Network Router User Manual

Release 11.0 Release Notes and User Guide Supplement
Issue 1, March 2011 Page 41
10.3 BENCHMARK DEFINITION
In a complex system, any measurement depends on system configuration, traffic mix, various
settings, and measurement techniques, and so to have reproducible results a “benchmark” is
defined.
System configuration
The PMP benchmark system consists of 3 SMs and 1 Advantage AP, as shown in Figure 8 on
page 42. Traffic generation and measurement equipment is connected to both SMs and the AP.
Traffic is generated such that any one packet attempts to traverse an SM and then the AP, or the
AP and then an SM. No SM-to-SM traffic is included in the benchmark. RF conditions are
maintained such that all links run at max rate (2X or 3X).
Traffic mix/Packet size
All generated packets have a size of 64 Bytes. The packet format used is a valid Ethernet/IP
packet. The performance of interest is performance near a 50% Downlink/Uplink Load Ratio.
PMP Settings
 Downlink Data: 50%
 Control Slots: 2
 Range: 2 miles
 Max rate (2X or 3X) Enabled
 Encryption: Enabled (DES modules)
 MIR: 20,000 kbits/sec sustained rate and 500,000 kbits burst allocation
(defaults)
 CIR: 0 (default)
 NAT: Disabled (default)
 VLAN: Disabled (default)
 High Priority: Disabled (default)
PTP Settings
 Downlink Data: 50%
 Max rate (2X or 3X) Enabled
 Encryption: Enabled (DES modules)
Measurement technique
1. Send a specific number of frames at a specific rate through SMs and AP (uplinks) and AP
and SM (downlink) simultaneously. This is the Offered Load. Count the frames that are
received correctly at both sides. This is the Carried Load. Repeat this through the load
rates of interest. Review the results, noting where the packet loss (the difference between
the Offered Load and Carried Load) is essentially zero (<0.001%).
2. Confirm results by running longer tests at selected load rates.
3. Confirm results by varying Downlink/Uplink Load Ratios to ensure no significant changes
around the 50% benchmark.