max no of subscribers for 10 MHZ channel width

RouterOS Wireless Networking
1

Hello





I am having a 5GHZ WISP system. each subscirber will connect on a speed of about 512Kbps ,

now I am suffering form interference in my area on 20 MHZ channel width. so thought of switching to 10 MHZ channel width.


\

  • will the maximum number of simulatiounously connecting ( active) CPE’s will be less for 10 MHZ channel width?

how many CPE’s can be connected on 10 MHZ channel width (each one connects on speed of 512Kbps)?

  • will the maximum practical throughput for 10 MHZ channel width be 25 Mbps?





    Regards
2

5GHz-A :

10MHz channel support 40 Client max on 802.11, if all the users simultaneous want 512k and if all clients have 100/100% CCQ, and hardware-retries set to 7 on both CPE and AP.

On Nstreme the band is little lower (max 38 client @ 512k), but the connections are more stabile = no retransmissions

On NV2 I think you go up to 50 users @ 512k

5GHz-N:

Really there is no differencies from 5GHz-a, the 10MHz limit is valid for both protocol
but the N protocol can go up to 32.5Mbit if the chip support this channel width.

you can add ~10 users for all type of modulation, with this protocol:
802.11 max 52
Nsteme max 50
NV2 max 64 users.

Remember: this data suppose all perfect link!

3

thank you
how it will be if I dont use mikrotik protocols (NV2, nstreme)

4

Already I have written on post: 802.11

If you use other hardware, than MikroTik, not all have support to 10MHz, and on 802.11 protolol, you can connect less client @512k than MikroTik

5

sorry I didn’t see what you wrote earlier about 802.11


thanks again :slight_smile:

6

Hi rextended,

How did you calculate the above? What I mean is, is there a standard formula for the above or can you point to an RFC?

Rgds,
Mark.

7

There is not a formula, there is my WISP experience from 2007…


Sorry I left one question unanswered:

will the maximum practical throughput for 10 MHZ channel width be 25 Mbps

No, on 802.11a with 10MHz channel (if chip support that) the theoretical link are @ 24Mbit,
obviously the real stream are ~18Mbit bi-directional

With NV2 the value are adaptive: the time not use for transmission from CPE to AP can be use for traffic from AP to client.

8

Hi rextended,

Thank you for the quick answer.

Buon fine settimana,
Mark.

9

Bonġu Mark,
è la prima volta che ho contatti con un Maltese. :astonished:
Ho molto rispetto nei confronti del to Popolo.

Buon lavoro,
Ċaw. :wink:

(Ah, Jiddispjacini non conosco il Malti, era solo per cortesia e ringraziamento, e spero di non aver scritto qualche cavolata in questo post! :blush: )

10

another question about 10 MHZ width


some people say it is true that 10MHZ is more clear from interference and more stable , but ping spikes and latency can more occur with 10MHZ because of the width of 10MHZ is narrower , and oise of 10MHZ hits your signal at the same frequency , all the channel that will drop and latency spikes can occur.

is that true or the real origin of that problem could be because of overloading 10MHZ channel with lots of users so when all of them do download the latency and ping spikes occur?

what is your opinion about that?

11

It’s ture, undeniably.

12

could you explain a little about it?

13

You have already explained that in your post!

14

there are two opinions that i dont know which one is right

is the latency , spikes because of
A- 10MHZ generallycould be more affected by noise ?

B- or it is because of overloading 10 MHZ with the bandwidth that it can’t withstand ( more than 30Mbps)

if it is because A , then w can’t do anything, but if it is because of B, we can prevent it by simply not overloading the 10MHZ channel with more than 25 or 30bps.


i want to know which one is right in order to take the right decision to swith to 10 MHZ or not

thanks

15

Sorry, I’m not English, I have understand now.

The right answer are B (in noise-less environment)

In perfect link 10MHz can carry max 20Mbit/s on non-NV2 setup

NV2 increase the download when are less upload from CPE

About the noise:
less MHz channel = less data and more signal strenght
more MHz channel = more data and less signal strength
more data for MHz (like the future 802.11ac) = more interferencies TO other operators & more interferencies FROM other opertors.

16

thank you

17

gracias :slight_smile: