Long-Distance WiFi Device

[s/v Jedi]

some observations:

1. Ubiquity now has good, stable firmware for their M-series radios. If you buy now, I would certainly opt for the M-series, like BulletM2hp or/and NanoStationM2.

2. More and more MiMo systems are coming on-line. The Ubiquity M-series is MiMo too and the NanoStationM2 has a 2x2 antenna array. This means you get both horizontal and vertical polarized signals. My tests show a MUCH better signal on horizontal polarization for boats. My reasoning is both due to the horizontal water surface and the many vertical obstructions like masts, rigging etc.

3. For height of the radio/antenna, the Fresnell zone is the important factor... much more important than just the angles. As it would be crazy for a crew to use Fresnell calculation software while hoisting a radio, it's best to just hoist it a bit, select and connect to your choice of station and then find the optimum height by watching the signal/noise ratio.

4. The biggest mistake I see is using way too much power. -50 Db is hot... certainly hot enough. In that test that was linked you see something like a -38 dB signal which gets a 1.8% CCQ as punishment... the receiver is totally overloaded, greatly degrading the performance of the link. -50 to -65 dB might have taken it to 80-100% CCQ and a superfast link. This means the antenna had too much gain.
A good way to adjust output power is by lowering it until you loose the link and then adding 3 dB to the last working output level.

5. The M-series radios can be used with the free downloadable spectrum analyzer software from the Ubiquity website. This is the perfect, real-time way of testing your radio+antenna setup.

6. In a marina or anchored in stable tradewind, the NanoStationM2 is just perfect. Don't bother with a Bullet + Yagi etc. because the NanoStationM2 outperforms it by an embarrassing amount. It is also super easy to quickly mount somewhere and without any exterior metal parts. Those metal parts (antenna bracket, N-type connector) are the weak points of a Bullet on a boat.

If you swing and are close (within 500 meters) to the access point, the amazing PicoStation is fantastic. The M2 model for the PicoStation is out now.
The PicoStationM2 is also perfect for the on-board access-point/router.

The Bullet gets very little use nowadays aboard Jedi while it was our #1 before. We only use it for long range when omni-directional is a must.

Ubiquity rules... just forget your plans about that super-high-power WiFi adapter you saw in the shop or the boat show... if it's any good, there's a Ubiquity radio inside it.

cheers,
Nick.

[richardhula]

Will second all your comments Nick

Just maybe worth adding that Ubiquiti M series products are 802.11n compatible.

This protocol has backward compatibility with 802.11b & g still regularly used by cruisers. Yes there are small extra overheads when using M series equipment in compatibility mode (the only one available with Bullet M2HP) but this appears to be insignificant.

MIMO (multiple simultaneous spatially separate radio links) is for the first time included within the 802.11n spec, the top end of which can provide up to 600Mbps connections with 4x4:4 equipment (four antenna's providing four links at both ends). Some manufacturers offered their own hybrid MIMO systems with dual antenna wireless routers using 802.11g but it was never to a single standard.

Note the Bullet M2HP is not MIMO as it has only one antenna, so it can work at max MCS7 802.11n speed = 65Mbps when connected to 802.11n AP. As this is 10 times the best cable broadband speed you will probably get at home, its not really an issue.

[mtnsapo]

Hello guys, jsut got through this entire thread! good info, i am about to liveaboard and cruise, is the Ubiquity NanoStationM2 still teh best bang for teh buck? How hard would it be for a SEMI techinical guy to get it up and running?

Thanks!

Brian

[s/v Jedi]

Originally Posted by mtnsapo

Hello guys, jsut got through this entire thread! good info, i am about to liveaboard and cruise, is the Ubiquity NanoStationM2 still teh best bang for teh buck? How hard would it be for a SEMI techinical guy to get it up and running?

Hi Brian,

Yes it is There's new firmware out which makes it better again, although in station mode it was perfect already. You can download manuals from their website too.

ciao!
Nick.

[Gashmore]

Ubiquiti makes some of the best and easiest to set up wifi gear on the market but from what I have read about the NanostationM2 it would not be my first choice on a boat. First it requires 24V which might be hard to get on a 12V battery bank. Second it has an internal antenna and no external connection. The 10.4db gain internal antenna sounds good but unless you are tied to a dock that 55 degree horozontal beam width can be a PITA. Especially swinging on a ball or at anchor.

As to the choice of the M series it depends on your location. Around the US coast where /n is getting more available the M series would be appropriate but in the Caribbean the Bullet 2HP would be a better choice. N has hardly penetrated down there yet and the 2HP has slightly higher power and receive sensitivity in the b/g modes than the M series.

The Bullet 2HP has several advantages. First it runs on anything from 10 to 24VDC so you can hard wire it to the breaker panel. Second it has a sturdy waterproof N connector on top so you can screw a good 8dbi omni right on it with no coax loss. I have installed 12 so far for friends. One is 2,000 nm out of Cape Town enroute to Brazil today after a year long circumnavigatiopn and his Bullet has performed with out a single glitch in every anchorage. Got a report the other day from a boat on the Thornless Passage. He was getting a usable signal out of Myaguez from just east of Mona Island and had good coverage out of Cubebra from just off Fajardo almost all the way to ST. Thomas. That is a bit unusual but he mounted the Bullet and an 8dbi antenna at the masthead.

[mtnsapo]

cool, thanks fyi, i am on a cat, so no heel for me

brian

[thatboatguy]

Originally Posted by mtnsapo

cool, thanks fyi, i am on a cat, so no heel for me

brian

Brain, it's not the healing, it's the swinging.

At the moment we are anchored in the Indian River Lagoon and getting our wifi from a shore station about 2 miles away via a M2HP and a ravioli cantanna. I have a 2HP which is going to the masthead with a 15dbi Hawking antenna, probably next week some time. The cantenna rocks but is highly directional in nature and I do have to go into the cockpit to change it's orientation from time to time. I have solved this problem and will discuss it further when it's ready to deploy, but not any time soon. In the mean time, a masthead 2HP with an omni is still the best bang for buck and relatively easy to set-up and use.

On a side note, the Hawking antenna at around 4 feet will effectively turn our Ketch into a Schooner! Worth the extra money just for the coolness factor. heh heh

George

[Gashmore]

That Hawking 15dbi is expensive but a real killer. Good SWR through the band and a pretty wide e-plane for a 15dbi. As a plus you could add a reel seat and a couple of eyes and it would double as a standup rod.

[mlahrkamp]

Just a few thoughts (I appreciate that some of these points have already been raised):
1.
Using a Bullet2HP at full power with a 15dbi gain antenna can produce enough power to overdrive a hotspot when close in. I would encourage you to only use as much power as is needed so as not to ruin the band for all other local users. Additionally, to much gain (focusing) at masthead can result in not connecting to a nearby hotspot as it may be below your antenna's narrow beam. Lastly, FCC regulation limit the amount of EIRP power allowed to 36dbm. A 15dbi antenna coupled to a 1000mW radio will produce 45dbm effective output power - way over the limit. That's like being at a party and one person is talking through a megaphone disrupting everyone every time he speaks.

2.
The company I work for has now provided over 500 Bullet2HP based systems, many installed at the masthead. This is not for range, since 99% of the time you will be less than 1 mile from a hotspot. The real benefit is to get a clearer line of sight to the hotspot and to minimize the fresnel zone effect. Water absorbs, not reflects wi-fi signals.

3.
I don't believe there's much value in N mode devices since your ultimate Internet speed will be limited by the Internet service speed of the hotspot which will be only a few mbits/sec. Why connect at 108mbps when the hotspot only has a 3mbps ISP service attached to it. If you look at device specifications, including the Bullet2HP, you'll note that the max power and best receive sensitivity happen when operating at lower speeds and when in B mode, and not G or N.

4.
Certainly using a Nanostation2 can provide a better connection, but they are a directional antenna that requires aiming and setup. They are not a simple 'turn on and play' device. However, if one likes that sort of tinkering, they can be a better way to go for connecting to more distant hotspots.

5.
Wi-Fi is two way radio. You can boost the power that you transmit, but unless you also increase the receiver's sensitivity, it won't do you much good. By way of example: a 5W handheld VHF will not transmit as far as a 25W VHF. If you add a 25W booster to the handheld, you're now on a level playing field. However, further increasing it to say 100W will make it broadcast further, but you'll still only hear at the range capability of the other 25W radio. What ends up happening is that you walk over transmissions of radios that should normally be outside your range and end up disrupting others.

Anyway, just our experiences.

Mike

[thatboatguy]

All good points Mike,

But your vhf analogy misses one thing. The hand held radio with a booster still has the rubber ducky antenna for recieve. As you point out, receiver sensitivity is just as important as transmitting power since it's a two way street. I can't do anything about the shore-side equipment, but I can make certain I have the best radio gear aboard and the versatility to make it work the way I need it to. Right now the best antenna onboard is a home made cantanna that has to be periodically hand aimed. And I'm talking receive here, not power out. The Hawking performs nearly as well, stood up in the cockpit with lots of stuff in it's way. I'm pretty sure it's going to perform very well at the mast head but I wont really know until it's up there.

I also know where that power setting is in Air OS... heh heh And your point is well taken. That can be adjusted downward with a few clicks as needed.

I will make this one general sweeping statement concerning wifi. It *never* pans out in reality the way it does on paper. Experimentation is the only way to find out what works and what does not. For that reason, people like you with tons of in the field experience (and Bob at Island time) get my ear, every time.

As for ending up "looking over the top of a close signal". I have not yet been able to make that happen. I'm building pretty reliable (if anecdotal) evidence of that as we always have an Alpha running in the cockpit hooked up to the nav-computer and not once in a thousand miles of ICW has there been a time when it saw a station that the mast head omni did not. This is yet to be seen with the Hawking.

George

[Gashmore]

Mike is 100% correct. It is very easy for a 2HP and a high gain antenna to over power a hotspot. Fortunately the Bullet has a nice little slider in the software to adjust maximum power and I always explain how and why to adjust it to meet conditions.

Unlike Mike's company I do this as a hobby and work for free sailing time and beer. Most of the friends I put systems together for are long term cruisers who spend little time in marinas and need 3 to 5 mile range. Often the only available signal is an open AP in a villa somewhere up the hill. For that the 2HP and an 8dbi antenna is more than adequate.

An omni antenna radiates in a doughnut pattern. The more you flatten the doughnut the higher the gain. Beam width is defined as the angle where gain drops by half (from 15 to 12db for example) so even a very flat doughnut will pick up a signal if the access point is close. OTOH, at long range the signal strength will drop considerably as the boat rocks. 8dbi for monos and 12dbi for cats seems to be the happy medium.

Especially in the Caribbean, speed between the boat and shore means very little if Cable and Wireless doesn't keep their strings tight and the tin cans clean on the back haul side. Except for St. Martin 90% of the down island traffic flows on a single fiber optic cable operated by ECFS that comes ashore at Brewers Bay, Tortola. From there it is routed to either Bermuda in the CARAC cable or Puerto Rico. The ECFS cable is nearing capacity so don't expect the response times you get in the US.

A side note if your cruising ground is the Caribbean. C&W and Digicel both have 3G on all the former English possessions and one $90/month account covers all islands with no roaming fees. Several of my installations on crewed charter boats use the Bullet with a Cradlepoint or similar 3G router. When out of wifi range they use the cellular system.

Another thing to watch is that Digicel and C&W are in a race to add WiMax. Not enough info yet to determine if it will be a viable option for boats but it is something to keep an eye on.

[s/v Jedi]

The NanoStationM2 does NOT need 24V. All regular Ubiquity radios have the same power requirements and the NanoStations are just the same as the Bullets on that front. I power it from 12V house bank using a $10 DC POE injector.

All Ubiquity M-series radios are HP (High Power) radios. They are also the newer generation and support 802.11b/g just fine.

The (old) Nanostation2 (instead of M2) does have an external antenna connector but for use on a boat this makes no sense as a Bullet would be the preferred option then.

