Aligning an 80GHZ link at a mile and other licensed backhauls

Recently we had a teaching moment for a couple of folks who had not had much experience with aligning higher frequency antennas with very tight beamwidths.  This particular day we were aligning 2 foot Siklu 80GHZ antennas.

One of the questions we often get asked is how do you align these? These questions are usually asked by someone who is familiar with aligning 5ghz antennas with a 10 or 20 degree beam which you can eyeball and has tried a microwave shot. They find out it is much harder.  The higher you go in frequency the tighter and smaller the beam is.  Distance also affects how far off you can be.  Think of it as a laser pointer.  If you have ever taken a laser pointer out at night and shone it a long distance you will notice even the slightest movement will cause it to jump inches, even feet.  Keep laser pointer analogy in mind for this next section.

In order to understand alignment, we need to understand lobes on an antenna. An antenna is just a device that focuses radiation in a direction.  In a licensed microwave setup, these antennas focus the radiation in a tighter “beam”.  Let’s go back to our laser pointer analogy.  Some laser pointers project a smaller dot at 10 feet than others.  Same for antennas.   The diagram below shows what is called the main lobe and the side lobe.

The way to get the best signal is to get both dishes locked on to the main lobe. Sounds easy right? With higher frequencies, you are talking about millimeter waves. This means the main lobe may only be 3mm wide, about the size of this text on a laptop screen.  Now imagine trying to keep that 3mm beam in the center of a paper plate at a mile.  On top of that, the difference between the main lobe and locking onto a side lobe could be the difference of 1-2mm. A slight wind can move a dish 2mm.

To give you a real-world example. A 2ft 23 GHz antenna having 3 dB beamwidth of 1.6 degrees. Allowing for a path length of about 2.5 miles (this is licensed 23GHZ) the actual beamwidth at the receiving antenna is around 370 ft and is, therefore, likely to be greater than the height of the tower. If the antenna’s out of horizontal by even a couple of degrees to start, the antennas will miss by around 460 ft and not be able to “see” each other. This can be amplified as frequency and distance increase.

This is all fine and dandy, but what about the practical world? How do I align the thing?
It all starts with the FCC path coordination paperwork you will receive on your licensed link. There is a wealth of information in here.  It tells you all of the following:
-Your mounting height (this is typically already known)
-Your heading (more on this in a bit)
-The antenna angle downtilt or uptilt (very important)
-The expected signal target

Armed with this information you will have all of the information you need to align the link.  From this point, the philosophical side of things kicks in.  Some tower climbers are good with using a compass to get their exact bearings.  Others have high dollar tools to do it all via GPS such as microwave path alignment from Sunsight.

What everyone doing alignment should have in their toolkit are the following:
-A small magnetic bubble Level. We want to make sure we start with a level mount.  We would be fighting an uphill battle if the pipe or standoff we are mounting to is not level.

-An angle Finder is very helpful for determining the antenna down or uptilt per the path calculation.

Obviously, the above tools are just one of many examples.  There are more expensive ones and bare bones ones.  Tools are only as good as the person using them.

-Ratcheting wrenches for the left and right and up and down adjustments.
Having ratcheting wrenches makes fine-tuning a very easy process.  You will see why later.

-A good hands-free communication method.  Depending on the tower FM communications may or may not work.  Cell phones may or may not work. Being able to talk to the crew on the other end is crucial.  And yes, to make this smooth you want a crew on the other end.

Aligning backhauls, especially microwave, is a skilled trade.  With any skilled trade, you will get all kinds of tips and tricks of the trade.  Some you may use, others you may not.  Ask any Carpenter, Drywaller, or Mason and they will tell you little tips and tricks. They probably all are great and will work, but you may only use some of them.  I am going to tell you mine. You may find others you like better.

We always start with a google earth plot of the path. I call this Phase 1.  The goal of phase 1 is to get the radios talking.  We make sure the line is exactly on the two points, not just approximate.  If the backhaul it on the left side of the tower, we draw the line to/from the left side of the tower.  We then pick 2-3 landmarks along the path as we can.  We start with something close to the tower the climber should be able to see.

