Table of Contents
Advertisement
Packet over SDH (POS)
Throughput Testing
What determines the packet rate?
• Channel bandwidth (STM-0e/STM-1/STM-4/STM-16).
• Packet size.
• HDLC byte stuffing.
• Inter packet gap.
Available bandwidth
The available bandwidth (in bytes/second) for packet transmission for each channel
is as follows:
• AU-3
• AU4
• AU-4-4C 74880000
• AU-4-16C 299520000
Packet Size
For IP, the packet size is specified in terms of the IP datagram size including the 20
byte IP header. The actual packet size transmitted includes 4 bytes of PPP/HDLC
overhead plus the FCS (HDLC Frame Check Sum) which is 2 bytes for FCS16 or 4
bytes for FCS32. Hence, for a 75 byte IP datagram with FCS32 the actual packet
size is 75 + 4 + 4 = 83 bytes.
The packet size may be further modified by HDLC stuffing. This will add one extra
byte every time the flag (7E) or escape (7D) octet appears in the packet.
The actual transmitted packet rate (in packets/second) will then be:
channel bandwidth ÷ (packet size (including overhead and stuffing) + gap size).
For our example (see display on previous page), ignoring the effect of HDLC byte
stuffing, we would expect a packet rate of 299520000 ÷ (83 + 17) = 2995200 pkts/
sec. However, due to the effects of HDLC stuffing, this figure may be slightly
reduced. To eliminate the uncertainty due to stuffing, it is possible to choose
carefully the IP header and payload so that no stuffing takes place. The procedure
given on the next page shows how this can be done:
Artisan Technology Group - Quality Instrumentation ... Guaranteed | (888) 88-SOURCE | www.artisantg.com
6048000
18720000
209
Advertisement
Table of Contents
