Re: Ubiquity wifi access point

The antenna pattern charts are thoroughly confusing because of MIMO (also known as 802.11n). In order to distinguish between the two independent signals or beams used by the MIMO in the NanoBeam, Ubiquiti calls one signal "vertical" and the other "horizontal" in honor of their relative polarization. The direction, relative gain, and pattern are called "azimuth" and "elevation". When used with

802.11a/b/g which use only one signal, I think (which means I'm too lazy to check) that only the "vertical azimuth" and "vertical elevation" patterns are used. In other words, you're using the wrong pair of beam patterns. Fortunately, the other two are sufficiently close so your numbers will be unchanged.

There is no distance. That's not how an antenna pattern works. The outer ring is defined as 0dB, which is a reference level established by the strongest signal in any direction. Every other point on the pattern is at a signal level that is less than 0dB.

Doing it that way is highly convenient. For example, let's pretend you aim the dish antenna at a distant receiver. You maximize the signal by swinging the antenna. If it's designed correctly, the peak signal should be when the dish is pointed directly at the distant receiver. If not, you have what's called a "bore sight error". You then scribble down the maximum signal level (in dBm). You can then use the antenna pattern to calculate the signal level at any point of the compass (azimuth). If the outer ring were a distance, which makes all the points on the plot absolute values instead of relative values, you couldn't do that.

Perhaps it would be helpful if I explained why I have a preference for calculation over "it just works". One problem with my calculations are that they are the best case numbers. Everything that brings the numbers closer to reality makes the range and coverage worse. If I threw in reflections, Fresnel diffraction, interference, component variations, tolerances, and just plain lies on the data sheets, the range would be less and the coverage patterns smaller (or less coverage). The question my calculations answer is "Under ideal conditions, does this thing have a chance of working"? If it doesn't pass that test, I usually don't bother trying it and look for a better solution.

Of course, something unexpected will always appear that were not considered in the calculations. In this case, it was a steel fence. I still don't know exactly where it is located or what effects it might have on Fresnel zone diffraction and any reflections. I'm also not sure about how much the brick wall attenuates the signal. Looking at the photo, I'm wondering if the 150ft is a measurement or an estimate. Any or all of these can screw things up, but probably won't because of the rather large fade margin. If I don't run the numbers first, to see if the system has a chance of working, I'll never know if I have a huge fade margin where almost any change will not have a major impact, or if I'm skirting the bitter edge of failure with a marginal system.

Incidentally, I just realized a made another screwup. My fade margin calculations were from: Ubiquiti Nanostation LocoM2 -> Typical smartphone I'm so used to symmetrical systems, where I have identical bridge radios at both ends of the link, that I forgot to calculate the fade margin going in the opposite direction. The LocoM2 transmits at +23dB while the smartphone might manage +20dB. In this case, a loss of -3dB of fade margin will have little effect on the rather large fade margin, but I should have calculated it anyway.

I don't understand it either. The first few slides somewhat explain how AirMax (TDMA MIMO) works: I vaguely recall reading a white paper on the topic, but can't find it right now. Very little on the UBNT forum on the topic.

At this point, I usually dig through the available patents for clues, but dinner beckons:

If he has both another LocoM2 radio, and a mixed bag of client computers, he's screwed. It's one or the other, not both at the same time. In his case, there is no 2nd LocoM2, so the obvious choice would be to turn OFF AirMax, thus losing the speed benefits of

802.11n.

Now that's interesting because I never understood the polar charts.

Is this somewhat correct?

1. Vertical = Mimo #1
2. Horizontal = Mimo #2
3. Azimuth = radiation pattern when viewed from above
4. Vertical Elevation = radiation pattern when viewed from the side

Um. OK. I guess I just proved what I said, which is that I had never really understood those damn things! :)

I did notice they were in negative decibels, but I had figured that that was in dbM which is always (in our experience with our type of radios) in negative decibels simply because of the large size of the 0dBm point of

1mW.

I didn't realize that the negative wasn't that, until you just said it.

I like that you provide a practical use example!

Makes all the sense in the world. Especially if you're doing it for a business. And if you're buying the equipment.