The NanoStation has an internal MiMo 2x2 directional antenna array. MiMo improves link quality even when the "other side" is non-MiMo and it certainly does not have any negative effect when the other side is non-MiMo. What is more important: a directional antenna provides a great improvement in signal-to-noise ratio compared to omni-directional antenna's. The NanoStationM2 antenna opening angle is in the 55-60 degree range which means that interference from the other 300 degrees around you is suppressed by design. The other side of it is that you must point it to the access-point used. This still means that you can swing 55-60 degrees at anchor before you get a -3dB attenuation of the radio signal. If you are within a couple hundred yards of the access-point, you can swing much more; I have seen a link between a netbook with internal WiFi and a NanoStationM2 that was 500 yards away and at an 80 degree angle and it worked just fine. I think cruisers are smart enough to point an antenna and would't call that "tinkering".
Also, when both sides have MiMo antenna arrays the results are lifted by another magnitude because all boat-shore connections I have seen get a much better signal level on horizontal polarization; 6-9dB better (which is 2-3x as "loud"). So, it's not just the 802.11n mode; the antenna design is more important.

The sturdy N-type connector of the Bullets is also their weak point. I have had it completely seized to an antenna after a couple of months use. The NanoStations have no metal parts at all and are also much more resistant to water ingress (Bullets need tape). Also, the antenna's that are used with the Bullets have metal parts too and often just galvanized mounting hardware with dramatic results of corrosion... needing acid cleaners to get the rust spots from the fiberglass etc. I've seen it all happen.

I have about 10 Ubiquity radios aboard so yes I'm weird. When we move to a new spot and I want WiFi, the NanoStationM2 is my first choice and I'll try to find a way to use it. Only when I swing wildly at anchor (like with the tides or morning/afternoon 180 degree windshifts) I will use a PicoStation (which is now also available as the new M-series) for short range (up to 500 yards or so) or the BulletM2 for longer range. This shows that every radio has it's own advantages/disadvantages but also that the NanoStation is my preferred radio.

I think that most of the people that state that the Bullets are the best models are also people that only have a Bullet. How can you make such a statement when you never tried a NanoStation or PicoStation?!

I also think that most of the people that state that their "brand X" is the best, never had an Ubiquity radio or their "brand-X" has an Ubiquity radio hidden inside (I have seen that on some $500 marine WiFi products!).

cheers,
Nick.

[mlahrkamp]

Some background:
FCC regulations allow no more than 36dbm for point-to-multipoint communications. That means a hotspot and any number of clients.
FCC regulations require 1dbm power reduction for each 3dbi gain starting at 6dBi for point-to-point. That means a directional pointed at a single other directional, typically found on communications towers.

Now in looking at the Ubiquiti website, they categorize products as Airmax ISP Solutions (NanostationM, BulletM, etc) and they also have another category called 802.11 a/b/g ISP Solutions (Nanostation, Bullet, etc).

Looking at specs for:
Nanostation2 - 26dBm power output+10dbi antenna gain = 36dBm
PicoStation2HP - 29dBm power output+6dBi antenna gain = 35dBm
Notice how they stay within the limit?

Now look at:
NanostationM2 - Built in ~11dBi gain antenna and a 28dBm max power output. The antenna gain is almost 6dBi of additional gain beyond the base 6db, or 2 steps of 3dBi. If following the FCC regulations and using it in a point-to-point system, one needs to reduce a 30dBm max transmitter by 2dBm=28dBm. Notice this is what their max spec is. But when connecting to a point-to-multipoint system, the power would need to be further reduced to 25dBm (36dBm max allowed power-11dBi antenna gain=25dBm).

So what's happening is that the NanostationM2, intended to be a carrier communications system product and operating on the same unlicenced 2.4ghz band is being used as a general wi-fi receiver and when operated at full power is putting out about 39dBm (28dBm + 11dBi). It therefore has the potential for pushing the signal a bit further. Using my previous example: you're at a party and one person constantly talks in a loud voice. There is some disruption but not nearly as much as a megaphone 15dBi antenna coupled to a fully powered Bullet2HP.

In looking at the spec sheets for the Pico2HP and the PicoM2 both have the same output power and 6dBi antenna and neither can exceed FCC power limits, either point-to-multipoint or point-to-point.