In our photo above we have picked out two reference points close to the tower the climber can see.  The first is the clump of trees on the climbers left.  The path passes “just to the right” of the edge of the end of the trees.  The second reference is the intersection of the county roads about 2-3 miles out.  Our path should be just to the right of those.  That point of reference is more of a sanity check. More than anything. The climber at the other end has a similar printout.   I have found communication during this process works best if both climbers and someone logged to at least one radio on the ground with a laptop are on a conference bridge.  Many radios have lights, tones, or multimeter outputs to indicate signal.  Some modern radios only have web-interfaces and apps.  Hold a phone while trying to align can be cumbersome.  This is where the guy on the ground can take some load off what the climbers are doing.

Regardless of the mechanics of the radio, the goal of Phase 1 is to establish a radio link, no matter how bad it is. Now, here is where the real meat and potatoes of backhaul alignment come into play.  This is a very deliberate and calculated process.  Your goal at the end of the entire alignment process is to end up with the following diagram

What many folks don’t realize is it is possible to establish a signal on a side lobe. So how do you know if you are on a side lobe? Here is how we start phase 2. This is what I call fine-tuning. Real original huh? Depending on good, or lucky you were during phase 1 you may have a long way to go or a short way to go to meet target.  Remember that in your paperwork we talked about earlier?  One side and one side only starts moving their fine adjustment on their antenna to the left and right and up and down.  This is typically called sweeping.  The key thing to note here is you need to find the very edges of the radio signal, not just the lobe you happen to be on.

Let’s take a real-world example to explain how sweeping affects main and side lobes.  At the start of this article, we mentioned an 80ghz link.  With our phase 1 rough alignment, we were able to get linked at a -86.  The target was a -32.   The first side to start alignment started sweeping to the right, signal started going from a -86 down to a -72 rather quickly. This was using very small turns of the adjustment.  The ratcheting wrench was only clicking 1-2 times for each 2-3 db of signal change. Once it reached a -72 it started climbing back up.   The climber then kept going to the right to find the edge of the signal, not just the lobe we were on.  The signal started getting worse until we were back into the upper 80’s.

Now, the climber brings the alignment back to the left, and stops at the -72 and makes a mental note of where that is in relationship to the overall placement of the dish, etc.  Some mounts have distinct notches, some guys use markers, others just remember.  Now the climber continues on to the left and the -72 gets worse and goes back down to the -86 and continues to get worse.  So the climber, at least for now, has found the sweet spot for the left and right alignment.  The climber also knows this will probably change, but has found it for now.   Climber repeats the same procedure for the up and down. Due to the anglefinder, the climbers have with them they feel pretty confident they are fairly close with the up and down so they do not adjust the up and down travel as much as the procedure goes on.

Next, the other side does the same procedure the first side did. They do the left to right and get the signal down to a -62. Essentially, what the climbers are trying to do is find the center, which will contain the strongest signal, by sweeping past the other signals.  Keep in mind there may be only millimeters separating these other lobes.  Due to physics, and the shape of the signal, the first lobe is actually stronger than the edges of the main beam.

Say what? The first lobe is stronger than the edges of the main beam? Yes, but not stronger than the main beam.  Let’s go back to our installers. They have each had a go around at alignment and are only at a -62.  On a 5ghz backhaul that would be respectable, depending on your noise floor. But we are 30db away from our target of -32. Some climbers, incorrectly I might add, try to do a shortcut by scanning in an x pattern instead of x and y-axis separately. This makes it easier to lock onto a side lobe.

80ghz backhaul

So now our first climber goes back to making the left and right adjustments.   At this point, the installer finds something odd.  He has gotten the signal down to a -55, but that’s the best he can do. Even a small turn jumps the signal up    Then our installer remembers the above statement.  The first lobe is always stronger than the edges of the main beam.  He gets the signal back down to a -55 and turns the alignment over to the other side.

Here is a very important thing to note.  Both of our installers have now “gotten a feel” for the few turns needed to adjust the signal on these dishes.  To them compared to 5ghz dishes, these are very tiny and almost insignificant movements. But they sure make a difference in signal.  Now our installer at tower B has his second alignment session.  As he is making adjustments the signal is not changing.  He is moving his wrench for what seems like forever and the signal is barely moving, Any other time their signal would have been a -90 or dropped.  What has happened here? The main lobe of one side has locked onto the first lobe because it is always stronger.  Since the main lobe is bigger it seems like it takes forever to make any change.  If we had a guy on the laptop he was probably also probably seeing very mismatched data rates.  One side was probably much higher than the other by a large margin.

Then boom, all of a sudden the signal goes from a -55 to a -42.  A 17 db jump!   We can now tell we are on the main lobe.  If the laptop person looks at the data rates now they should be more balanced.

Data Rates on a Mimosa B11 Rates properly aligned but not fine-tuned

At this point, it is just a simple matter of each side making finer and finer adjustments back and forth to get the signal down.  If you think of the above circle/crosshair you are making smaller and smaller adjustments to nudge toward the center of the circle. This is where the ratcheting wrenches help by giving a very measured amount of travel.  This helps with the whole feel of alignment.  Much of it is feel to see how much you can move the adjustment mechanisms to make the numbers move.  Sometimes it may be a single click of the wrench.  Sometimes it may be one or two.  It just depends.  As you get closer and closer to target you are moving the adjustment less and less.

As you get closer and closer to target you need to be thinking about how tightening down the adjustment bolts will affect the alignment.  Even tightening them down snug can affect the signal.  That extra amount movement to tighten them down can move them slightly past their alignment center.  You may need to take into account the amount of travel it takes to tighten down the adjustment bolt into account on smaller dishes.  If it takes a half turn of the bolt to get it tight you may need to stop a half turn and tighten “into” target.  As you tighten it down fully that is where you end up in align.  If you wait until you are in align and then snug it completely down, the force of snugging it down may pull it past and you will end up with a worse signal.

This article sprinkled in some examples from a real-world install, with some theory, with some practical knowledge. Your mileage and experience will vary.  Your experience with 6ghz vs 80ghz will vary as well. Each frequency will have it’s own quirks and tricks.

DHCP Starvation attack

DHCP starvation attacks are designed to deplete all of the addresses within the DHCP scope on a particular segment. Subsequently, a legitimate user is denied an IP address requested via DHCP and thus is not able to access the network.  Yersinia is one such free hacking tool that performs automated DHCP starvation attacks. DHCP starvation may be purely a DoS mechanism or may be used in conjunction with a malicious rogue server attack to redirect traffic to a malicious computer ready to intercept traffic. Imagine a user filling up the dhcp pool and then re-directing users to their own DHCP server.

How do you fix this?
802.11 has several mechanisms built in. DHCP Proxy is one way. Port security is another. If you are running Mikrotik there are some scripts which can alert you to rogue DHCP servers, but that is an after-the-fact kind of thing.


Apple USB to Serial

Several weeks ago I had to replace my USB-to-Serial adaptor.  After much research, Facebook posting/discussion I bought a Ugreen USB adaptor from Amazon.  After a few times, I started having issues.  I figured, like most network engineers it was software.  Long story short it was a hardware failure and a replacement one fixed it.

However, in my searching I came across a little gem simply called Serial from Decisive tactics.   What sold me on it was the ability to do profiles.  Many times I am connecting to Cisco and the old way is the terminal emulator defaults to something that is not 9600. So I have to go into preferences, change it, apply a few times and I am good. Lots of wasted clicks.  With Serial I can select my profile and off I go.

AIrCube Part 2

Part 1 of our AirCube article can be found here

One of the best features about the new aircube is the interface.  On Facebook and other places the aircube has been criticized for it’s lack of features. I believe this is where the simple interface really makes the unit shine.   One thing many people don’t realize is your typical home router really doesn’t have a true firewall. Most routers have features that are firewall like. Most “firewalls” are security by obscurity.  The ability to close off ports is a by-product of a nat router.

Setup was very easy.  I downloaded the ap, scanned the QR code, and then went into the phones wifi and connected to the AirCube wifi. Once I plugged it into my home router I was online.

One of the first things I always do when testing a new device is upgrade the firmware, unless I have a specific firmware version for whatever reason.  Upgrading the firmware on the unit was very easy.  I like the fact you can see the changelog notes with a single click.

Once upgraded the simplicity of the setup really shines through.  By default, the AirCube is in an access point mode. This is probably the default mode which will be the most beneficial for the ISPs out there.  I will explain why in a little bit.

The interface speaks for itself, I could go through screen by screen and spoon feed you the very simple setup screens, but you would get bored very quickly.  It is truly a very minimalist product and interface.

One of the coolest features I like about this product is the scan feature.  For some reason the way it is presented on screen resonates with me.  It makes seeing the frequencies in use handy.  While not a chart or graph, it still accomplishes the same function.

I know this little product has received some ripping from operators on various groups and forums.  However, I think it fills a very basic need.  That need is an easy to configure device which allows devices to access the Internet.  Nat firewalls can be done by the provider, and the user never has to touch them.  This is a trend many operators are headed toward anyway.  Let the provider manage the endpoint for the customer. The customer ends up calling the ISP to do port forwarding or if they have a virus anyway.  If the customer wants to add things in the house, it should be simple, and not add extra layers of NAT and firewall rules.  This is a perfect fit for the Cube.  So don’t dismiss it because it doesn’t have all the fancy features some router have.  Save that for other product lines that you, as the ISP, can manage.


MTIN announces forthcoming 2018 Pricing changes

In an effort to better serve customers MTIN is changing some of our pricing structure. We are moving toward a limited availability structure to guarantee availability to our core customer base while recognizing a one-size-fits-all solution isn’t for everyone.

Tier 1: Customers without a service contract or retainer
$109 per hour during normal business hours (8am-5pm EST time)
$135 per hour for after hours and weekend
First come first serve. No SLA.

Tier II Support
Normal Business Hours (8am-5pm EST time)                           $89 per hour
Late night & Weekend support                                                 $125 per hour
Must Purchase in 5-hour blocks ahead of time to meet response times below for any non-scheduled work.  After hours work can be scheduled ahead of time for normal business hour pricing.

Tier II emergency response times for Tier II customers with a time balance.
Normal working hours (3 hour maximum)
Late night and weekend (4 hour maximum)

Tier III Tech on call Plan
This plan is designed for those who need a high-level engineer on a WISP budget. These slots are extremely limited.

5 hour blocks per week available
6 month minimum commitment
as low as $150 per week paid bi-weekly (1/2 off normal rate)
Discounts are given for multi 5 hour blocks
response times much lower
Contact MTIN for complete details

Contracts available
-Priority support
-Faster response times
-Late night & Weekend support rates don’t apply
-Access to backend monitoring and other services
Contact MTIN on details on contracts

Policy details

1.Late night and Weekend is defined as. 9PM-9AM EST MONDAY-FRIDAY & 8PM EST FRIDAY -9AM Monday. For West coast and customer in other time frames work can be schedule to meet your needs and not be charge for after hours.

2.All customers who don’t have pre-purchased time will be served on a best effort service. Priority will be given to contract customers, and then customers with a balance, and finally to “walk-in” customers.

3.All times stated are maximum times for response. Depending on workload, times are typically much less.

Looking into to 2018

As the snow is melting off during the 50 degree days I have a moment to pause for some of the things coming in 2018.

One of the things we will be working on is some tweaks to our pricing and support structure.  Over the past year, we have been moving to a deeper relationship with our clients.  This helps me provide better support as well add stability to income.  MTIN will be rolling into a more personalized service for clients in the upcoming months.  This will mean more one-on-one attention to clients. With this, we will still be doing “walk-in clients” but only as time allows.   Customers with service contracts and retainer balances will get priority.

Secondly, partnerships with other companies will become a way we will be able to provide customers with a great experience.


We look forward to this new year!

MTIN is now a FLexOptic Reseller

MTIN typically is not a reseller for many product lines, for several reasons.  We like to be vendor agnostic and not chasing sales commissions on products, and we are not in the business of stocking product.

Having said this, we now have a reseller relationship with  They have optics you can code for a huge variety of manufacturers.  WISP clients will be intersted to know they support the following vendors:
and a whole bunch more. There are over 150 vendors supported.

The optics are coded with a product called Flexbox. The flexbox has several features to it such as coding, wavelength tuning of DWDM, distance analyzer, power measurement, and diagnostics.

FLEXBOX series - Configure Universal Transceivers | CSFP, SFP, SFP+, XFP, QSFP+, QSFP28, SFP28, CFP, CFP2, CFP4

We are working on some reviews, how-tos and other tutorials for these products. At the very least we are recommending everyone have a few optics of the form factors you use for compatibility troubleshooting.  If you have a device that you wonder if it is recognizing your optics correctly you can pull out this kit, code an optic for your device, and go on with troubleshooting.   Very handy for vendor optic issues.

If this is something you are interested in send us an e-mail for a quote on a starter kit and look for more information coming soon.

