WLAN: Maximum Transmission Power (ETSI)

Different radio bands require different regulatory power limits. Each device you can buy, must not exceed any power limits given by the regulatory domain you want to deploy it in. The values given here are taken from the ETSI standards and apply to european countries. Some of the terms related to regulatory demands or transmission power used in this post are explained in this post.

2.4 GHz

There exists two EIRP power limits for the 2.4 GHz band, one for 802.11b rates with CCK modulation (1, 2, 5.5 and 11 Mbps) and one for 802.11g/n rates with OFDM modulation. The limit is set to 20 dBm (100 mW) for OFDM and 18 dBm (63 mW) for CCK.

The spectral power limitation of 10 dBm/MHz (10 mW/MHz) causes the lower power limit for 802.11b. As the spectral mask of the CCK modulation looks more like a sombrero, we see a high spectral power per MHz at the center and a lower one at the edges. So if you don’t lower the Tx power generally to 18 dBm, you exceed the spectral power limitation at the center of a 802.11b 20 MHz channel. For OFDM, the spectral mask looks more like a rectangular, so the power is nearly distributed equally, with an idealistic 7 dBm/MHz (5 mW/MHz) over a 20 MHz channel for example, and the maximum power limit of 20 dBm can be used.

5 GHz

Since the 5 GHz band is divided into two different ETSI Radio Local Area Network (RLAN) bands of 5150 to 5350 MHz and 5470 to 5725 MHz, which can be compared to the FCC Unlicensed National Information Infrastructure (U-NII) bands in the US, each band can have different power limits. As 802.11 only uses OFDM modulation in this radio band, there are no modulation specific regulations, only frequency specific.

RLAN band 1 (5150 to 5350 MHz)

Indoor only sub-band I (5150 – 5250 MHz)

The first RLAN sub-band includes the channels 36 to 48 and has an EIRP power limit to 23 dBm (200 mW). These channels are considered for indoor only usage and do not require any Dynamic Frquency Selection (DFS) or Transmit Power Control (TPC) features. It is comparable to FCC U-NII-1.

Indoor only sub-band II (5250 – 5350 MHz)

In the second sub-band of the RLAN band 1 with channels 52 to 64, the ETSI has set the EIRP power limit to 23 dBm (200 mW) for devices with TPC and 20 dBm (100 mW) for devices without TPC. For a device with TPC, the mean EIRP at the lowest power level of the TPC range must not exceed 17 dBm (50 mW). This band requires DFS support and is comparable to FCC U-NII-2.

RLAN band 2 (5470 to 5725 MHz)

Channels from 100 to 140 are part of the second RLAN band and have an EIRP power limit of 30 dBm (1000 mW) for TPC and 27 dBm (500 mW) for non-TPC devices or 20 dBm (100 mW) for devices without any TPC or DFS support. The mean EIRP power level for a slave device with TPC must not exceed 24 dBm at the the lowest TPC power level if the device is also capable of radar detection or 17 dBm otherwise. This band can be used for in- and outdoor deployments as well and is comparable to FCC U-NII-2e.

RLAN band 3 / Broadband Radio Access Networks (BRAN) (5725 – 5875 MHz)

Comparable to the FCC U-NII-3 (5725 – 5825 MHz) band with a higher upper frequency range, the ETSI has defined the channels 155 to 171 (155, 159, 163, 167, 172) for Broadband Wireless Access (BWA) use. The idea is to give internet access to locations without any wired access network available. The maximum EIRP output power has been set to 36 dBm (4000 mW) with the limitation of RF power into antenna of 30 dBm (1000 mW).

Furthermore, with the ERC Recommendation 70-03, it is allowed to deploy WLAN in this band with a maximum EIRP output power of 14 dB (25 mW) in this frequency band.

Clarification: The RF output power is defined as the mean equivalent isotropic radiated power (e.i.r.p.) of the equipment during a transmission burst. In general, the limits are valid for the device with antenna gain and cable loss and not only the output power of WLAN module.

ETSI documents used for this post

EN 300 328 v1.8.1 for 2.4 GHz

EN 301 893 v1.7.1 and EN 302 502 v1.2.1 for 5 GHz

Document 32005D0513 Article 4 on Indoor only usage of RLAN band 1.

ERC Recommendation 70-03 for short range devices in U-NII-3

Advice: The author hopes that the values given here are correct. If you can prove otherwise, feel free to comment or contact me directly.

*Update (2014-11-27): Use the term EIRP power instead of Tx power, I also added a clarification.*

*Update (2014-12-02): Substituted U-NII band definitions with ETSI RLAN bands.*

*Update (2014-12-09): Added link to post with term definitions, better definition of “lowest” power level with TPC*

*Update (2016-01-07): Correct indoor only use for frequency range 5250 – 5350 MHz, thanks to comment by Stefan Schneider.

*Update (2018-09-21): Added paragraph regarding usage of the U-NII-3 band with 25 mW EIRP.

22 thoughts on “WLAN: Maximum Transmission Power (ETSI)

  1. If you want to be 110% correct UNII is an American “Standard” term used. In ETSI the official names of thos bands are different. Yo may need to know/be aware this when reading EU contrives standard.

    Like

  2. This is a little bit confusing.
    U-NII-2e (5470 – 5725 MHz)

    restriction 1
    Channels from 100 to 140 are part of the U-NII-2e band and have a EIRP power limit of 30 dBm (1000 mW) for TPC and 27 dBm (500 mW) for non-TPC devices

    Had to read it several time to get the slave restrictions.
    You may also explain waht a slave device is so people get it correct.

    Restriction 2
    or 20 dBm (100 mW) for non-TPC and non-DFS slave devices.

    The “””lowest power”””” level of TPC must not exceed 24 dBm for slave devices that are at least able to detect radar interference or 17 dBm otherwise. This band can be used for in- and outdoor deployments as well.

    lowest seems strange. Please explain.

    Liked by 1 person

  3. Thank you mtroi for the awesome article. It is hard to find a good summary fir EU Wi-Fi regulations.

    I was look where the indoor regulations is stated in the EU document.
    I finally found this from the 5 Ghz docs you referenced: http://eur-lex.europa.eu/legal-content/EN/TXT/?uri=CELEX:32005D0513 (Article 4.)

    Just commenting this here as it may be useful to other people.

    As a second thought, I find it interesting that my Android phone (Nexus 6) can create a Wi-Fi hotspot/AP in the 5150 – 5350 MHz range (i.e. 36, 40, 44, 48 channel). I wonder if it Ok to have the hotspot enabled when I am in a car. 🙂

    Like

    • Hi Stefan,
      thank you very much for your feedback, I did post this to have an overview about ETSI regulations myself as it is rather painful looking stuff up in the reg. documents.

      We can hope that future phones will check via sensors if they are indoor and only then allow the use of these channels. I guess a car can count as indoor environment. 😛

      Like

  4. I’m concerned about your line “There exists two EIRP power limits for the 2.4 GHz band, one for 802.11b rates with CCK modulation (1, 2, 5.5 and 11 Mbps) and one for 802.11g/n rates with OFDM modulation. The limit is set to 20 dBm (100 mW) for OFDM and 18 dBm (63 mW) for CCK.”

    Do you have a reference for where you found this information?

    Like

    • Please disregard, I found the reference document you used to make this post:

      Wireless Transceiver Design: Mastering the Design of Modern Wireless Equipment and Systems

      Like

  5. Hi, very good and clear writing.

    Do you know if there are future plans for expanding the 5GHz wifi spectrum in the ETSI domain.
    Like the FCC is looking for expanding/adding more channels to the 5GHz wifi spectrum.

    Cheers, Kenny

    Like

    • Hi Kenny,
      that is a good question and as far as I know of, no such expansion is going to happen in the near future. But it’s been a long time since I checked, so I will try to get an update on that and maybe make a new post or at least a comment about any ETSI plans on enabling more unlimited spectrum in 5 GHz.

      kind regards
      mtroi

      Like

    • The ETSI spectrum domain is regulated by the European Conference of Postal and telecommunications administration (CEPT) with their sub-comittee Electronic Communications Comittee (ECC). After looking at this report http://www.erodocdb.dk/Docs/doc98/official/pdf/CEPTREP064.PDF we can see that CEPT is also investigating the use of spectrum between channels 64 and 100 as well as higher than 144 (5350 to 5470 MHz and 5725 to 5925 MHz) for radio access.

      Since this spectrum is also used by satellite systems, e.g. for earth exploration, the CEPT is currently asking work/task groups to come up with harmonised conditions to reduce interference between these systems and a Radio Local Area Network (RLAN) device like WLAN/Wi-Fi. Just think of DFS radar detection patterns, suitable power levels and that stuff.

      Is this a fair enough information for you, Kenny?

      Like

  6. Dear Author,
    Firstly Thank you for the work and making it available.
    My Question is concerning the use of BRAN (5725 – 5875 MHz):
    I would like to set-up a wireless bridge, 700m apart, here in Sweden in BRAN band.
    Is this a license free spectrum permitting non-commercial private use in EU? just like the RLAN1or2.
    Thanks in advance.

    Like

    • Hi dba,
      you should take a look at PTS.se and their frequency plan to find the information you need regarding the use of wireless equipment in the BRAN frequency band. Also take a look here: https://t.co/jvqnY4Y4hm

      My understanding is, that you need a permission, just like in germany. You don’t license a certain channel, but get the permission to use any channel in that band. Since I don’t understand swedish, the translations I got might give me some wrong information, so please check it out yourself.

      kind regards,
      mtroi

      *Update* I got a notice from a swedish contact. Look at §157 of http://ec.europa.eu/growth/tools-databases/tris/sv/index.cfm/search/?trisaction=search.detail&year=2015&num=327&dLang=EN :
      5.725-5.875 GHz: Radio transmitters for unspecified areas of application. Maximum power: 25 mW EIRP (14 dBm) – so you need really high gain directional antennas for your application.

      Like

      • mtroi, Thank you very much for the kind response.