Regarding tinkering: Using a directional device such as the NanostationM2 does require knowledge on the part of the operator in order to be a good citizen on a hotspot network. The antenna needs to be pointed in the general right direction and consideration has to be given to the amount of power being driven. This is not a simple 'flip a switch and you're on' system. And I suspect that without proper knowledge as to the various intended product roles, the average individual who just wants to surf may not operate the products correctly and produce Wi-Fi polution through overdriving, etc. Given that a Bullet2HP coupled with a 6dBi omni is capable of delivering 5+mile connectivity, there really shouldn't be any need to go to the extra complexity of directional. What's more important than power is an unobstructed line of sight and being out of the fresnel zone. Height matters. So, instead of driving 36+dBm directional at rail level, try omni at masthead with reduced power - assuming you follow FCC requirements, I guarantee you'll get equal or better results.

BTW: We use NanoStation2, Pico2, Pico2HP, Bullet2HP, and possibly soon BulletM2. Current testing shows performance with it that is a bit less than the Bullet2HP - again, that can be seen in the spec sheets for both output power and receiver sensitivity. We haven't yet found a solid benefit that it offers over the Bullet2HP considering a hotspot's ISP connection is usually just a few mbps and can't take advantage of the higher speeds the BulletM2 offers.

Regarding antennas: We tested 6, 9 and 12dBi gain. When properly driven (30dBm+6dBi, 27dBm+9dBi, 24dBm+12dBi), they all connect. Difference then is physical size and whether you wish to be a good netizen.

Sorry for the long post.

Mike

[s/v Jedi]

Originally Posted by mlahrkamp

NanostationM2 - Built in ~11dBi gain antenna and a 28dBm max power output.
[...]
But when connecting to a point-to-multipoint system, the power would need to be further reduced to 25dBm

This is correct but why would that be a negative for using a NanoStationM2 ? Also, the US has one of the highest allowances of output power in the world, meaning that, to stay within local regulations in most other countries, you have to reduce power for almost every Ubiquity product.

So what's happening is that the NanostationM2, [...] is being used as a general wi-fi receiver and when operated at full power is putting out about 39dBm (28dBm + 11dBi).

And how is that different from Bullet2HP's at max power coupled to 12-15dB antenna's like so many cruisers use? All Ubiquity radios are semi carrier-class devices.

Regarding tinkering: Using a directional device such as the NanostationM2 does require knowledge on the part of the operator in order to be a good citizen on a hotspot network. The antenna needs to be pointed in the general right direction...

I think you underestimate the brain power of the average cruiser. Even after loosing many brain cells due to rum consumption, they still manage to point stuff in generally the right direction

and consideration has to be given to the amount of power being driven.

which is just as difficult as for a Bullet. They use the exact same software and thus the same slider for power output. Even when everybody would couple a Bullet to a low to medium gain antenna so that they stay within legal power limits, this can still be way too much power if they are close to the access-point, with all the same nasty consequences you list like WiFi pollution through overdriving etc.

Also, an Omni with a Bullet at the masthead, while providing better Fresnel zone clearance, does not match well to a close-by accesspoint with a high-gain omni which is mounted much lower, because it will be outside the radiation pattern of that access-point antenna. In this case, even a PicoStation at rail level will outperform it.

BTW: We use NanoStation2, Pico2, Pico2HP, Bullet2HP, and possibly soon BulletM2. Current testing shows performance with it that is a bit less than the Bullet2HP

How can you test that when you don't have the BulletM2 yet? My NanoStationM2 outperforms both my Bullet2hp and BulletM2 coupled to an 8dBi omni hands down. I can't detect any difference between the Bullet2hp and the BulletM2 when I put it in 802.11g mode. When I allow it to go to 802.11n mode I see performance improve. This is with the accesspoints in regular operation with about 30-60 other stations active in the marina & hotel.

I just finished the installation of a WiFi system in the marina (+ hotel) here in Panama with 9x 2.4GHz and 2x 5GHz access points. I used PicoStation2hp (all at low power levels), NanoStationM2, NanoStationM5 and BulletM5 radios. I used a tripod mounted NanoStationM2 on top of my SUV while driving around the complex to test coverage etc. When we found the most troubled spot in the complex (behind the travelift :-) I could not establish a solid link with a Bullet at any setting while the NanoStationM2 went up to 117Mbps with 80% link quality without a single drop-out (because it had a good signal quality on horizontal polarization and only a marginal one at vertical polarization... the strange thing was that the Bullet didn't improve when turned 90 degrees to horizonal polarization).

