Is EIRP being correctly adjusted in current ROS?

pamribeirox · February 5, 2020, 2:35pm

In our “wAP ac” ROS 6.45.7 “/interface wireless info country-info portugal” reports:

  ranges: 2402-2482/b,g,gn20,gn40(20dBm)
          2417-2457/g-turbo(20dBm)
          5170-5250/a,an20,an40,ac20,ac40,ac80,ac160,ac80+80(23dBm)/passive,indoor
          5170-5330/a,an20,an40,ac20,ac40,ac80,ac160,ac80+80(20dBm)/dfs,passive,indoor
          5250-5330/a,an20,an40,ac20,ac40,ac80,ac160,ac80+80(20dBm)/dfs,passive,indoor
          5490-5710/a,an20,an40,ac20,ac40,ac80,ac160,ac80+80(27dBm)/dfs,passive
          5190-5310/a-turbo(20dBm)/dfs
          5180-5300/a-turbo(20dBm)/dfs
          5520-5680/a-turbo(27dBm)/dfs,passive
          5510-5670/a-turbo(27dBm)/dfs,passive
          902-927/b,g,g-turbo,gn20,gn40(30dBm)

But with the interface is correctly configured for the country regulatory domain the total power reported is only 18dBm with the card producing 15dBm and the default (minimum) antenna-gain is 2dB (information from WEBFIG)

/interface wireless monitor [find default-name=wlan1] once 
                 status: running-ap
                channel: 2437/20/gn(18dBm)
      wireless-protocol: 802.11
            noise-floor: -95dBm
         overall-tx-ccq: 81%
     registered-clients: 1
  authenticated-clients: 1
            wmm-enabled: yes
      current-tx-powers: 1Mbps:15(15/18),2Mbps:15(15/18),5.5Mbps:15(15/18),11Mbps:15(15/18),6Mbps:15(15/18),9Mbps:15(15/18),12Mbps:15(15/18),18Mbps:15(15/18),24Mbps:15(15/18),36Mbps:15(15/18),48Mbps:15(15/18),54Mbps:15(15/18),HT20-0:15(15/18),HT20-1:15(15/18),HT20-2:15(15/18),HT20-3:15(15/18),HT20-4:15(15/18),HT20-5:15(15/18),HT20-6:15(15/18),HT20-7:15(15/18)
    notify-external-fdb: yes

If the maximum EIRP is 20dBm and the antennas in the device are reported to have 2dB of gain in this device, the card should produce 18dBm itself so the additional antenna gain will result in a signal within the regulatory limits. Am I missing something?

As a test, if the antenna gain is defined to be as an example 10dB, the reported “Total Tx Power” is reduced to 10dBm so, it seems the effect of the antenna gain is being subtracted to the “Total Tx Power” instead to the “Tx Power” (card).
I also don’t understand the origin of the 3dB difference reported between the “Tx Power” and the “Total Tx Power” it should be the antenna gain no? …

This results in worst conditions of connectivity to Mikrotik devices as compared to other manufacturers in similar conditions.
Our users are pointing to the Mikrotik devices telling “the signal were better with the old network” …

regards.

mkx · February 5, 2020, 6:54pm

Your device has two Tx “chains”, that’s two transmitters (a few MT devices have 3 or 4 transmitrers and a few have only one transmitter). EIRP is total power and that total power has to be divided to number of transmitters, hence on your device each transmitter can transmit with half of total power. Factor of 2 can be expressed as 3dB and factor 0.5 can be expressed as -3dB.

Some time ago most of WiFi vendors did not care about country regulations and used maximum Tx power available from hardware, could be up to 30dBm per chain. Which meant up to 15dB higher wireless signal from AP than it’s available now.

pamribeirox · February 6, 2020, 11:51am

Thank you for that detail I was missing …
I’ve confirmed that the “Tx Power” raises to 18dBm when I disable one of the Tx chains. The “Total Tx Power” remains at 18dBm in this situation.

So I’m concluding that when configured to some regulatory domain with “Tx Power Mode” as default and “Frequency Mode” as “regulatory-domain”, based on this variables:

AG - “Antenna Gain”
TTXP - “Total Tx Power” (sum of power from each contributing chain)
TXP - “Tx Power” (card output for each chain)
NTXC - number of “Tx Chains”
EIRP as the Effective Isotropic Radiated Power referred in the regulatory documentation.

TTXP[dBm] = EIRP[dBm] - AG[dB]
TXP[dBm] = TTXP[dBm] - 10*LOG(NTXC)

With these base values:
AG = 2
EIRP = 20dBm
NTXC = 2

And results in:
TTXP = 20 - 2 = 18dBm
TXP = 18 - 10*LOG(2) = 18 - 3 = 15dBm (as seen in the device)

One thing remains inconsistent (at least for me!).
The legacy rates from IEEE802.11g (OFDM 6-54Mbit/s), IEEE802.11b (5.5 and 11MBit/s) and IEEE802.11 (1 and 2Mbit/s) only use one Tx chain so, these rates should be using higher card output power TXP=18dBm no?

regards.

mkx · February 6, 2020, 12:25pm

IMHO this is a feature.

Consider the following: BSSID beacons are transmitted at minimum allowed speed depending on AP setup (can be at rates equaling 1Mbps) and it’s actually beacon signal strength that defines AP’s range. Thus if lowest possible speeds were transmitted with considerably higher power than the real data transmissions, then the AP coverage would be (artificially!) larger and many client devices would receive shitty service. Even if Tx power for all rates is kept the same, high-rate AP coverage is still smaller than low-rate coverage due to higher required SINR on reception side, additionally Rx sensitivity for higher rates is usually somehow worse, hence SINR can drop sharply on AP coverage edge.
And the above consideration actually applies to each Tx chain individually regardless of total number of Tx chains. With larger number of Tx chains (more than 2) where inter-chain separation over the air becomes an issue[], the SINR issue gets even worse as AP signal interferes itself.
[] with 2 chains the separation is easily done using different polarization planes but there are only 2 polarization planes for EM field in our 3-D space so this technique can’t be effectively used for more than 2 chains

Similar consideration applies when AP doesn’t adhere to country limits while client device does … which means that UL can be much worse than DL. Hence forcing too high Tx power on AP is usually a bad practice.

PackElend · July 10, 2021, 8:14pm

You may have look here as well: Wireless Tips and Tricks for RouterOS v6
There is a good illustration on “Tx Power” and “Total Tx Power”

thanks for summarizing like this, it raised some question marks at my side as well, as I went into details.
“Chains” and “Antenna Gain” has caused still confusion until this very moment.

I tried it in my word to verify if I understand it correctly.

Mulitple “Chains” is used to do MIMO but EIRP still applies. MIMO rises the possible bandwidth going trough the wireless but overall power is not raised due to EIRP, so it is trade of between those two extremes:

Range: one antenna with full power output (20dBm,100mW @ 2,4 GHz / 23 dBm, 200 mW @ 5 GHz)
Bandwith: e.g. 10 antennas, so if you have few devices each of them could his own antenna but they have to be close as power output:
per antenna 10mW 2,4 GHz, 20 mW 5 GHz

Antenna Gain itself should be regarded to optimize/limit the range of each AP individually. Have Tx-Power set at 20 dBm / 23 dBm and set antenna gain at whatever value needed to limit its range depending on its location as it is not set by CAPsMAN.

It shall be said that Antenna Gain itself is not adding power, it just concentrates the energy output by doing beam-forming.

the following references were quite helpful for me, to get these things sorted: