Straightforward out-of-the-box solution for extending WiFi range

Hi, Do you work fpor Ubiquiti? Nothing new you are doing there.

Hi Tony,

Two things I've always wanted to do:

1. Increase the range of my wifi signal *inside* the house
2. Connect to access points which are miles away

What I was documented is the first step, which is to increase the range. The second step is to connect to an access point that is miles away from that laptop.

I haven't done *that* yet, but, the potential now exists.

First, I had to be able to connect to an access point, period. It took me a long time to figure out *how* to do that.

Now that I've done that, my *next* test is to see if I can connect to my home broadband router from a couple of miles away.

It will be fun, to see if I can do that (but I need to find my inverter because this particular solution requires 120volts).

I'll report back if/when I'm successful.

There's a store in Chicago that specializes in DX WiFi. They are called

The noise floor probably consists of other routers on the same frequency. This is that 3 miles path, so you are bound to hear all sorts of crud.

There are ways to get omni gain, and directional gain, based on the antenna type. For point to point long distance, you want horizontal polarization. The cross polarization spec of the antenna will reject vertically polarized signals.

"Horizontal polarization is used over longer distances to reduce interference by vertically polarized equipment radiating other radio noise, which is often predominantly vertically polarized."

They then say do whatever you want (horizontal or vertical), but horizontal will pick up less crud.

I never did circular polarization on wifi, but it would reduce multipath. The reflected signal would be reduce by the cross polarization since the reflected signal has the opposite polarization. But the circularly polarize antenna will pick up both horizontal and vertical signals, just reduced by 3dB since they are not circularly polarized.

Most people just get two sector antennas and set them up horizontally polarized.

You mean work on both bands at the same time? They kind of get tangled in knots. I'm pretty sure these dual band units have band separation filters, and they are lossy due to cost constraints. That is the filter has an insertion loss which can be quite significant (say 4dB). They don't have individual lowband and highband antennas, so they are duplexing. The out of band signal is probably bigger than the noise floor.

A doubt you can get line of sight to a Starbuck. Trees are a pain!

You would probably have to use circular polarization, like a helix feeding a dish.

Defcon has done wifi shootouts, but they used similar gear on both sides. I'm not sure if anyone ever tried to set a record for long distance interception. I've done 10 miles in the desert, but there is nothing in the way of the signal and and no competing signals. If a square law holds, 20 miles would be 1/4 the power. I get my power and voltage ratios mixed up, but probably is would need 6dB more gain in the antenna. Doable. I'm at 16dB, and gear exists at 22dB.

Hi Miso,

My fault for not being clear.

What I had meant by "double the decibels" is that I've noticed that seemingly, the same equipment (like the dish for the Rocket M2 versus the seemingly exact same dish for the Rocket M5) always has a 3 decibel (i.e., doubling) higher gain on the 5GHz equipment as on the 2.4GHz equipment.

So, without actually looking it up, I always assumed that a given set of equipment is double the power if it's 5GHz.

But, maybe I'm wrong. Googling, I found this gain calculator:

Reading the "Antenna" section, I see a parabola is a parabola, so, I can safely assume the Rocket M2 (2.4GHz) and the Rocket M5 (5GHz) do use the same dish reflector.

Plugging in the numbers, into the "Parabolic Antennas" calculator of

So, for any given size equipment, I always assumed (perhaps incorrectly), that the 5GHz equipment inherently had twice the gain.

Hi Miso, This is interesting, the polarization issue. I really never thought about polarization, so, I'm reading (and re-reading) your post, to get the facts into my head.

For example, I had never realized that, for long distances, one polarization would be any different than the other - but - as you noted - most things are vertically polarized - so - to reduce noise - I'd want the opposite polarization for long distances.

Thanks for that insight! You seem to be years ahead of me in knowing what's what.

I'm so new to polarization, that I had never heard of circular polarization.

Googling, I find this "patch antenna", which describes what you said about penetrating obstructions:

Here's what they say which agrees with what you had said: "circular polarization means the antenna is better able to receive a clear signal due to the fact that the signal is both vertically and horizontally polarized at the same time. This effect enables the signal to penetrate through small obstructions and "twist" its way through trees, leaves and small structures. "

Hi Miso,

I'm a retired accountant.

I have to do the math!

And, it has to all add up!

:)

BTW, they must expect us to use just the feedhorn for the Nanobridge M2, because it shows right here in the router configuration that I can select the 3dBi gain of the feedhorn alone, without the additional 15dBi gain of the reflector dish:

Just by way of testing, I clipped the 23dBi NanoBridge M2, with just the

With the laptop on my lap, I held the feedhorn like a flashlight, pointing the invisible beam directly at the router at the other side of the house, pointing up at the ceiling, pointing down at the floor, and pointing directly away from the router (with the cat5 "tail" of the feedhorn denoting direction).

