How Many Cameras Can I Run On A Ubiquity Nanostation M5?

I'm gonna try to do the calculations to see how many 3MP cameras a pair of Ubiquity LocoM5's will comfortably handle, but in the meantime I was wondering the maximum load anyone has successfully put on the M5 series (Loco or its bigger Nanostation cousin).

I've read the IPVM article on this product and looks like at least three cameras is possible, but how about 5 or 10? I realize it depends on the megapixels, frame rate, etc, but I'd like to get some real-world feedback if anybody has some.

For my purposes, distance is not a factor since I'm only going a couple hundred yards with clear line of sight. My plan was a couple LocoM5's & a switch for that out-building, but my client now wants to add a few more cameras than was originally planned.

Any input is greatly appreciated.


I realize it depends on the megapixels, frame rate, etc

Technically, no, none of that matters. What matters is bitrate. Yes, bitrate is going to mostly be a factor of frame-rate and resolution, but the nanostation (like any piece of network gear) does not care at all about fps, just mbps.

Distance and obstructions between the unit can also affect maximum throughput, as will other wifi gear in the area.

For the units you're talking about in an average setting you should be able to get about 45Mbps throughput. With a little bit of consideration paid to the install (picking a clear RF channel, proper aiming, positioning away from walls), you can get about 50-55 Mbps. You likely won't ever see more than 60Mbps of real throughput without a lot of luck.

How many cameras you can carry over 45-50Mbps will vary based on the camera itself (see other tests IPVM has done for bandwidth), but you should be able to get about 5-8 cameras on that link.

Maximum 60 Mbps?

Wow, that's a far cry from the 150 Mbps Ubiquiti touts on their website.

Would be feasible to use two pairs of radios?

That's an interesting thought.

Would each pair of M5's need it's own switch?

Would there be any interference between the two sets of nanostations, being they would be so close to each other? Although I guess setting them to different channels should take care of that.

I had a dealer use two sets of nanos side by side each using almost 40Mbps with no problem with interference. We could have had two Nanos pointing to one AP nano but since it was only 100Mb ethernet connection we didnt want to take the chance.

I would call his success more luck than logic. Most people wouldn't recommend 2 sets of transceivers that close together, but rather upgrading to the next tier of equipment that is intended for higher throughput.

I agree with A Mfg. If one head end radio won't suffice, you need a better radio.

However, if you need more than one for directional reasons, you can use an RF Armor shield kit.

RF Armor Website

We have these deployed with 12 3MP and 4 D1 cameras on a single link. We cap our 3MP bitrate at 4mbps VBR. The D1 is capped at 1024kbps VBR. We haven't seen a single issue, but it is using a Dahua DVR, not a real VMS, so there isn't much as far as reporting issues. Last time I checked at night, we were not close to saturating that link. The distance for that link is under 100'.

We have a second location that uses the same radios with 3 3MP cams, again VBR 4mbps capped. These have a Dahua NVR, so again, reporting is minimal, but have yet to have any issues, other than needing one reboot in 3 years in service.

A third location was recently installed with 5 3MP VBR 4mbps capped per radio (two sets) and at that location we are running DW Spectrum. Have yet to receive a single email about network issues. If you're not familiar with DWS, it usually nags you about network issues often. This site has been very quiet. So much so I checked the logs to be sure.

We have run 5 X 4MP cams from 200' to two back to back radios easily accommodating them and no issues.

As of today we are running 7 on ptp nanos m2 1/4 mile from each other at 48dcbms, in a unpolluted 2ghz area.

Thanks for all the replies. As always the input is very helpful.

Just to confirm, real world one way throughput of Nanostation M5s over short distances is in the order of 50Mbit/s, as opposed to the figures touted by Ubiquiti. This is primarily due to the processor performance (can't handle higher packets per second). So bigger packets = larger throughput.

Ubiquiti have recently launched the Nanobeam family of units. Similar pricing to Nanostations but with 560MHz processors instead of 400MHz. These should offer a significant throughput improvement, although the hard upper limit will be set by the 100Mbit/s Ethernet port.

Andrew,

Thanks so much for that input.

Did you mean to say the PowerBeam units? I just went to the Ubiquiti site and those are listed with the higher capacity. These PowerBeams look like the perfect solution for my project. The Nanostaions are just to close for comfort considering their much lower real-world throughput.

After reading the specs on the PowerBeams, it looks like they are all-encompassing units like the Nanostations in that there is no need for a seperate Rocket unit or the like. Am I correct?

Not quite. Ubiquiti has slightly confused matter by (a) rebranding some products and (b) having one product range (PowerBeam) with two different throughput capabilities.

