BIG project - need advice please!

[Matce]

Basically we are in the process of getting a beachfront connected to WIFI.

Attached is a rough drawing of the situation.

As a note why I am not asking ubiquiti directly: I did and they won't answer. Absolutely worst customer service ever. I anybody knows a company with similar prices, I am more than happy to switch everything over



Point 1 and 2 are our main towers. We are feeding internet to tower 1 thru a Cisco RV016 connected to 5 LTE devices..

Tower 1 and 2 will be connected by a Rocket5AC PTP in combination with a RocketDish 5G30 (so we need 1 unit each per tower, correct?). The distance between towers is about 2.5km without obstruction.

We then distribute the signal towards the beach thru Rocket 5AC PTMP combined with 2 Sector antennas AM-M-V5G-Ti (limited to 60degree and below each other to have a wider vertical signal-band) on each tower. Will we need 4 or 2 Rocket 5AC total then?

The distance between the Sector-antennas and the receiving antennas is a maximum of 1km.



On the beachfront (marked by twentyfive x) we will have the access points.

We would like to “grab the signal” with a NanoBeam NBE-M5-300 connected to a Unifi AP Outdoor+ access point. We will have 25 of those combinations at about 80m distance from each other.

We are thinking of connecting each one 8dBi and 15dBi antenna to the APs to have better coverage.



Will this network work the way we planned it?

Would you have any revisions?

We can have 5Ghz components to “distribute” the signal and then have 2.4Ghz to our clients?

How many clients can we connect?

Each access points holds how many clients?



Thanks in advance

 

[netwrks]

Did you post on the ubiquity forum? They are usually pretty quick to answer any posts..

http://community.ubnt.com/

 

[Matce]

Did you post on the ubiquity forum? They are usually pretty quick to answer any posts..

http://community.ubnt.com/

yes I did. ZERO replies

 

[System Error Message]

Some here have mentioned that ubiquiti access points only handle 10-15 clients. Mikrotik also offers similar but with more hardware resources.

Go for the highest tx power and a dish for point to point networks. You will need multiple dishes pointing to each other and designed for 5Ghz.

in wifi bridging ofcourse you can use 5 Ghz to bridge and 2.4Ghz for clients but the issue is that the radios share the same antennas (or the radio might be dual band itself) so it makes it difficult if you set up point to point bridge while also using it as a regular AP.

If you have a clear line of sight i would suggest laser communication.

 

[abailey]

Usually you get tons of replies over at the Ubiquiti forum. After going over to the forum and looking at your post you have over 150 views and no responses. I believe it is because you have given a very complicated scenario with few real details. One question I have is are all your client sites (the 80 or so sites on the beach) located pretty evenly between tower one and tower 2, as it appears in your drawing? I would change your question a little bit on the forum giving many more details of what you are actually trying to do and then ask for recommendations. Don't try to make people give you an answer with what you have chosen to use (as I don't believe you have chosen the best equipment or the best set-up and maybe others believe that way also). Details like how much of the beach are you trying to cover, the entire 2.5km area? How much bandwidth are you starting with (bandwidth to the internet). How many people do you think will be attaching to each node? In other words your project is extremely complicated and really needs an expert to do it right. You could spend a ton of money and end up with very poor beach wifi if the planning and equipment is not right.

 

[Matce]

Usually you get tons of replies over at the Ubiquiti forum. After going over to the forum and looking at your post you have over 150 views and no responses. I believe it is because you have given a very complicated scenario with few real details. One question I have is are all your client sites (the 80 or so sites on the beach) located pretty evenly between tower one and tower 2, as it appears in your drawing? I would change your question a little bit on the forum giving many more details of what you are actually trying to do and then ask for recommendations. Don't try to make people give you an answer with what you have chosen to use (as I don't believe you have chosen the best equipment or the best set-up and maybe others believe that way also). Details like how much of the beach are you trying to cover, the entire 2.5km area? How much bandwidth are you starting with (bandwidth to the internet). How many people do you think will be attaching to each node? In other words your project is extremely complicated and really needs an expert to do it right. You could spend a ton of money and end up with very poor beach wifi if the planning and equipment is not right.

Hi,
thanks for the input!
We would need to cover the entire beach area... its about 2.5km long and really no more than 100m wide (this is including INSIDE of restaurants/bars). For now, I would need capacity to connect a total of 1500ish peope - I figured, once everything is up and running, we can replace the APs with better ones, that can handle more capacity. At this point, to be perfectly honest, we are running on a tight budget.

 

[System Error Message]

you are going to need a really good router and a lot of APs (depending on internet capacity).

 

[thiggins]

Hi,
thanks for the input!
We would need to cover the entire beach area... its about 2.5km long and really no more than 100m wide (this is including INSIDE of restaurants/bars). For now, I would need capacity to connect a total of 1500ish peope - I figured, once everything is up and running, we can replace the APs with better ones, that can handle more capacity. At this point, to be perfectly honest, we are running on a tight budget.

This project needs a pro. If you can't afford one, don't do it, or roll it back in scope. Trying to design it based on free advice from forums is not the way to go. GET A PRO.

 

[azazel1024]

Yeah, definitely need someone who's job it is to deploy networks like this.

That said, to chime in only on the a couple of aspects.

If you are attempting to deploy the signal, I would use higher gain antennas, probably Yagi or parabolic to the point of distribution. I would reduce the density of the access points and go with higher gain antennas. Ideally you'd have 1:1 on the bridge to access point. The best setup you could have would be at each access point, having a single wireless bridge with a high gain antenna pointed back at the wireless distribution point, which would then have either a sector antenna or something like an 8-12dBi Omni to receive.

If using 5GHz, one of the things you could do is to have multiple wireless bridges at the distribution point to spread the wireless spectrum load.

THAT being said, you need to find out what the "backbone" to the internet is going to accommodate. If all you are realistically going to have is a 20Mbps link, then your wifi performance requirements are going to be very low (and also you are going to completely crush it with people at the beach trying to use it). If you have a 200Mbps link, different story.

If the AP spacing is 80 meters, then you do NOT want to use 15dBi antennas. At my house I have 7dBi antennas on my WDR3600 and I can get a fairly strong usable signal at 40 meters, easy. Even alternating channels, you'd possibly have some signal overlap 3 APs down (very weak overlap, but overlap). I would use 8-9dBi if the spacing is every 80 meters. I'd only use 12-15dBi if the spacing was 100-150 meters. You are also going to have to watch the beach slope and antenna positioning VERY closely with anything over 9dBi. Heck, even with 7dBi you'll have to watch it if the beach shelves sharply.

Unless this is a very low population beach, but a big area you are probably going to get crushed. These days I see lots of people just playing around on their phone or tablet at the beach. I am sure most aren't streaming Netflix, but with the area you are talking about, I can't help envisioning a few thousand people at maybe 50-100Kbps each...

 

[L&LD]

The budget amount is important to note here. At least a range.

Hi,
thanks for the input!
We would need to cover the entire beach area... its about 2.5km long and really no more than 100m wide (this is including INSIDE of restaurants/bars). For now, I would need capacity to connect a total of 1500ish peope - I figured, once everything is up and running, we can replace the APs with better ones, that can handle more capacity. At this point, to be perfectly honest, we are running on a tight budget.

 

[Matce]

Yeah, definitely need someone who's job it is to deploy networks like this.

That said, to chime in only on the a couple of aspects.

If you are attempting to deploy the signal, I would use higher gain antennas, probably Yagi or parabolic to the point of distribution. I would reduce the density of the access points and go with higher gain antennas. Ideally you'd have 1:1 on the bridge to access point. The best setup you could have would be at each access point, having a single wireless bridge with a high gain antenna pointed back at the wireless distribution point, which would then have either a sector antenna or something like an 8-12dBi Omni to receive.

If using 5GHz, one of the things you could do is to have multiple wireless bridges at the distribution point to spread the wireless spectrum load.

THAT being said, you need to find out what the "backbone" to the internet is going to accommodate. If all you are realistically going to have is a 20Mbps link, then your wifi performance requirements are going to be very low (and also you are going to completely crush it with people at the beach trying to use it). If you have a 200Mbps link, different story.

If the AP spacing is 80 meters, then you do NOT want to use 15dBi antennas. At my house I have 7dBi antennas on my WDR3600 and I can get a fairly strong usable signal at 40 meters, easy. Even alternating channels, you'd possibly have some signal overlap 3 APs down (very weak overlap, but overlap). I would use 8-9dBi if the spacing is every 80 meters. I'd only use 12-15dBi if the spacing was 100-150 meters. You are also going to have to watch the beach slope and antenna positioning VERY closely with anything over 9dBi. Heck, even with 7dBi you'll have to watch it if the beach shelves sharply.

Unless this is a very low population beach, but a big area you are probably going to get crushed. These days I see lots of people just playing around on their phone or tablet at the beach. I am sure most aren't streaming Netflix, but with the area you are talking about, I can't help envisioning a few thousand people at maybe 50-100Kbps each...

thanks for your input. duly noted!

The budget amount is important to note here. At least a range.

15.000 USD for equipment
Laborcost is VERY cheap here, so that is not an issue.

 

[System Error Message]

thanks for your input. duly noted!



15.000 USD for equipment
Laborcost is VERY cheap here, so that is not an issue.

There are many ways you can set things up but stick to a few important things.
Wiring LAN is better than WLANs so be creative. You can use fibre optics, VDSL and so on. If you have to use point to point wifi to your gateway than make sure that the APs doing point to point wifi are only doing such so you can angle all their antennas to each other. You connect the P2P AP to the standard AP via ethernet. This gives you better bridging performance and cuts out antenna confusion and makes it easier to plan. Wireless bridge example ((([normal AP])))----[P2P AP])))))))) (((((((([P2P AP]-----Router will give you the best wifi bridging performance.

Choose your equipment wisely. Make sure it supports the features and load that you want. Look at the hardware specs such as RAM, CPU, etc. Ubiquiti APs can only support 15 clients so you need something else. If you want to use SFP mikrotik APs (9xx) have them and POE in. Make sure to choose the right antennas because 5Ghz dishes for point to point are different from 2.4Ghz ones. Make sure the equipment is well protected in weatherproof box and against theft and lightning. If you are giving wifi to one area and not all around the AP you can improve reception using directional antenna.

You than need a good router to connect everything. You will wire/link your APs to a central switch/router/switch + APs. QoS and bandwidth management is crucial so i suggest using pfsense or mikrotik or cisco or juniper that can perform advanced QoS and per user control including security against the pineapple hack.

For your wifi performance i suggest using the lower channels of each bands for distributing wifi. P2P links can use the upper channels. Inform others with wifi in surrounding area if you have to, tell them to stay off the lower channels (you can tempt them saying the upper channels give better throughput which is partially true).

Advertised range of omni directional wifi is 100M but i would say it would perform about 50meters unless you use better APs.

 

[abailey]

Choose your equipment wisely. Make sure it supports the features and load that you want. Look at the hardware specs such as RAM, CPU, etc. Ubiquiti APs can only support 15 clients so you need something else.

Not sure where you came up with this. If you go to the Ubiquiti forum you will see several implementations where there are 100 or more connections to a Ubiquiti AP. Ubiquiti makes good stuff, but like any manufacturer their are always compromises (usually based on price). I know you are a MikroTik fanboy but they are not always the best solution to every problem.

 

[System Error Message]

Not sure where you came up with this. If you go to the Ubiquiti forum you will see several implementations where there are 100 or more connections to a Ubiquiti AP. Ubiquiti makes good stuff, but like any manufacturer their are always compromises (usually based on price). I know you are a MikroTik fanboy but they are not always the best solution to every problem.

Some of ubiquiti APs dont support that many clients but the rest do. I also do recommend ubiquiti routers and switches too. The issue with Ubiquiti is that some of their hardware is either a hit or a miss. There are many complaints about how buggy their edgerouter lite is but their edgerouter 5 is much much better even though it uses the same hardware inside.

As i advised to the OP is to check the hardware specs which gives an idea of how many clients it can support. Clients, not connections. Some businesses have used a single netgear R7000s with more than 100 clients but it has 256MB of ram so thats not an issue. Some outdoor wireless AP from ubiquiti only have 4 to 8MB of ram. I know the radio handles wireless task but the clients are handled by the CPU so the system itself has to keep track of clients based on their mac addresses and routing. In most APs the radio only performs layer 1 while the CPU is left to perform layer 2. In APs like the ASUS AC87U has tiny CPUs to perform layer 2 for wifi and does the layer 2 conversion to connect directly to the switch chip. This is why for many APs it has a limited number of clients based on ram while CPU affects throughput for bridging. Even when i recommended ubiquiti APs back than they were much cheaper and reliable but they would only work reliably up to 15 clients and had issues handling more at the same time. The APs only had 4MB of ram and this was deployed outdoor in a college. The maximum i ever saw in the admin stats was 16 connected at a time, never more even though the population density per AP was way higher. The older enterprise based APs had no issues dealing with more clients but they were very unreliable outdoors (despite being an outdoor AP and costing 10x ubiquiti) and getting destroyed easily by lightning even with all the grounding and protection. they also had much more ram too.