Client subnet in DNS requests

Some Light Reading:

Many Authoritative nameservers today return different replies based
   on the perceived topological location of the user.  These servers use
   the IP address of the incoming query to identify that location.
   Since most queries come from intermediate recursive resolvers, the
   source address is that of the recursive rather than of the query

   Traditionally and probably still in the majority of instances,
   recursive resolvers are reasonably close in the topological sense to
   the stub resolvers or forwarders that are the source of queries.  For
   these resolvers, using their own IP address is sufficient for
   authority servers that tailor responses based upon location of the

   Increasingly though a class of remote recursive servers has arisen
   that serves query sources without regard to topology.  The motivation
   for a query source to use a remote recursive server varies but is
   usually because of some enhanced experience, such as greater cache
   security or applying policies regarding where users may connect.
   (Although political censorship usually comes to mind here, the same
   actions may be used by a parent when setting controls on where a
   minor may connect.)  When using a remote recursive server, there can
   no longer be any assumption of close proximity between the originator
   and the recursive, leading to less than optimal replies from the
   authority servers.

   A similar situation exists within some ISPs where the recursive
   servers are topologically distant from some edges of the ISP network,
   resulting in less than optimal replies from the authority servers.

   This draft defines an EDNS0 option to convey network information that
   is relevant to the message but not otherwise included in the
   datagram.  This will provide the mechanism to carry sufficient
   network information about the originator for the authority server to
   tailor responses.  It also provides for the authority server to
   indicate the scope of network addresses that the tailored answer is
   intended.  This EDNS0 option is intended for those recursive and
   authority servers that would benefit from the extension and not for
   general purpose deployment.  It is completely optional and can safely
   be ignored by servers that choose not to implement it or enable it.

   This draft also includes guidelines on how to best cache those
   results and provides recommendations on when this protocol extension
   should be used.

For those of you running BIND here is some practical information

UBNT Air Cube first impressions

I have been meaning to start this review for several weeks.  Due to the holidays and sickness that has not happened until now.  Recently Ubiquiti Networks released the airCubeAC. I won’t bore you with all the stats, just some of the highlights.  For the complete list go here…

-AC radio containing 5ghz and 2.4 Radios (AC Model)
-4 Gigabit ethernet ports
-Supports POE in and Out

One of the first things you notice about the modern UBNT products like this is the nice retail looking package.  This could be on the shelf of Best Buy, or on the shelf of any computer shop. The packaging is modern and eye-catching.

After unboxing we find a very minimal packaging.

All that is contained in the packaging is the airCube itself, quick start guide, and the power cord. One of the first things I noticed as I went to plug this in was the length of the power cord.  Too many companies give you a short power cord you are always fighting against.  This cord has to be 7-8 feet long. In addition, the power plug is a compact size to fit into most surge protectors with ease.  It’s the little design features like this which can really make a product shine.

While waiting for it to boot a quick tour around the outside reveals the four gigabit ethernet ports, one of them being the WAN port.

The quickstart guide was very helpful, except for the terminology used for the UMOBILE app. On the IOS store, I finally figured out the UNMS app was the correct one to use. This might be confusing for some folks. Maybe newer documentation reflects the change in the naming.

I connected the Cube to my home network and fired up the app, the wizard was very helpful in getting me connected to the Cube.

The use of the QR code to customize the instructions is a very nice time saver.  I was up and connected within 40 seconds.  Most of that time was switching over to my settings to connect to the wifi and switching back to the app. A nice feature would be launching the settings app for you.  Not sure if such system calls are allowed on iOS but something to consider.  On a side note, there is Puerto Rico listed as a country yet again. Not sure why this is a recurring theme with UBNT.

Anytime I get a new device like this one of the first things I do is upgrade the firmware to the latest. This was a very easy process. The app even had a little orange information thing directing me to go check it. The addition of the changelog within the app is a very nice touch. The total firmware upgrade took about 2 minutes.

I made the mistake of switching out of the app before the upgrade was done. The unit was not reporting the firmware was upgraded, and when I tried to upgrade again it gave me an error. Hitting logout on the app and logging back in refreshed the app and confirmed I was indeed at the latest firmware.

It’s getting late, but I wanted to get this out there and get the ball rolling.  Look for part 2 coming shortly when I go over the interface in detail. For now, I will leave you with my first impression summary.

The airCube has many nice physical features.  The long power cable makes the flexibility of installation easy.  No longer do you have to set it in an awkward place just because the power cable did not reach.  It does POE in and out, so you could power the unit with a wireless CPE POE if you were a WISP running UBNT gear. This would save on a power plug because you would only need one for your outdoor radio and the airCube. However, if you are deploying these with non-UBNT gear, or simply in a home with fiber or cable the small power plug makes for a neat and compact installation.

Setup was easy, minus the documentation issue on the app to get.  This is probably simply the app being updated for whatever reason and the documentation that came with my Cube being behind.

Look for part two coming soon.