BROADCOM CONFIDENTIAL
Regulatory FCC Band Edge Testing of Broadcom 802.11a/g/n Radios
Application Note
Revision 1.0
February 10th, 2007
1. Introduction
Broadcom 802.11a/g/n WLAN radios use packet based signaling for regulatory testing. The transmission is forced in various ways (see section 2) to ensure up to a maximum of 96% duty cycle. These test modes provide a case which is representative of the worst case real life transmission.
The earlier 802.11b specification detailed using a 100% duty cycle modulated test mode for EVM testin
g. In many cases some of the early WLAN test methodology was based upon using this continuous test mode. The FCC “Marker Delta Method” for band edge measurement is just one technique that is based upon a continuous test signal.
With the introduction of OFDM signaling and smarter test equipment the more recent 802.11a/g/n specifications detail packet based test modes and gated measurement techniques. In these cases test techniques previously dependant upon a continuous test signal are no longer appropriate. Note: Broadcom WLAN radios do not support 100% duty cycle OFDM modulated signaling.
The FCC band edge requirement is just one case where care must be taken to ensure that the test procedure used is appropriate for packet based signaling. This document has been created as a test guide to ensure that appropriate band edge test methodology is used and that equipment is not penalized, in terms of maximum power, due to the packet based signaling being used for test.
2. Packet Based Signaling
Where the test requirement is based upon an average measurement technique (and the measurement includes no associated duty cycle correction) the worst case mode is determined by the mode of operation with the highest duty cycle. Note: FCC Average Band edge testing against the
54dBuV/m limit is just one case where this applies.
To ensure that the measurement is tested and compliant in the worst case real life transmission mode a number of things need to be done to force the highest duty cycle:
- Packet Length set to worst case for host environment (see section 4 under epi_ttcp)
- Reduced header mode (e.g. gmode performance)
- It is also important to ensure that there is no interfering traffic that could affect the duty cycle under test.
- “wl amdpu 1”
- “wl frameburst 1”
3. Worst case Modulation Schemes/Rates.
The following modes have the worst case duty cycles and should be used for average based regulatory testing such as Band edge:
802.11b: 1Mbps
802.11g: 6Mbps
802.11a: 6Mbps
802.11n (20MHz): MCS0 6.5Mbps
802.11n (40MHz): MCS32 6Mbps (where supported) or MCS0 13.5Mbps
4.
The application can be used to transmit UDP packets from the equipment under test either to a Broadcast IP address (STA testing) or to an associated IP address (AP testing).
STA example: epi_ttcp –tsuHfm –l 8760 –n 9999999999 192.168.66.255
5. FCC Band edge requirements
FCC 15.209(a) Band Edge and Spurious Emissions
Average Measurements performed in 1MHz RBW, 10Hz VBW, Limit = 54dBuV/m as per - 15.209(a)
Peak Measurements performed in 1MHz RBW/VBW, Limit = 74dBuV/m as per 15.35(b) Peak and average measurements only apply for emissions in the restricted bands defined under 15.205(a).
FCC Emissions mask based upon measurement @ 3m.
6. Incorrect Band edge measurement techniques for Broadcom products incorporating “Packet Based Signaling”
Incorrect Procedure A.
In the past labs have been confused by the following FCC statement and used it as the basis for increasing VBW during “Average” band edge testing. Increasing the VBW makes the band edge limit more difficult to meet and therefore this is bad practice and will force the applicant of the product to reduce the power levels to meet band edge compliance unnecessarily.
The procedure above addresses systems that have a low duty cycle pulse transmission, such as radar, and can be modified for continuous transmission. This was included in the FCC procedures to make it easier to measure these types of systems that may exhibit pulse de-sensitization. In these cases you make an average (10Hz VBW) measurement of the continuous transmission and additionally apply the duty cycle correction factor, this
is then assessed against the 54dBuV/m limit.
The intent of this is NOT to apply to products with pulsed (or packet based transmission) that do not exhibit pulse de-sensitization. Furthermore the VBW is not intended to be modified as part of this procedure. Broadcom 802.11a/g/n radios do not exhibit pulse de-sensitization at any data rate supported. This can be witnessed by increasing the VBW from 10Hz in steps and observing the waveform under test. You can see no degradation in signal at the lower rates.
Note: ANSI 63.10 is a new standard being developed as the basis for testing of this type of product. The draft document clarifies that “Pulse De-Sensitization is only applicable to “Peak” signals and not “Average”
Incorrect Procedure B.
The FCC Marker Delta Method is not appropriate for Non-Continuous Transmission and should never be used for testing Broadcom 802.11a/g/n radios. Where equipment under test does not use a continuous transmission the “Average” correction factor is lower than required to make a correct reading. Use of this method will force the applicant to reduce power on the band edge unnecessarily.
7. Correct Band edge measurement technique for Broadcom products incorporating “Packet Based Signaling”
The most common concern is with regards to band edge requirements where power may be limited. This power limitation is typically worst case with CH9 40MHz signaling where the restricted band on the band edge is 2400-2483.5MHz.
Below is an example of the most appropriate measurements procedure for Broadcom products with regards to max transmit power prior to band edge limits, see also notes on the next page.
豫公网安备41160202000603
站长QQ:729038198
关于我们
投诉建议