        Apparently some ongoing work in draft stage concerning 5.725-5.875 GHz in Sweden hopefully aligned with EU/ETSI. It is dissapointing to learn there is such a huge difference between Sweden Vs ETSI for 5.725-5.875 14dBm Vs 36dBm 😦

        My Scenario:

        I am considering setting-up a wireless ethernet bridge 700m apart for private use, bought Ubiquiti NanoM5.

        Ubiquiti GUI has different country profiles predefined. Allowed output power level depends on country profile selected and frequency. I saw there is difference between country profiles Sweden and Germany, in my case it’s the German profile that barely establishes a usable bridge connection.

        Germany country profile allows max power 20dBm@ 5735 MHz(DFS). Internal antenna gain of 16dBI.

        Whereas Sweden profile max power = 14dBM@5700 MHz. This power level is insufficient for establishing the link ( due to NLOS, trees in between).Frequencies >5700MHz not available for selection in the GUI, but now based on your updated information they don’t matter anyway, cos the power level is even lower for frequencies higher than 5725 MHz as per Swedish regulations.

        Its a bit dissapointing to learn there could be such a big difference within EU ? I mean Sweden Vs ETSI 14 dBm Vs 36dBm.

        Just wondering about your suggestion of high gain antenna, please help me understand.
        Based on my understanding from different forums, higher the gain of attenna lesser the allowed output power ( to meet country specifi E.I.R.P). So in the end Attenna Gain benefit lost due to lower power to antenna. Perhaps there is a recieving sensitivity gain, but the signal might have hard time reaching the destination. I am considering trying out a 22dBi Antenna system,but based on priliminary results above (Usable connection at 20dBm with 16 dbi Antenna ), i am not sure if this is something even worth trying ?

        I also have difficulties in understanding why a high gain antenna ( relatively complex compared to omini directional) gets penalized by allowing lower amount of power to feed it, by the EIRP. Becouse in the end the whole point of getting to farther distances by focusing the beam ins’t lost ?

        Thank you,

        Like

  7. Hi dba,
    my guess is that the ETSI documents define the frame, but each country can define its own regulations as long as the frame is not “exceeded”. But nonetheless it would be great to see a harmonised standard frame throughout the EU.

    The regulated total output power, which is power from the “total power = WLAN module output power – cable loss + antenna gain”, is always based on the power of an “Equivalent Isotropically Radiated Power” (EIRP). I described the relevant terms in this post: https://wlan1nde.wordpress.com/2014/12/09/wlan-transmission-power-terms/

    So if you add an antenna with a directional focus to the system, the system is treated as if it sends the same amount of power in all directions. Otheriwse, the limit is always exceeded and the sense of regulating powers doesn’t make sense. It should not matter in which direction power is transmitted, as long as you respect the power limit. Did you get it? 🙂

    As you tested your bridge with the german country settings and the result was barely usuable, the 22 dBi antenna might not do any good with swedish country setting, although the transmitted as well as received energy is focused a bit more. You should test it as long as it is not to difficult to install.

    Anyway, your scenario with NLOS and trees (consider the Fresnel zone as well!), seems to require a higher position for the APs or a different access technology. :-/

    kind regards,
    mtroi

    Like

  8. Thank you very much mtroi,

    “different access technology”, wondering if you can suggest any technologies here ? 

    Unfortunately conventional connectivity not an promising option i.e. Fiber& Cable DSL not available. Telephone ADSL perhaps a possibility but one operator monopoly
    Pricing way too high for a mere 10 Mbps. Presently using a 3G/4G connection, but very limited in usage due to a low GB limit.

    Getting better LOS is very challenging, I am a tenant with no easy access to the roof top, causing all this hassles with LOS.

    Perhaps 900MHz link would be good for NLOS applications, but I guess that is licensed band in EU ( Sweden ) ? 2.4GHz wifi, I think would not be much better in terms of NLOS, and added comes a crowded band with too many users.

    Seems like I am hitting end of the road in this regards….

    Like

  9. You could try to get a connection on channels 100 to 140 with your 22 dBi antenna due to 30 dBm power limit. There is also 802.11ah (aka Wi-Fi HaLow) for sub 1GHz applications, but I haven’t looked for any ETSI regulations or even equipment.

    Sorry, I can’t recommend any other wireless transmission option to you. 😦

    Like

    • Wikipedia says 25 mW for SRD (short range devices like amateur radio), but if you take a look at ETSI EN 302 502 V1.2.1 (2008-07) chapter 4.2 table 1, it says that for 20 MHz channel bandwidth, the mean EIRP (output at the antenna) can go up to 36 dBm (4000 mW).
      The ECC report, that is cited for the 25 mW limitation, claims this in Annex 1 for non-specific short range devices. Since WLAN is a broadband radio access network, it is not non-specific and EN 302 502 can be applied. Nonetheless, here in Germany, you need to get a license from the “Bundesnetzagentur” (federal network agency) to operate a WLAN device in this band.

      Like

      • Seems like I’m wrong, you can deploy WLAN as “non-specific” short range device in UNII-3 (up to 25 mW / 14 dB EIRP). I’ll update the post.

        Like

Leave a comment

This site uses Akismet to reduce spam. Learn how your comment data is processed.