WiFi Hardware Advice

Hardware MikroTik hardware general
1

Dear All,

I’ve been asked to make a quota for wireless solution to provide a local soccer field with wifi for guests around alle the fields. In the light masts around the fields is power which can be used but no ethernet of fiber is available. So I want to offer a point to multipoint solution to all the light masts and from there an accespoint which distributes the wifi for the guests. At most there will be around 50 clients connected to a singe AP but usually less. The AP’s wil be hanging about 3 meter of the ground in light masts.

The dimensions of the ground are roughly 250 meters by 100 meters. At the lowest base of the grounds is a building where the internet originates. I’m planning on putting a mANTBox ax 15s as the base station on 5Ghz but Im a little lost as for the receiving part. I’m thinking of two units per pole, 1 for the wifi backbone and 1 for the AP wifi side for all the clients.

I’ve been playing with the Hamina wireless calculator but they don’t have all the mikrotik hardware and the free version only limits to 3 devices. That did however gave me the rough calculation of six AP’s around the soccersfields and 2 AP with in building to provide wifi.

Can anyone help me with the selection of the wifi access point needed and tips and tricks in general for this kind of deployment?

My mikrotik experience is mostly with firewalling, VPN’s and LTE and lesser with wifi.

Thanks for al the help in advance.

2

I think you should post a sketch of the area to be covered and where the light poles are.
You sure about the 3 meters off the ground? (usually light masts are much taller)

3

The height on which the hardware will be mounted is variable. If there is any advice on which height an ap should be mounted for optimal performance I would like know.

I’ll inserts an drawing of the situation later this evening

4

Of course it depends at what height the connected devices are and on the kind of device, the 3 m seems to me a bit too low for three reasons:

  1. possibility of interference with some obstacle (no direct line of sight in case of - say - people using signs)
  2. depending on the devices used, and antenna gain, the “thickness of the doughnut” may be much larger than three meters and you could have reflections from the ground (that could be a good thing or a bad thing)
  3. the AP’s should be out of reach of someone with a reasonably long (1 m or so) stick to protect from possible vandalism acts.

Anyway, I have seen often 4 to 6 m or higher recommended, example (HUAWEI):
https://support.huawei.com/enterprise/en/doc/EDOC1000060368/1b04ea99/what-installation-modes-are-supported-by-indoor-and-outdoor-aps-what-are-the-recommended-installation-heights

5

“So I want to offer a point to multipoint solution” - its a hard no as network backbone. If you estimate is right, 50 stations to AP is a lot, you need to cherish every airtime millisecond you can.

As mentioned, we need more information. Start by more detailed location information, you can use satellite image of this stadium and write over it dimensions, where wifi is needed, where those poles are, where will be internet origin point, etc.

6

Yep, the “magic” number I have often seen cited for enterprise wifi network planning is 30 clients per AP.

Or, if you prefer, for schools, one AP (wired) per classroom.

“Good” devices can - in theory - support up to 128, 256 or 512 clients (some even 1024), but - in practice - most vendors recommend 30 or so max “active” ones, only as an example see this Zixel document:
https://support.zyxel.eu/hc/en-us/articles/360005190180-Access-Points-Wireless-maximum-number-of-users-that-can-be-connected-to-Zyxel-Wi-Fi-Access-Points

And of course it also depends on the bandwidth that the PTMP can provide.

7

You can push more clients, but only if modulations on most stations are good, every 5ghz radio has clean channel to work whit (2.4ghz of), there is traffic shaping (including mcast/bcast filtering etc), good roaming etc. Also throughput you end up getting is acceptable.

Only wireless option as networks backbone I can recommend is 60ghz connections between poles and base station.

8

image
image727×738 95 KB

Hereby the layout of the complex.

Are there any PtMp 60Ghz solutions or can I use a same hardware which is provided in a wireless wire set?
Im just not sure about the width of the beam. The two accespoints the fartest away are about 100 meters apart and about 300 meters from the base station.

