NanoStation vs Canopy

[jefftebo]

I'm new to the nanostation. I have an all Canopy network and want to try to add the Nanostation5 to a few of my towers. Any advice from anyone would be helpful. How do you manage several towers and AP on each tower without timing from a CMM or CTM? I have billing thorugh Freeside and I've just purchased a Gatespot for a few mobile home communities to handle billing and authentication. Will the nano's be as reliable as the Canopy system I have in place?

[WHT]

Short answer...They won't work and no.

The Canopy is a proprietary system and won't support other 802.11b/g radios. So you'll have to install some traditional radios.

Will an all UBNT system meet your expectations of an all Canopy system?
* Canopy AP array and thirty stations = $25,000
* UNBT array and 30 stations = $2,700
Most likely, no.

A properly deployed $5,000 UBNT system could give a good run against the Canopy.

YMMV - You Money May Vary

[m2806]

Hi all,

I work with the two systems in 2 different WISP, the UBNT performance is much higher that the canopy (in a very high noise floor area) but doesn't have a CMM ( Cluster Management Module ) witch is basically an advance switch with a GPS for AP synchronization (i use a 3com switch to connect 6 NS5 to make a 360degre AP and work very well)

Every canopy pmp100 AP can handle up to 200 cpe( as motorola says) against 35 on NS5 ( pmp100 is the standard canopy point to multipoint)

Latency on pmp100 is around 7ms up to 10ms for 2ms on nanostation5 (pmp200 best know as canopy advantage is 2ms )

canopy maximum throughput is 14mb/s vs 24mb/s for ns5

Canopy is Direct sequence and NS5 is OFDM ( canopy have the pmp400 with is OFDM but very expensive)

NS5 can be AP AP-WDS Station Station-WDS but not canopy witch have AP( Access Point )SM ( Subscriber Module CPE) and BH ( Backhaul)

NS5 cost 79usd and canopy around 150 to 200 for the SM(CPE ) 1000 to 1500 for 1 Access Point and 1000 to 1400 for a CMM

Canopy have a Prism server for network monitoring ( very expensive licenses )but i use Dude NMS from mikrotik witch is free

Both can work in bridge and NAT mode but only NS have firewall and p2p blocking system based on ipp2p.org

NS5 has Led to help you pointing and for indication , canopy need an external headset to point with a beep system :S

With all this data i think that the choice is clear : Ubiquity ROCKS ;)

[dGlueguy]

A couple of corrections
A CMM ( Cluster Management Module ) which is basically an advance switch with a GPS for AP synchronization (i use a 3com switch to connect 6 NS5 to make a 360degre AP and work very well)

Not really. The CMM's main function is to synchronize the APs so that they will not interfere with each other. A properly timed Canopy cluster can have up to 12 APs on the same tower. There are other GPS timing devices available that allow you to use any switch you might like


Latency on pmp100 is around 7ms up to 10ms for 2ms on nanostation5 (pmp200 best know as canopy advantage is 2ms )

And the Canopy will stay with that latency even under heavy load. The NSx latency will vary wildly under load. Try the same test with 50 or 80 or 100 CPEs attached.


Canopy is Direct sequence and NS5 is OFDM ( canopy have the pmp400 which is OFDM but very expensive)

Not true. Canopy PMP100 is FM (Not DSS). Very tolerant of noise. Can handle a carrier/interference ratio as low as 3db. NSx needs at least 10db under similar conditions.


Canopy have a Prism server for network monitoring ( very expensive licenses )but i use Dude NMS from mikrotik which is free

Prism is not required. Other monitoring solutions work just fine (Nagios, Cacti, PRTG, Dude, etc. etc.)

[rconaway]

dglueguy is correct. Each product has advantages and disadvantages. Given that, I can't imagine any scenario where having 80-100 users sharing 7Mbps is a good thing for the price. Although we carry Motorola, I can usually design a Ubiquiti based system capable of functioning as well or better for 1/10 the cost.

If you are limited to a single tower covering a high-density area, then Motorola may be your only option. However, it's not hard to get at least 8 UBNT base stations on a single roof with no interference and 13 with a little more work and stay completely legal. That gives you up to 390 possible users with less than $1100 in radios at retail.

This is my hot button.

[jetout]

i think Nanostation5 better

[rconaway]

I think if anyone has installed a Motorola system, they should jump in here. I have friends that have 1000 users or more on their Motorola systems and they swear by them. I will say that the Motorola will have a lower support time than the Ubiquiti. The management firmware in the Motorola is definitely stronger. The CPE side in Ubiquiti still has a lot of manual adjustments that can make big changes in performance. Is it worth the price difference though? That's the question.

[dGlueguy]

OK, maybe we don't have 1000's of Canopy SMs installed, but way more of them than Ubiquity. And pound-for-pound, dollar-for-dollar, we have fewer management/support issues with Canopy.

For us, it all boils down to the cost of a truck roll. We have almost zero call-outs/truck rolls on the Canopy portion of our network. It just works. 95% of the calls we get are the subs' SOHO router, be it Linksys, D-Link, or Netgear. And that is usually handled with the unplug/wait 15 seconds/plug it back in routine.

In 5 years of operation, we have had to replace 4 Canopy subscriber modules; 2 were from trees falling on them, and 2 were from PG&E power surges.

We also operate near other operators that use 802.11x gear of one kind or another. I know from talking to the subscribers that have switched over to our service (from them), that our network is way more reliable than the other ones; and more consistent.