So to clarify:

The new NanoBeam M product range is effectively a direct replacement for Nanostations and Nanostation Locos. 5GHz only, small, low cost and fully integrated. The new model has better antennas with cleaner radiation patterns and faster processors than the old models. See http://www.ubnt.com/airmax/nanobeamm/

The PowerBeam range is intended for longer distance links. For 5GHz it has larger dishes (12 inch, 16 inch or 20 inch) delivering higher gain. Radios and dishes come as separate components that need to be fitted together and unit prices are higher. Throughput performance depends upon the model. The 12 inch and 16 inch variants have effectively the same radio and processor as the Nanobeam Ms, so maximum performance should be similar but with less dropoff over distance. PowerBeams also have gigabit Ethernet ports. The 20 inch variant has a better processor and the radio technology is based upon 802.11ac rather than 802.11n. This is the only variant for which Ubiquiti quotes a 450 Mbps throughput. Again what people have seen in the real world is a quite a bit lower but it is still in the order of 250 Mbps

Thanks for that clarification.

If I understand correctly, the new NanoBeam M, even though listed at the same 150 Mbps will be much closer to that spec in real-world terms than the previous Nanostations.

Sorry for belaboring on the details, but as you can probably tell I don't have much experience in this type of wireless equipment so I want to make sure I understand the nuances.

I have a large project in process that includes a seperate garage building that was supposed to have 3-4 cameras (hence the Nanostaion setup). Now the client has added two more cameras, and last Friday dropped a bombshell on me. They now want to have the whole back truck lot covered, which will also tie into the garage setup and hence feed through wirelessly back to the main building.

So now, even though the PowerBeam is overkill distance wise, I'm actually contemplating it to handle the extra capacity in case the client adds the back-lot to the project. I had a feeling this was going to happen, so I ran a gigabit line to the antenna site on the main building just in case.

If you have any other suggestions for shorter distance (~200m), higher throughput wireless equipment, I'd be very appreciative.

Jerome,

Do you have a bitrate cap planned? How much data do you anticipate traveling through this wireless link? Have you surveyed the area for existing WiFi networks? Using 5GHz should help in congested areas, but one shouldn't assume 5GHz is always clear. You should do a site survey to be sure. Also, if it is that critical to the project, buy a couple of radios and run some testing. That way you have real world examples.

Thanks for the suggestions Jon.

I've got 3MP Hikvision bullets that I do plan on capping, but there are too many variables at the moment to have a hard set number. While the client is making up their mind about the back lot, I'm boning-up as much as possible on stronger alternatives to what I had planned.

Heck, I even looked at trenching over to the garage since it is all grass & compacted dirt driveway between the two buildings. But, it's a rural area with a clear path between buildings, and there is only a small wi-fi network nearby in the office. So wireless should work fine, especially if I bump the system up to a PowerBeam.

I assume truck radios are too far from the 5GHz range to come into play?

Do you have any idea about a total number of cams you need to support via the wireless backhaul?

My best guess at the moment is 4 for inside the garage, and probably 6 for the back lot, so a total of 10. Hopefully, I won't have to cap the cameras too far down because the clients are now spoiled with the video quality of the main system which is up & running (20 cameras). We'll see, I've already told them there are limits to what can be installed given their wireless budget.

Ten cams capped at 4Mbps each should work fine over a single set of M5s. But that won't give you any room for future growth.

Yeah, that was also my (although somewhat inexperienced) assessment. The PowerBeam is not that much more of a bump in budget, so I think I'm gonna go with the extra throughput.

To get the full radio rates of the ac product, you need a clear 80MHz channel. Assuming you are installing in the US, you might find that difficult given the small amount of spectrum you currently have available. So in practice you may be limited to 60 or 40 MHz. Nevertheless even at 40MHz you should be able to reach sustained user rates well in excess of 100 Mbps. So you are probably right to go with the flexibility offered by the Powerbeam. The larger dishes will also provide more directivity and hence greater isolation from external interference.

For the price there is no real alternative that will offer better performance. If you are willing to pay quite a lot more then you can move to carrier grade 5GHz point to point solutions such as Infinet Wireless or Cambium. The main advantages of these products are custom radio modules that provide more robust connectivity, better quality antennas and automated zero outage channel change in the event of interference.

The final option is to use millimetric (60GHz). Over a 200m distance the Siklu EH600 will provide reliable full duplex Gigabit capacity, all in a small package weighing around 4 pounds and costing around $5000 per link. Here in Europe the band is licence free - not sure of the rules in the USA.

Thanks so much for the alternate options Andrew.

It does sound like Ubiquiti is my only option price-wise. Even though my client has never coughed at my invoices, I think spending $5000 on a wireless link might exceed their comfort zone. I'm going to take all the info you guys have generously given me and going to research further a PowerBeam setup.

On a side note: Even though they drive me batty at times with all the changes, these are great clients and pay me T&M for this even though I am still learning. This means I get to learn on the job and put what I learn here at IPVM to practice. In turn, I keep my invoices fair and look out for their best interest. It's not often you get this kind of trust working both ways in client/contractor relationships. Very refreshing.