Luckily, for me, I had the radio equipment lying around so it was OK to just try it and then worry if it worked or not.

I hope Rod Speed has the courtesy of letting us know, in the future, if it worked or not, and how well it worked (if he implements anything from this).

Yep, that's correct.

Nope. I didn't say "dBm" which is a numerical value equal to 1 milliwatt into 50 ohms. I said "dB" which is a ratio. 0dB would be a ratio of 1:1. Negative dB means a less than 1 ratio, positive dB is more than one.

Unfortunately, there is quite a bit of creativity in generating antenna plots: When the outer ring is 0dB (not dBm), it's a "normalized plot". If the radial scale is logarithmic, it called "ARRL style". When a linear scale is used (mostly in antenna advertisements), it makes the pattern look far more directional than it deserves.

I used to have a tech working for me that would spend company time using some of the most sophisticated test equipment available to work on company products. However, after hours, he would drag out his ham radio and CAP radios, plug in a light bulb for a dummy load, and tune for maximum brightness. You kinda remind me of him.

Yeah, that would be interesting. However, the tradition is to only post problems, complaints, errors, rants, topic drift, and new surprises. When things actually work as advertised or calculated, it's deadly silence. This kinda takes the fun out of trying to be helpful. Ideally, I would like to see field measurements and range tests to check my calcs and guestimates. That may have happened once or twice in the distant past, but I don't recall. These daze, I get "It just works" or "I did something completely different" which I find rather un-educational and disgusting.

Oh well. Labor Day vacation is now over and it's a back to work day. Good luck trying to get my attention again and Bah Humbug.

Thanks for helping Rod Speed.

Let's hope he reports back with details.

Rod knows where to find me (since I care more than most people to understand the immense differences between iOS and Android), and I know where to find you (since you care a lot about radios & instrumentation, but far less do you get into conversations about iOS differences even though you know the hardware differences in their radios far better than I do).

:)

Roy Tremblay wrote

Yes I will.

Unlikely not to even if I say end up a cripple or something, I would just get her son to do the physical work.

Jeff Liebermann wrote

I don't operate like that.

And sometimes you learn something from what doesn't work the way you thought it would too.

I have in fact done that with the medion repeater.

That's harder given that I don't have the hardware to do that.

I'm currently helping quite a few with our VDSL2 rollout and that has the big advantage that the modem synch rate is trivially available to see how well the estimate worked out. And with the RSP who has shown up from Singapore with a big splash here with very low prices. And with the current price war with cellphone telcos which has produced $10/month for unlimited calls and texts to any landline or cellphone in the country with varying amounts of data, around 1GB on a 4G+ service.

Irritating, anyway.

Rod Speed actually wrote:

To Rod Speed and anyone who wants to set up a ubiquiti radio as a receiver far from the house.

We had a half dozen spare Ubiquiti Rocket M2 radios on hand, because we swapped them all out for Rocket M5 radios instead (for the lower noise).

So a bunch of us grabbed a few and set them up outside, a few hundred feet from the house, pointing over the air back at the normal home SOHO routers.

At about 200 feet line of sight from the barn to the house, we got signal strength of over minus 65 decibels, which is pretty good, and when we tested speeds, they were asymmetric (even though all our feeds are symmetric) at 43 Mbps down and 17 Mbps up, which is fine for what we're doing (since we get our Internet over the air via WISP).

The key to set up the Ubiquiti radio as a receiver is "station" and "bridge" for the two tabs, "wireless" and "network".

The configuration file has the radio set up with the following cfg file which can be loaded into any Ubiquiti Rocket M2 (and probably other Ubiquiti AirOS radios).