[rconaway]

I understand. It's could be a pay me now or pay me later scenario. From what I understand, the ROI on the Motorola gear installed is about 9 months for my friends system but you are right, they have no problems and it's highly reliable. We have some NS equipment that has been out for almost a year and we also have had no issues with installed equipment so far (except for the woodpecker issue yesterday). Our ROI in the same scenario would be about 5 months.

However, take into account the cost of the base stations in the ROI and that number changes dramatically. Considering that a 5 radio NS5 system costs $400 in radios to deliver 125mbps versus a similiar Canopy system at half that bandwidth and the Ubiquiti wins hands down. Throw in both systems under full load and the Motorola wins. This is why us engineering types get paid the big bucks.

[dGlueguy]

I'll go along with that. I surely understand the rationale for inexpensive equipment; as long as it does the job. However, if you look at the cost of deploying a cell site, those operators would slather all over themselves to get an access point at the cost of Canopy. Perspective I guess.

After we looked at it, we determined that the cost of the equipment is not the most important factor; even though it stares you in the face every time you buy a batch. The real costs are support and truck rolls; they add up really fast, and are pure negative.

We operate on a bare-bones staff, and the support load is minimal, to non-existent.

I'll add that we backhaul through a Dragonwave Airpair. Awsome devices, and they are pushing the equivalent of an OC4, full duplex. The cost of one pair of those is an order of magnitude more expensive than a Canopy AP, and then some.

[mwarren]

Motorola Canopy is about like Windows. You have to upgrade these things constantly and the some firmware releases are good some are bad. They still haven't figured out the true transmit power and each firmware change reflects this. The speeds are an aggregate of 3.3 on the 900 Mhz spectrum with a 70/30. Typically you will see 1.5 dwn 500 up. Very suseptable to localized interferance. Very expensive to extend range. 2.4 is better but again it is very expensive to extend the range. With that being said the timing and color coding are great features and proves to over power other like equipment. Using WIFI with WDS an individual home or business could act as a low cost repeater station dramatically improving your distance as well as mobility.

[dGlueguy]

You have to upgrade these things constantly and the some firmware releases are good some are bad.
That's really no different than any other vendor. We've gone through the same thing with Tranzeo, Trango, and Ubiquity. Calling it unique to Motorola is a bit like the pot calling the kettle black.

The better part of it though is that a significant portion of the hardware is implemented in FPGA, and we've done significant upgrades through down-loadable hardware upgrades.

They still haven't figured out the true transmit power and each firmware change reflects this.
Not sure what you're speaking of here, but because of the FM nature of the design, "power level" loses some of its meaning, especially compared to the OFDM radios. We have not see this as an issue.

The speeds are an aggregate of 3.3 on the 900 Mhz spectrum with a 70/30. Typically you will see 1.5 dwn 500 up.
Our aggregates on 900 are more like 4.4Mb. We are getting a consistent 3Mb down, and 1.4 up on our 900 Canopy.

But each to their own. We like our Canopy equipment a lot, and still have a lot fewer issues with it than any other equipment we've used.

[rconaway]

I have to tell you guys, after we figured out the ACK and bandwdith relationship on 3.2.3, our support on the Ubiquiti radios is also now pretty much zero. On the other hand, Motorola doesn't give you anywhere near the flexibility of field design that Ubiquiti does. That means Ubiquti goes in every location a Motorola can go in but the reverse can't be said. It also means it can be done at 1/30th of the cost or more. I have 260 users, about 60 I think on UBNT, and some users are 5 radios over 2-3 hops before hitting the authentication server and Internet. My latency is 6ms tops under normal load or about 1.5-3ms per radio. My SkyPilot radios at $5000 are 8ms per radio.

I no longer think that there is a comparison with Canopy. If the system requires that I move AP's out from the central location because some users can't see it, then Ubiquiti has a solution. We stick up a backhual, add another omni or sector, and continue up the terrain. Can't do that with Canopy for less than a couple thousand bucks or more. On our biggest system, we have 2 main bases with an NS5 connecting them, and 5 remote base stations to get around tree issues. The system is totally stable and performance is excellent. Most of our clients are now 3.3.1 and all of those were updated remotely.

[mike95826]

You are all right. This is the same debate that goes on in any technical field. Price vs Features vs Reliability vs Customer type. Another example in a closely related field also involving Motorola is two way radios. Motorola also builds "high end" "high price" "proprietary" equipment for THAT market. The questions and some of the answers are the same. Do I need to buy Motorola to build a usable, reliable system...No. Who is the customer and will they need the "features" "slightly more reliability" knowing that the price tag will be that much higher? It depends upon who the customer is. If the customer is a fortune 500 company or a large government agency that will use the system for public safety then a few dollars would "brake the bank" but if you are a new small business that just needs to use radio to dispatch a few trucks then it may make a difference to the bank account.
In my opinion, A start up WISP that provides service to mostly homes would be best to start with a lower cost product and maybe when the business becomes more profitable (when the customer count warrants) then go to the "higher end" more proprietary equipment.

Just my $0.02.... before inflation...

[xphat]

I agree with everyone here basically.

Ive worked with Canopy Systems and now im using Ubquiti, and i haveto say when building out a system for market penetration you really cant get better than ubnt.

All you do is just to look at where your market is expanding, find a suitable location, put up a new AP there and there you go, added capacity, added range, excellent QOS to everyone... move on.

We started doing this WISP thing a couple months ago and believe me ive learnt a helluva lot.. its all about designing your network to do what you want it to do and to go where the customers are. And when it comes to the best bang for your buck.. you really cant get better than ubnt.