I’m hoping for some solid advice. Thanks already in advance.

9

The “base station” (that is actually a “base AP”) could be one or (I believe better) two:
https://mikrotik.com/product/cubesa_60pro_ac

one serving the left 3 CPE’s and one serving the right 3 CPE’s.

The corresponding 6 stations/CPE should be the “normal” cubes:
https://mikrotik.com/product/cube_60pro_ac

Coupled to 6 access points.
Assuming that the soccer players don’t need wi-fi in the field, instead of the omnidirectional (360°) AP’s (as sketched in the drawing) I would think at wAP AX’s:
https://mikrotik.com/product/wap_ax
that should cover more like 90-120 degrees (according to brochure 180)
In the worst case you might need to add another two on the first two masts (nearest to the base station) or on the same pole (IMHO better as they will be cabled) as the base station to better cover the first green area.

Of course the internal Ap’s can be anything, if ceiling mounted, cAP Ax:
https://mikrotik.com/product/cap_ax

10

Thank you very much for the advice. I used a 60Ghz connection before but not at these distances, thought that wouldnt work.

Your right that the players of on the field dont use wifi, however around the fields will be spectaters and they what to use the Wifi. So bassicly your saying that its better to use directional AP instead of omnidirectional AP because of client density?

I came up with a new drawing accoring to our suggestions. Every “half” cirkel represents an wap AX not inclusing the inside AP’s which will be cAP ax’s.

image
image397×672 76.4 KB

11

Yes, I thought that you wanted wi-fi coverage “around” the field, but your last scheme is clear, and of course you can play a little with the orientation of the wAP ax’s.

Generally speaking the more wide is the radiation angle from antennas the more the output power is “dispersed”, so having devices that cover narrower angles should provide better signal (in the covered areas).

The wap AX has a pattern (see here):

that you seem like having well represented in your sketch.

The “wireless wire” link between the “wide” Cube pro SA (if I remember correctly, but needs to be checked, it has a 60° angle) and the “narrow” Cube Pro’s (11°) should be good for up to 500-600 m, your case is more like 300-400 m, so it should be fine, and (again it should be re-checked) a Cube Pro SA should be able to “feed” up to 8 devices.

BUT personally I would think of using two of them because of the bandwidth.
Having two Cube Pro SA’s, each feeding 3 or 4 Cube Pro’s seems to me more reasonable.

A litlle word of warning, there have been reports of issues with weatherproofing with Cube Pro’s, incorrect distribution of liquid gasket, though it probably depends on production batches, see:

Again, personally, I would add an external additional caulking with good sealant before installation, better be safe than sorry.

I would try using a geotextile tape and some liquid rubber around the whole device, at least here it is available also in a light grey colour (besides black and red) that shouldn’t be too noticeable.

12

I was also thinking about 2 60GHz APs. One reason is bandwidth; the second is beam width - it’s not that wide, you’ll need good modulations. Research 60GHz models well, pick the most suited ones, and when installing, aim them properly.

WiFi distribution wise, WAP AX should be a solid choice, but make sure it's closer to stations. Also think about line of sight. For example, maybe mount on a pole at 2.5m instead of 3m. Go out to the field and imagine it full of people — does your device have clear or mostly clean LOS to the AP? I tested a WAP AX a while ago with regulatory domain restrictions, and range was fine up to 40–50m. The beam is a bit directional, but even from angles, it still worked fine.

Still hard to recommend without proper site pictures and more details. You might need manboxes if modulations are poor at locations with higher station density or greater need.

13

Not my install... But this is how another member covers a camp ground.

Point to Multipoint
Point to Multipoint1920×1440 123 KB

14

One must be careful when deploying high-gain antennas too close together. However, if you need 360-degree coverage, this setup can actually be a good option. It appears the 60 GHz backbone is running pole to pole - which works well and eliminates issues with beam width and distance to the access point (AP).

That said, it creates a bottleneck in overall throughput and uses more 60 GHz channels.