I'm new here, so please excuse the ignorance & if this is a nauseatingly frequently asked question.

Is there a good way to estimate wifi 2.4Ghz penetration ahead of time? I can visually approximate how many access points are going to be needed based on previous experience, but I'm wondering, is there a better way to do this?

In terms of equipment, I'm looking at the cAP or wAP units, either lite or ac...

thanks!

based on my experience, if you need to have bigger coverage, get those with the "H" mark (high powered), e.g. RB951-2Hnd, and avoid those Lite model.

A single, RB951-2Hnd basically can cover an open area as huge as 45ft x 60ft (or more or less around that) and you would still get half the signal strength at the edge of that said area size, as long as you place the AP strategically.

The answer always depends on the local situation (not described), type of clients (not described), their behavior (not described), expected functionality - mainly throughput (not described) and other restrictions (not described).

In this case whatever you do might be good or bad....

The answer always depends on the local situation (not described), type of clients (not described), their behavior (not described), expected functionality - mainly throughput (not described) and other restrictions (not described).

OK, fair point.

It's a residential block of 8 units, 4 on each side, with a eating hall in the middle (9 main buildings), length is 175 feet, width is 65 feet, or 100 feet if you include the hall.

Being a residential block, clients will range from PC's to smart TV's, Xbox's, tablets & other mobile devices & the odd IoT device.

behavior will include total bandwidth of about 1.5TB a month, (total) speed will only need to be about 20Mb/s, mostly accessing consumer url's. Maybe the occasional VPN.

restrictions (since you asked), some blacklisted IP's, individual client bandwidth limited to 10Mb/s, hopefully some traffic shaping rules in place.

That's a general sketch, nothing too elaborate. My question specifically is how would I approximate how many AP's to install. Mentally, I'm thinking either 2, to cover the width, or 4 or 5 to cover each main portion of the building. 5 seems like overkill, but better than having dead spots... Thoughts?

thanks!

Ok. I would suggest one ap (whatever, according to budget) per building, connect by wire to central router /switch. For such total bandwidth of 20mbits CRS switches (CRS125 for example) might be powerful enough to make natting and queueing. When a switch start to be a bottleneck, leave it for switching task only and put a router (750Gr3 or hap ac2 for example) in front of it. Don't forget to regulate down the tx power of each ap.

Hi Jarda,

Hmmm, maybe I should have given more information. I estimate that only 30 or so clients will be connected at any one time...

I don't mean to second guess your suggestion. Maybe let me check again, wouldn't a CRS125 be overkill?

And to be clear, you're suggesting one access point per apartment (the apartments are small 30 feet x 20 feet)?

thanks!

You wrote 9 buildings. If they are small, built from wood and close together maybe one wap between two neighboring could be enough. CRS is not an overkill, you need to have enough ports to connect all devices together. Sure you can use cheap dumb switch but you loose manageability. In such case you would need additional router for sure. But combination of small router with dumb switch could be cheaper than single CRS. On the other side, CRS has cable test function that will tell you where the cable is broken and how... Very useful. Look at RB2011 which is generally the same like CRS125 (different inner switch chips only).

Cheapest and yet maybe working solution in your case could be rb941. One as router and 2-3 others in repeating mode around. 10mbits can be achievable by such combination.

Thank you so very much for taking the time to make suggestions etc. much appreciated!

I think that gives me a reasonable idea how to proceed... Only question in the back of my mind is whether it would be useful to run "the Dude" as well. I'm new to Mikrotik, so I'm still feeling my way forward. Am I right thinking that the Dude would need to run on the gateway router, or could I somehow bridge in a hEX RB750Gr3 (thinking cheapest options here)?

thanks!

Dude can run anywhere in the network. Even in remote location if you join the networks by tunnels. Dude is not necessary but nice to have. RB750Gr3 or rbd52g are the cheapest devices that can run it (use sd card or USB flash for database). You can also run dude virtually on any pc you have in CHR for free.

Back to the network. You can do it cheaply and improve continuously further or you can do it perfectly without need of a touch in next years. Your choice.

Back to the network. You can do it cheaply and improve continuously further or you can do it perfectly without need of a touch in next years. Your choice.

I have really appreciated your input & it has helped me think these issues through better, especially regarding future-proofing the system. I'll probably go with your suggestion of something like the CRS125.

I'll do some reading & might need to start a thread asking how to anticipate/read the throughput specs on routers/switches.

Good start. Try to find someone familiar with mikrotik in your neighborhood to get some local knowledge and help. It will be much more beneficial than theoretical ideas from the forum.