With the Nanobridge M2 transmit power dialed down to 6dB, which is the lowest it would go, I got the following signal strengths:

-38dBm, -40dBm, -52dBm, & -61dBm

All of these are fantastic signal strength numbers for connectivity, but, they're still too strong to make much of a conclusion about usefulness.

Since my signal strength is so powerful, even at the very lowest possible settings, I'm going to need to go a few hundred yards to maybe a mile away from my home broadband router to see how much the directionality really makes a difference.

Wifi is just like ham radio. Same issues.

The sector antennas are always able to do horizontal or vertical polarization, but sometime the pattern isn't appropriate to more than one polarization.

For example, say you bought a sector antenna that was vertically polarized and designed for coverage over a wide horizontal region but a narrow vertical region. That makes sense for a lot of vertically polarized applications, like you are covering a yard. But if you rotate that antenna so it is now horizontally polarized, you now have narrow coverage in the horizontal direction and wide coverage in the vertical direction, which is pretty useless.

For long distance, you want a laser beam. Since that can't be achieved, you generally want it tight in both elevation and azimuth. That is beaming is often done with dishes.

If you look at microwave links, they often have two dishes. One is horizontally polarized, the other is vertically polarized, That gives them isolation between the dishes. Generally one is TX and the other RX. You don't want your TX signal leaking into the receiver.

I haven't done much non line of sight wifi. But in SIGINT, you deal with scattered signals. Reflections, knife edge diffraction, even atmospheric ducting. So you never know the polarization. Circular works well in that case.

When you design comm gear, well perhaps only back in the days I was doing so, you would put in "hooks" in the gear to see the constellation or eye pattern, depending on the complexity of the modulation. The hook would be some secret multiplexing of pins on the chip so the internal demod could be viewed. Often as simple as just providing hooks to drive dacs. These schemes provide great insight to the effect of noise and multipath. It wouldn't surprise me if the developers of these wifi chips had special drivers just for that purpose.

The wiki is OK on this topic:

The constellation is the more interesting of the two. You can watch your phase lock quality by seeing the constellation rotate back and forth. If a dot crosses a boundary, that is a bit error.

The eye pattern is useful in aiming the antenna. You move the antenna around until the eye is as wide as possible. A wide eye means the bits are less likely to be in error.

I did my MSEE with emphasis on communications and worked in the field a bit on baseband mod/demod, Not RF but wireline. But the theory works for both. Wireline reached a dead end, so I moved on to other design areas.

It is a bit more complicated than that. Meditate on a very simple antenna, say a quarter wave monopole. The signal in the aether has a field strength, generally expressed in volts per meter. If you compare two quarter wave monopoles, with one being twice the frequency of the other, it will also be half the size. [Wavelength being inversely related to frequency.] Same field strength, half the size, hence half the received signal being sent to the receiver. So generally lower frequencies are better than higher frequencies if you want range, but only for the same type of antenna.

Now there are things you can do to improve the reception of high frequencies. When I mentioned that the higher frequency antenna is smaller, that refers to the aperture of the antenna. The smaller antenna has less aperture. To increase aperture, you make a more complicated antenna. Stack simple antennas to make one that have more aperture. You now have a bigger sponge by combining smaller sponges.

This all can be related to photographic lenses. The opening of the front element of the lens is the aperture. The focal length of the lens is the gain. This all gets confusing, hence the use of various calculators to predict received signal level.

Now for a dish, and I may not be correct here, but I believe the aperture of the dish is just the damn diameter of the dish. Whatever frequency you chose, the aperture is the same. The gain however goes up with the frequency. But I think you don't gain any signal strength when the dust settles. For one thing, your feed horn for the dish is now smaller, i.e. it lost aperture.

This would be a good Jeff L. question, though he seems missing in action.

Generally there are no free lunches in electronics. The only benefit to going higher in frequency is there is less atmospheric noise. At least to a point. In satellite reception, you can have the noise of the sun kill your signal.

This makes sense, as everything in radio transmission is a tradeoff.

Again, you have a great perspective on how to explain why we use the dishes for point to point over long distances!

Luckily, in WiFi, you know everything. (At least you do!). :)

But that isn't a WAP with integral antenna. If the device just has a RF port, then you need to know the characteristics of of the antenna to get the effective radiated power.

I'm not sure what the limit is in those ISM (free) bands, but there probably is some limit and the vendor doesn't want you to exceed it.

Maybe you can do the first wifi moon bounce.

Actually the network issues can be confusing. But that isn't radio. For instance when you do your multiple WAP scheme, you need to insure the DHCP ranges don't overlap.