------------ cut here -------------- bridge.status=enabled tshaper.status=disabled gui.network.advanced.status=enabled dhcp6c.status=disabled resolv.nameserver.status=enabled resolv.nameserver.2.status=disabled resolv.nameserver.2.ip= resolv.nameserver.1.status=disabled resolv.nameserver.1.ip= resolv.status=disabled ppp.status=disabled vlan.status=disabled users.status=enabled users.1.status=enabled users.1.password=ubnt users.1.name=ubnt system.eirp.status=disabled system.cfg.version=65546 sshd.status=enabled sshd.port=22 route.1.devname=br0 route.1.status=enabled route.1.comment= route.1.gateway=192.168.1.1 route.1.netmask=0 route.1.ip=0.0.0.0 route.status=enabled netmode=bridge netconf.3.up=enabled netconf.3.netmask=255.255.255.0 netconf.3.ip=192.168.1.20 netconf.3.hwaddr.mac= netconf.3.hwaddr.status=disabled netconf.3.alias.2.comment=Safety netconf.3.alias.2.netmask=255.0.0.0 netconf.3.alias.2.ip=10.0.0.200 netconf.3.alias.2.status=enabled netconf.3.alias.1.comment=Safety netconf.3.alias.1.netmask=255.255.255.0 netconf.3.alias.1.ip=192.168.0.20 netconf.3.alias.1.status=enabled netconf.3.autoip.status=enabled netconf.3.role=mlan netconf.3.mtu=1500 netconf.3.devname=br0 netconf.3.status=enabled netconf.2.up=enabled netconf.2.promisc=enabled netconf.2.netmask=255.255.255.0 netconf.2.ip=0.0.0.0 netconf.2.allmulti=enabled netconf.2.hwaddr.mac= netconf.2.hwaddr.status=disabled netconf.2.autoip.status=disabled netconf.2.role=bridge_port netconf.2.mtu=1500 netconf.2.devname=ath0 netconf.2.status=enabled netconf.1.up=enabled netconf.1.promisc=enabled netconf.1.netmask=255.255.255.0 netconf.1.ip=0.0.0.0 netconf.1.hwaddr.mac= netconf.1.hwaddr.status=disabled netconf.1.autoip.status=disabled netconf.1.role=bridge_port netconf.1.mtu=1500 netconf.1.devname=eth0 netconf.1.status=enabled netconf.status=enabled httpd.status=enabled httpd.https.status=enabled httpd.https.port=443ebtables.sys.vlan.status=disabledebtables.sys.status=enabledebtables.sys.eap.status=disabledebtables.sys.eap.1.status=enabledebtables.sys.eap.1.devname=ath0ebtables.sys.arpnat.status=enabledebtables.sys.arpnat.1.status=enabledebtables.sys.arpnat.1.devname=ath0 ebtables.status=enabled dhcpd.status=disabled dhcpc.1.status=disabled dhcpc.1.devname=br0 dhcpc.status=disabled radio.status=enabled radio.countrycode=840 radio.rate_module=atheros radio.1.txpower=24 radio.1.subsystemid=0xe1b2 radio.1.status=enabled radio.1.reg_obey=enabled radio.1.pollingpri= radio.1.pollingnoack=0 radio.1.polling=enabled radio.1.obey=enabled radio.1.mode=managed radio.1.ieee_mode=11nght40 radio.1.dfs.status=enabled radio.1.devname=ath0 radio.1.cwm.mode=1 radio.1.cwm.enable=0 radio.1.countrycode=840 radio.1.acktimeout=25 radio.1.ackdistance=600 radio.1.ack.auto=enabled radio.1.forbiasauto=1 radio.1.rate.auto=enabled radio.1.rate.mcs=15 radio.1.mcastrate=15 radio.1.chanbw=40 radio.1.antenna.id=6 radio.1.antenna.gain=24 radio.1.cable.loss=0 wireless.status=enabled wireless.hideindoor.status=disabled wireless.1.status=enabled wireless.1.ssid=Birdfarm2Office wireless.1.security.type=none wireless.1.hide_ssid=disabled wireless.1.devname=ath0 wireless.1.addmtikie=enabled wireless.1.wds.status=disabled wireless.1.authmode=1 wireless.1.scan_list.status=disabled wireless.1.scan_list.channels= wireless.1.ap= aaa.1.status=disabled aaa.status=disabled wpasupplicant.device.1.status=disabled wpasupplicant.status=disabled bridge.1.comment= bridge.1.fd=1 bridge.1.port.2.devname=ath0 bridge.1.port.2.status=enabled bridge.1.port.1.devname=eth0 bridge.1.port.1.status=enabled bridge.1.stp.status=disabled bridge.1.devname=br0 bridge.1.status=enabled gui.language=en_US users.2.status=enabled users.2.password= users.2.name=guest users.2.gid=100 users.2.uid=100 users.2.shell=/bin/false resolv.host.1.status=enabled resolv.host.1.name=Rocket M2 system.date.status=enabled system.date.timestamp=092600002017 system.timezone=GMT+8 system.button.reset=enabled update.check.status=enabled system.latitude= system.longitude=

