to RTP streams because this traffic is already
compressed and would only INCREASE the sat bandwidth if compressed again.
• Additional statistics have been added to the compression statistics menu in order to
provide feedback on the compression efficiency that has been achieved.
• Payload Compression is selectable on a per route basis.
15.4.3.1 ADLC vs. LZS Compression Comparison
These numbers have been generated by using an internally created test program. This
program takes the target benchmark files and splits the files into payload size chunks and
compresses each chunk in a separate invocation of the compression algorithm. This is
important to note because most compression algorithms are applied to the entire file data
set as a single invocation of the compression algorithm which is easier for other types of
compression algorithms (LZS, GZIP in specific). This, of course, does not apply to
streamed packet data across an IP network (ftp transfer, for example).
15–7