User's Manual

80-VT321-1 B 13 Qualcomm Confidential and Proprietary
MAY CONTAIN U.S. AND INTERNATIONAL EXPORT CONTROLLED INFORMATION
5 Performance Test Configuration 1
This chapter provides some guidelines for throughput performance testing of Qualcomm
WCN1320-based reference designs.
2
3
5
6
7
8
9
10
11
12
13
15
16
17
18
19
20
21
22
23
24
5.1 Qualcomm recommendations for best performance 4
TCP/IP settings on systems sourcing and syncing data may impact throughput performance. For
example, on PCs loaded with Windows XP, Qualcomm recommend the following TCP/IP
settings:
 TcpMaxDupAcks-2
 TcpWindowSize-256960 or greater
These TCP settings can be done using Windows registry or a GUI-based tool like TCPOptimizer
available free on the Internet.
Qualcomm has observed in its performance testing that to completely saturate the wireless link it
may be required to use 10 or more TCP/UDP streams when using test tools like Chariot or IPerf.
5.2 Interference 14
As with all other 802.11-based equipment, interference (802.11 and non-802.11-based
interference from outside sources and ambient noise from PC platforms) impacts throughput
performance. In many environments, it may be necessary to use the 5 GHz band to find an
interference free channel for performance testing. Note that because of differences in propagation
characteristics between 2.4 and 5 GHz, maximum range in the 5 GHz band will typically be less
than maximum range in 2.4 GHz.
In an environment with heavy interference, better throughput results may be obtained using a
20 MHz mode instead of a 20/40 MHz mode. In an interference free environment, 20/40 MHz
mode produces better throughput results.