Summary We often get questions regarding whether it is possible to make measurements on radiated power (ERP) on a WLAN site which uses Frequency Hopping Spread Spectrum (FHSS) such as BreezeCom Pro.11 or Direct Sequence Spread Spectrum (DSSS) such as IEEE 802.11b. The question often occurs when talking about risks of getting caught for using higher output power than allowed (20 dBm ERP on the 2.4 GHz-bandet).Radiated power
Radiated power EIRP (dBm) is defined as the output power from the radio uteffekt Pout (dBm) - attenuation in cables L (dB) + antenna gain Gant (dBi).
PEIRP = Pout - L + Gant
An example is Lucent black card with 15 dBm output power 5 m RG223, with 0.9 dB/m attenuation and a 9 dBi patch antenna (VP165/24). This gives
15 - 5*0.9 + 9 = 19.5 dBm
Measurement method As a method for making measurements we use a spectrum analyser as receiver with a calibrated cable and antenna. We measure received signal for the frequency band of interest during a period of time, and use "peak hold", i.e. we collect top values. Gradually an image of the spectrum appears. There are two ways of getting a value on radiated power:
(a) absolut measurements
We make a link budget for the measurement according to as follows:
PRX (dBm) = PEIRP (dBm) - FSPL (dB) + GRX (dBi) - LRX (dB)
PEIRP is the unknown and searched output power from the device under test. Free space Path LOSS
FSPL is calculated as FSPL (dB) = 20*log10(4*pi*R/wl)
R = distance in metres
wl = wavelength in metres
The precision of this method comes from using a calibrated receiver / spectrum analyser and a calibrated cable and antenna. We perform measurements with a precision of ca +/- 3 dB.
(b) relative measurements
This method is similar to the one above, but more exact. We calibrate the measurement by applying a reference transmitter, with a known output power, on the same distance as the device under test. For this transmitter we use a calibrated cable and antenna, and make a complete calculation of the link budget.
The received signal from the reference transmitter is compared to the device under test. It is possible to get a precision of ca +/- 1 dB.
Choice of antenna, cable and distance The choice of antenna for the measurement is not critical, but it is important to use the dynamics of the spectrum analyser the best way, i.e. making measurements well above the noise floor. A cable can be measured with a precision of +/- 0.1 dB, while an antenna can be measured with a precision of ca +/- 1 dB.
The calculation of FSPL is very exact. However it is important to stay in the far-field from the device and the antenna under test. It is approximately defined as 10 wavelengths (1.2 m at 2400 MHz) for an antenna which is small compared to the wavelength or 4D2/wl for a large antenna, where D is the antennas size (aperture) in metres i.e. 12 m for a 60 cm parabolic dish. Allowed output power shall be measured for the direction where the antenna has its' maximum gain. It is not unusual that omnidirectional antennas has their maximum gain several degrees above the horizon.
The received signal shall be -32 dBm at 10 m distance from the object under test / the transmitter, if they hold 20 dBm EIRP and the measuring antenna is a 9 dBi panel antenna with 1 dB cable attenuation.
Making measurements within a tight budget
However, not everybody have access to a spectrum analyser. The reason for this article is to show that it is possible for a third party (or supervising authority) to make measurements on the output power from a link.
An alternative method of making measurements is to use Lucent Client Manager, monitoring software for Lucent WaveLan. It presents received signal strength in dBm and corresponds well with our measurements. We can deliver calibrated antennas and cables on request.
Measurement results
The pictures ../../../ below show spectrum measured with a HP 853A/8559 spectrum analyser. The window shows 2340-2540 MHz, i.e. 2440 MHz is in the middle of the image. The amplitude on the y-axis is 10 dB/square and the noise floor is, during this measurement, ca 60 dB under the signal. The amplitude is relative and does not contain information on absolute signal strength.
Products