------------ cut here --------------

Hi Rod Speed & Jeff Liebermann,

I just set up a spare Ubiquiti Rocket M2 radio & 28dBi dish antenna in station/bridge mode, which is what Rod Speed needs if he wants to use the radio at his neighbor's house, to pick up the weak signal from Rod's SOHO Wi-Fi router. We tested this configuration at about 200 feet and got perfectly acceptable signal strength.

I also set up a Ubiquiti Rocket M5 in Access Point mode, which is what Rod Speed is most likely to do, but this post is only about how easy it is to set up the Ubiquiti radios in station/bridge mode to *receive* signal from a SOHO router Wi-Fi access point.

I'm using the Ubiquiti Rocket M2 in station/bridge mode right now, connected to a desktop PC Ethernet port and then picking up the signal from my SOHO WiFi router, which is exactly what Rod's neighbor would be using if Rod opts to put a receiving dish and bucket router on the neighbor's property facing Rod's home.

Here is a photo of the radios that I'm playing with for this test.

Here is a photo of the radios set up in AP mode (where Rod Speed would broadcast his signal to his neighbor) and set up in station/bridge mode (where his neighbor would receive the weak signal from Rod Speed's SOHO WiFi router).

Up until today, I was using a Mikrotik RB411/R52n-M radio to pick up the weak signal from the SOHO WiFi router.

But we just got a set of spare Rocket M2 radios to play with so that's why I tested this out for Rod Speed.

For Rod Speed to test out what the neighbor would need to simply pick up the weak signal 300 feet away from Rod Speed's SOHO Wi-Fi router, here's all the neighbor needs to do.

1. Power up a Windows 10 PC (that's what I tested this on).
2. Connect the Ubiquiti Rocket M2 radio to the POE power "POE" port.
3. Reset the Ubiquiti radio to factory defaults (if needed).
4. Set the Windows 10 PC to a static IP address of 192.168.1.x (where x is anything not used, and not 20).
5. Connect the POE power supply LAN port to the Windows PC Ethernet port.
6. Log into the Ubiquiti radio at http://192.168.1.20 using the default login of "ubnt" and the password of "ubnt".
7. The radio will force you to set the country code & language and it will force you to accept the EULA checkbox.
8. The radio will force you to change the password, where it will take anything other than "ubnt" (e.g., "Ubnt" works just fine).
9. Go to the NETWORK tab and hit the "Select" button and select the SSID broadcast from the SOHO Wi-Fi router & enter the type of security and passphrase for that access point.
10. Hit "change" and "apply" and that's it. You're done!

The Windows 10 PC is now connected to the SOHO Wi-Fi router weak signal, and the Windows 10 PC is therefore instantly on the Internet.

In practice, the user can test this out at home, and then move the radio

300 feet away from the SOHO Wi-Fi router where the radio should still work pretty far out to connect to the weak SOHO router Wi-Fi signal.

Once the user establishes this works at 100 feet, 200 feet, 300 feet, etc., they can just put a router on the end of the radio, and they can wired or wirelessly connect any device they want to that router (such as a barn cam).

Here are screenshots of the relevant screens in the setup, but again, it's very simple because there is only one change that is required which is to set the radio to pick up the correct SOHO router Wi-Fi access point SSID, security type, and passphrase.

radio_001.jpg

radio_002.jpg

radio_003.jpg

radio_004.jpg

radio_005.jpg

radio_006.jpg

radio_007.jpg

radio_008.jpg

radio_009.jpg

Here are the screenshots of the setup to set this Ubiquiti Rocket M5 in the mode that Rod Speed wants.