 

[coxhaus]

My only thoughts are if you use pin point dishes on a beach with high winds is probably going to be problem. Lasers may be better if you don't have heavy fog or Hobie sails blocking the lasers. People drag all kinds of tall things up on the beach.I don't know where you are located. But you are beyond my scope.

 

[azazel1024]

My only thoughts are if you use pin point dishes on a beach with high winds is probably going to be problem. Lasers may be better if you don't have heavy fog or Hobie sails blocking the lasers. People drag all kinds of tall things up on the beach.I don't know where you are located. But you are beyond my scope.

Partly why I'd use Yagis. Less wind resistance and a somewhat broader half power beam pattern. Though parabolics can be fine too if you have a sturdy mount. With the distances discussed, a 14-16dBi yagi is just fine. It is only once you really start getting >>2km that you really need to be looking in to very high gain setups (and probably on both legs).

One thing to consider though, at a distance much over 1*km, you are going to need gear where you can change the packet wait, or whatever the heck it is called (having an old foggie moment). It is something like 80ns* IIRC, which works out to 1*km. Basically the sending router will transmit a certain number of packets and if doesn't hear an acknowledgement packet from the receiving router that it got the packets okay, it retransmits them.

This delay can be changed in some products and it is required for long links. Otherwise the wifi bridge/router/ap will thrash continuing to resend packets because it isn't getting the acknowledgement packet in time.

*I think that is the distance/pause. It may be 1 mile and whatever time that works out to. I just noticed that you do say that the longest distance is 2km in the setup. Not sure if there is a Wifi link that is actually the far, but even if it is 1km, that is really pushing it if the max is 1km (if 1 mile, then you'd really only be pushing things once you started getting up over 1500m. I like having a little wiggle room).

 

[Matce]

H

Partly why I'd use Yagis. Less wind resistance and a somewhat broader half power beam pattern. Though parabolics can be fine too if you have a sturdy mount. With the distances discussed, a 14-16dBi yagi is just fine. It is only once you really start getting >>2km that you really need to be looking in to very high gain setups (and probably on both legs).

One thing to consider though, at a distance much over 1*km, you are going to need gear where you can change the packet wait, or whatever the heck it is called (having an old foggie moment). It is something like 80ns* IIRC, which works out to 1*km. Basically the sending router will transmit a certain number of packets and if doesn't hear an acknowledgement packet from the receiving router that it got the packets okay, it retransmits them.

This delay can be changed in some products and it is required for long links. Otherwise the wifi bridge/router/ap will thrash continuing to resend packets because it isn't getting the acknowledgement packet in time.

*I think that is the distance/pause. It may be 1 mile and whatever time that works out to. I just noticed that you do say that the longest distance is 2km in the setup. Not sure if there is a Wifi link that is actually the far, but even if it is 1km, that is really pushing it if the max is 1km (if 1 mile, then you'd really only be pushing things once you started getting up over 1500m. I like having a little wiggle room).

Hey!

The sector-antenna is good for 52km+ and the Nanobeams for 5km+, so hopefully we are in the clear. We should have line-of-sight for almost all point-to-points.

thanks!

 

[Cloud200]

Depending on how cheap labor is by you it may be worth running fiber to each access point. I'm pretty sure mikrotik makes one that can take an sfp directly with no media converter.
Also, how are you going to power each WAP? Running cables via the same method may make sense.

 

[System Error Message]

OP tells me that its not viable to wire things up because people tend to trip over or cut cables by dragging along all sorts of weird stuff. Digging requires a permit as he is providing internet to places along a beach (all part of the same establishment). shallow digging has the same issue regarding weird stuff.

Mikrotik has inexpensive direct SFP to SFP cable, Ethernet SFP modules and different fibre optic modules. They also have SFP+ versions of modules and direct incase you need 10Gb/s.

 

[Cloud200]

Right

OP tells me that its not viable to wire things up because people tend to trip over or cut cables by dragging along all sorts of weird stuff. Digging requires a permit as he is providing internet to places along a beach (all part of the same establishment). shallow digging has the same issue regarding weird stuff.

Mikrotik has inexpensive direct SFP to SFP cable, Ethernet SFP modules and different fibre optic modules. They also have SFP+ versions of modules and direct incase you need 10Gb/s.

Right, but how are the WAPs going to be powered? Solar/Wind & battery?