For me this proves that a NanoStationM2 is the best choice for a marina environment.

cheers,
Nick.

[mlahrkamp]

Originally Posted by s/v Jedi

Even when everybody would ... stay within legal power limits, this can still be way too much power if they are close to the access-point, with all the same nasty consequences you list like WiFi pollution through overdriving etc.

Not quite true. The AGC circuits with the radios are capable of adjusting when the signal is within legal limits. By way of example, a 25W VHF docked next to another boat could call that boat (masthead antennas separated by maybe 20'), and they still work and sound OK. That's because the radios are designed to handle that level of power being put into the receiver. Boost one to 100W let's say, and I'll guarantee it'll saturate the other receiver.

Originally Posted by s/v Jedi

Also, an Omni with a Bullet at the masthead, while providing better Fresnel zone clearance, does not match well to a close-by accesspoint with a high-gain omni which is mounted much lower, because it will be outside the radiation pattern of that access-point antenna. In this case, even a PicoStation at rail level will outperform it.

Professionally installed access points will always be mounted up high and point down on the boats. This is to provide as much clear line of sight as possible to all boats. At masthead and with a 6dBi gain antenna, you'll pick up ground based access points no problem. I agree with you when something like a 12-15dBi antenna is up there. Of course if you're that close, you don't need a booster and your laptop will connect just fine on its own.

Originally Posted by s/v Jedi

How can you test that when you don't have the BulletM2 yet?

We've actually had them for almost a year now. We're just not offering them. Testing continues when time permits.

Originally Posted by s/v Jedi

My NanoStationM2 outperforms both my Bullet2hp and BulletM2 coupled to an 8dBi omni hands down.

I agree that the NanostationM2 will outperform - it should - it's putting out more power and the signal is highly directed as compared to an omni.

Originally Posted by s/v Jedi

I can't detect any difference between the Bullet2hp and the BulletM2 when I put it in 802.11g mode. When I allow it to go to 802.11n mode I see performance improve. This is with the accesspoints in regular operation with about 30-60 other stations active in the marina & hotel.

What performance is improving? Is it the speed? As I'd indicated, you're still limited to the speed that the ISP is providing to the hotspot. I doubt if connecting in N mode at 100+mbps makes a bit of difference as when connected in B mode at 5.5mbps. Looking at the specs for the Bullet2HP, when using G mode and 24mbps, the power output is 28dBm (630mW). Increasing speed to 54mbps sees the output power drop to 23dBm (200mW). Let's look at the receiver sensitivity: at 54mbps it's only -74dBm while at 6mbps, it's at a very good -94dBm. That's a heck of an improvement in receiving! Same thing for the BulletM2 (assume a 40mhz channel): at 13.5mbps, the power output is a respectable 28dBm (630mW) but at 135mbps, it's down to 22dBm (158mW). That's a big change! Same for receiver sensitivity: -96dBm at 13.5mbps dropping to -74dBm at 135mbps. For B mode at 1mbps, the power output of a Bullet2HP is 29dBm (800mW) and the receiver sensitivity is an excellent -97dBm. Sorry for the teckie talk. Observation: You may find that your best connection may just be a lower speed.

Originally Posted by s/v Jedi

When we found the most troubled spot in the complex (behind the travelift :-) I could not establish a solid link with a Bullet at any setting while the NanoStationM2 went up to 117Mbps with 80% link quality without a single drop-out

I agree, higher power with a directed beam. Like comparing a bare light bulb to a flashlight - and any small flashlight can outshine a bare 100W bulb - on a given spot.

Originally Posted by s/v Jedi

For me this proves that a NanoStationM2 is the best choice for a marina environment.

This comes back to the amount of effort that's needed to make it work. It needs to be pointed and power levels adjusted so that the hotspot doesn't saturate. From our experience, most people just want to flip a switch, select a hotspot and surf. But power to you Nick for going for the full gusto. If only all of our customers were as energetic.

Cheers,
Mike