Long-Distance WiFi Device

[richardhula]

Interesting results posted here regarding test of 9 & 12dB gain omni antenna's on a Bullet installation.

Too much emphasis seems to be placed on the number of AP's a particular installation can "see", but what's the point if they are too weak to connect to.

[bene505]

Originally Posted by richardhula

Interesting results posted here regarding test of 9 & 12dB gain omni antenna's on a Bullet installation.

Too much emphasis seems to be placed on the number of AP's a particular installation can "see", but what's the point if they are too weak to connect to.

That was my test. I hadn't gotten to posting it here yet. Thanks for linking it.

Regards,
Brad

[bene505]

Originally Posted by richardhula

Interesting results posted here regarding test of 9 & 12dB gain omni antenna's on a Bullet installation.

Too much emphasis seems to be placed on the number of AP's a particular installation can "see", but what's the point if they are too weak to connect to.

That is my test. I hadn't gotten to posting it here yet. I usually post to the 3 forums I frequent (anything-sailing.com, sailnet.com and cruisersforum.com) as a practice. Thanks for linking it.

CruisersForum has strict rules on how long you can edit your posts. I got hit with that time limit on this reply just now. (Mods, please delete my previous short reply to Richard.) The test results took me hours to post and I actually took them down and reposted them when I needed to change some of the analysis. I also find that I tweak grammar even a day later. So I typically post to CF last. (That, and not having any time limit is just more freedom.)

It may seem like too much emphasis simply because listing 50 access points takes up a few graphics. There are 3 reasons to show the access points that you can't currently connect to, during an antenna test:

1) signal strength isn't the only factor. In a lower noise environment (away from suburbia) a weaker signal will get you connected. And there were/are a lot of access points transmitting/interfering on the same channels. I mentioned that -75db was my arbitrary cut-off for determining if you can connect to an access point. The needed signal strength could be higher or lower depending on where you are.

2) so people can get an idea of just how many wifi access points are "out there" to connect to. Coupled with the capability to move the boat if you need better connectivity, knowing there are 36 more access points nearby gives you some confidence that if the first 14 don't work, you have some options. In my case, I already pay for service from one of the cable companies that have access points scattered about. When hurricane Earl passed close by, I moved the boat to the upwind side of the lake (and deployed a ton of anchors) and happily connected through that access point. It is nice to know they are there.

But most importantly, 3) it is the only good measure of the affect of the antenna's beamwidth during the test, and beamwidth is a major factor when choosing an antenna. Some of the weaker signals were within the 9db antenna's beamwidth and outside of the 12db antenna's beamwidth -- that affect would be more prominent if you are in hilly terrain or in an anchorage with some swell. The access points I couldn't "see" with the 12db antenna (but could see with the 9db antenna) are a clear demonstration of that beamwidth effect. When choosing, you'll want to think about how much rolling and pitching you do at anchor. And when you are rolling the most is perhaps when you want to connect the most -- to see where that storm is going on NexRad radar, or because it's poor weather for a shore excursion (or trip to the cafe to get urgent work done), or because you want to check-in on CF.

That's not to say those see-but-can't-connect metrics are the most important in an antenna comparison test. They are simply a big part of the decision on what to get. For me, where rolling/pitching at anchor are not a big factor, the 12db antenna let's me connect to access points that are 58% farther away. For others that have any consistent rolling, beamwidth would be a more important factor. And you just wouldn't get a sense of that part of the decision with a simple signal-strength test.

Regards,
Brad

[bene505]

Richard,

Saw your post over there just now. Thanks for the insight about beamwidth. At first I wasn't sure what was causing the variation in seen APs, since the terrain there is somewhat flat.

Regards,
Brad

[thatboatguy]

Interesting antenna discussion. Thanks to all. One thing that seems to be indicated but hasn't been mentioned in this current round of discussion is that it's very handy to have *some* kind of directional antenna for determining just where that almost-usable wifi signal really is. A real world example would be Nassau Harbor. We anchored near the east entrance and had a number of almost usable signals. I used the yagi, hand held, to locate one on shore, get our essential emails in/out and the next day we moved closer to it for a solid connection using the omni. One nice thing about the yagi I have is that is screws right to the top of the bullet (although we didn't have the bullet on that trip) and acts as a handle! One crew member keeps it aimed at the target ashore while the other works the email/weather etc.

[W32PAMELA]

Brad,

It's good to see some published antenna data. I've attempted to make some comparisons of a 12dbi antenna and an 8 dbi antenna myself and have ended up with data I felt was inconclusive.

I've found that wifi signals vary considerably from moment to moment, especially the weaker signals that boaters encounter when away from urban areas. This is often great enough to make an instantaneous signal strength readings meaningless.

The readings on the "Main" tab and scan window in the Bullet 2HP are snap shots rather than "live" real time readings. For instance your screen shots show the 9dbi antenna had a signal strength of -42 but the scan window showed a value of -59 for "LakeClub". Likewise the 12dbi antenna had a signal strength of -38 on the "Main" tab but a scan window reading of -71.

I've tried using the "antenna alignment" tool to get some data on how much signal strength was varying and found that there was too much overlap in the signal strength of the 2 antennas to say if a 12 was superior to an 8. I've found a similar problems in testing for maximum distance from an AP with 2 antennas of different gain.

[bene505]

Bob,

Nice web site you have. The picture of the complete setup with all the needed wires and the optional wireless router is a good summary of what's needed.



I too found some variation in the detected signal strength from moment to moment. I saw a 2bd difference from maybe 15 minutes earlier. To overcome this during testing I took my measurements for both antennas as close together in time as possible, given that I had to quickly bring one antenna down, swap it for the other antenna and get back to the salon to get the screen shots. I didn't see any big variations, but what you say has merit.

Of course, the way to overcome this when testing is to do a bunch of sampling and average-out the results. Truth-be-told, if I was to do more testing it would be to answer the "what height above the deck" question, to see if there was any real benefit in going all the way up the mast, or if it's good enough to mount a Bullet on a push pit, davit or arch.

The locked-in signal strength was significantly stronger than what was listed on the scanning results page. I was wondering if some astute person would notice that. WiFi uses spread spectrum, so I figure that we are seeing the results of processing gain. From some reading just now, it looks like the 802.11 standard specifies a minimum gain of 11db. We are clearly seeing much more than that.

Regards,
Brad

[YOGAO]

Surely we have a maths wizard on CF that could show the "field of vision" for 9 and 12db antennas at some standard distance, like a mile. Then you could determine how raising the antenna 50 - 60 feet above the water would help/hurt.

My offhand guess would be that the major advantage to raising the antenna is clearing shore-based obstacles and the fact that most on-shore APs are going to start out significantly above sea level. We typically find that raising the antenna 15 - 25 feet helps in this regard.

Fair Winds,
Mike

[bene505]

Originally Posted by YOGAO

Surely we have a maths wizard on CF that could show the "field of vision" for 9 and 12db antennas at some standard distance, like a mile. Then you could determine how raising the antenna 50 - 60 feet above the water would help/hurt.

My offhand guess would be that the major advantage to raising the antenna is clearing shore-based obstacles and the fact that most on-shore APs are going to start out significantly above sea level. We typically find that raising the antenna 15 - 25 feet helps in this regard.

Fair Winds,
Mike

The 12 db antenna has a beamwidth of 7.5 degrees. Assuming the antenna is vertical, that would be up 3.75 degrees above horizontal and down 3.75 degrees below horizontal. The sine of 3.75 degrees is 0.0654. That's 65 feet (total of 130 feet adding both up and down) every thousand feet. An access point 1 mile away could be 345 above the antenna and you'd still be within the beamwidth.