In this setup, the radio will hang off his SOHO router by cat5 cable and POE, and then this access point will paint the next few miles with his Internet signal such that a neighbor only a few hundred feet away should be able to connect to this powerful access point with small devices.

This happens to be a 5GHz 30 decibel rocket, but the procedure is exactly the same no matter what Ubiquiti radio Rod Speed chooses to make his access point that paints the neighbor's home (as per the calculations from Jeff).

ap_001 Security Certificate override at 192.168.1.20 (default)

ap_002 Log in to 192.168.1.20 port 80 as ubnt/ubnt

ap_003 Make sure AirMAX is not enabled

ap_004 You should be in Access Point/Bridge mode

ap_005 Choose the SSID & security & channel & width you want for the AP

ap_006 Choose any static IP address that you want for 192.168.1.whatever

ap_007 There's nothing to change on the Advanced tab

ap_008 There's nothing to change on the Services tab

ap_009 There's nothing to change on the System tab

As a more easily digested top-level summary, for Rod and the others: a. Station mode (the default) b. Access Point mode

1. Station mode means the access point locks on to any given SSID/passphrase, acting as a "station". For example, you can stand on the nearest hilltop and point the radio down into the city miles below, select the best signal strength open access point, and connect to the Internet (if you're lucky with signal strength both ways).
2. Access Point mode means the radio acts as an access point of your Internet to anyone (who can be miles away) who wants to connect to your access point. You can stick the radio on a hilltop, pointing at the city miles below, and everyone in the city can "see" your access point (if you're lucky with signal strength both ways).

More details for setup on Ubiquiti radios such as these in my basement:

1. The Ubiquiti radios, out of the box, default to "station" mode, where you can temporarily connect them by wire or WiFi to a mobile computing device to log in (192.168.1.20, ubnt/ubnt) and point them at any access point (even those that are miles away) and then lock on to either the SSID or the MAC address. That's it.

After that one-step setup of choosing the SSID to lock onto, you can plug

*anything* you want into the radio (e.g., a router, a camera, a computer, a mobile device, etc.) and it will be using the Internet of the SSID you're locked on to.

1. The Ubiquiti radios can easily be set up in Access Point mode, where you plug them into your router and then you can put this access point up to 300 feet away from the router, connected by that Ethernet cable.

This allows you to paint any part of your property, e.g., your pool or your barn or your front gate, etc., or even to paint an entire city miles away, with your access point.

In this photo below, you see that I have one powerful Rocket M2 (2.4GHz) which is set up in "station" mode, while the other powerful Rocket M5 (5GHz) is set up in "access point" mode.

Bear in mind that these radios can go for a dozen miles line of sight when connected to a similar radio, but the distance will be far less if the other radio is a cell phone, a router, or a less powerful access point.

The advantage, however, of these powerful Ubiquiti Rocket M2/M5 radios is that they have 24dBi and 30dBi antennas respectively, which, if you know how decibels work, is a huge increase in a weak noisy received signal strength.

However, even these two relatively weak 14dBi and 18dBi antennas can still go for miles line of sight under the right conditions on the other side.

None of those figures even counts the added power of at least 25 or so decibels (dbM) of power input into the antenna, so that gives you just an idea of how much more powerful, overall, these radios are compared to your typical SOHO router (which would be hard pressed to garner even 20dBm of EIRP overall).

Update for Jeff and Rod:

Using that bullet and planar antenna above, I finally got around to physically mounting the Ubiquiti Bullet M2HP connected to a 14dBi planar antenna outside on a pole stuck into the leftover hole from an ancient

15-foot wide satellite dish which was removed long ago.

Here are some LOS Android-based received-signal-strength measurements for Jeff Liebermann and Rod Speed where I set up *exactly* what Rod Speed wants to accomplish.

Those measurements are roughly at about 300 feet LOS (-50dBm to -60dBm) and at about 100 feet LOS (-35dBm to -40dBm) distance from the radio which is wired to the SOHO router and set up as an "access point" (which is what Rod wants to do).