The 9db antenna has 12 degrees of beamwidth, IIRC. 208 feet total (up and down from horizontal) every thousand feet. So an access point 1 mile away could be (1097 both up and down, or) 550 feet above you and you'd still be within the beamwidth.

Note that beamwidth is defined as where the signal is half as stong as the center (stongest) direction.

Regards,
Brad

[W32PAMELA]

Originally Posted by ;bene505

The 12 db antenna has a beamwidth of 7.5 degrees. Assuming the antenna is vertical, that would be up 3.75 degrees above horizontal and down 3.75 degrees below horizontal. The sine of 3.75 degrees is 0.0654. That's 65 feet (total of 130 feet adding both up and down) every thousand feet. An access point 1 mile away could be 345 above the antenna and you'd still be within the beamwidth.

Brad, I'm far from an expert in antenna design but I don't believe you can mount 2, 12 dbi omni's a mile apart with a difference in elevation of 345' and get anything close to the 9 to 12 db gain you would expect if your statement above is correct. If the beam width worked as an ever expanding cone then we wouldn't have any problem using 6dbi VHF whips on our masts. As I understand it these have around a 25 degree beam width so when heeled, vessels a few miles from us would still be within the -3db beam width and overall we would be better off than if we used the usual 3dbi whip.

[richardhula]

Brad, I have to agree with Bob's comments regarding beam width not being an ever expanding cone but more of a lobe.

Below is polar diagram of the 12dB 2.4Ghz omni antenna which I use. Note how it favours above horizontal which suits my lowish pushpit mounting.



Below is another which shows it better

[Laureat]

I am looking very siriously to buy wifi antenna adapter for my Sailboat in Nassau Bahamas. From Ca world wifi I found that product aou2410 looking interesting, and price is very afordable. Anybody using this wifi yet. Thank you for help
AOU2410 1 WATT USB WIFI ADAPTER

[CapitalistNYC]

I use a service called Clear. They are a standard residential/commercial internet provider with an interesting on the go device. The 4g internet signal is pulled from the cell towers to the device and fills the boat with wifi for many computers They are not launched in all US markets yet but you will get 3g nationwide (still fast enough). Not a solution outside us.

I get a cell signal miles off shore NYC/LI and can download charts or surf web from the boat. The service costs under 60/month and works great at starbucks, the car and at home.

[hellosailor]

"The 4g internet signal is pulled from the cell towers to the device and fills the boat with wifi for many computers "
That's called a "cellular Wifi hotspot" and several cellular companies sell them. Several cell phones can also provide that service, usually at an extra monthly charge.

An interesting note is that the ITU, who regulate international radio matters, say that there are currently NO CELLULAR SYSTEMS IN THE US THAT MEET THE DEFINITIONS FOR 4G SERVICE.

Among other things, 4G requires simultaneous voice and data connections, which Sprint and Verizon do not support at all. And then there's a matter of data speed, which apparently none of the US carriers support.

So if your carrier is offering you "4G" or even "3G"...apparently most of them are lying, as defined by the ITU definitions that govern the whole field. Which shouldn't surprise anyone, after all, if a cellco in the US moves their lips at all...Ayup, they're lying.

[CapitalistNYC]

Thanks HS

I cant speak to the credibility of Clears 4g claim. I have used many usb and spot services from wireless companies in the past and this is not the same. It is faster than most high speed services i have had for both home and business and much faster than any wireless company solution. Down and up speeds are comparable to verizon fios while miles off shore. It is a great solution for me but clearly not for all sailors.

Below is a pic of the device I have. You can add standard voice service that will alow your home phone to work offshore but my cell works fine most of the time.

[IMG]<a href="http://s1237.photobucket.com/albums/ff473/capitalistnyc/?action=view&current=clearwire-clear-spot-4g-wifi-hotspot-2.jpg" target="_blank"><img src="http://i1237.photobucket.com/albums/ff473/capitalistnyc/clearwire-clear-spot-4g-wifi-hotspot-2.jpg" border="0" alt="Photobucket"></a>[/IMG]