“Of Fox and Hounds” is a 1940s cartoon where a sheepdog, who is trying to catch a fox, is led around by a fox because he doesn’t recognize him as a fox (You should reread that first line again since I couldn’t figure out how to make it clearer). Unfortunately, the sheepdog is a little dim-witted and a little inexperienced in the job he is born to do. Figuring out what to do for expansion into new area right now has been pretty much similar to that. Between all the new products, the new products that aren’t meeting expectations or have feature delays, lack of real-world testing of some new features, changes in the 5GHz spectrum by the cell companies and the cable companies (and even Microsoft is throwing their hat in the ring now), and whatever additional damage the FCC might do because T-Tommy Wheeler isn’t sure who he has to kiss up to this week, I’m starting to feel like the poor sheepdog. AARRGGHHH, I’m so confuuusssseeeddd! If our industry were a GPS right now, I’d be driving off the cliff. What’s frustrating is that I have to keep re-evaluating this decision every few months as new things keep getting delayed (if you keep promising to give your girlfriend a ring and your delay excuse is that you are waiting for the jeweler to make it, eventually you are going to need to find a new girlfriend). I also need to dispel some misinformation that I keep hearing from new people contacting me that are getting into the industry. So using some logic, reading lots of forums, and grabbing enough Diet Coke for a long session, let’s get see if we can piece through it.
Because the number of variables here exceeds the number of emails Hillary Clinton deleted, I’m going to focus it down to a new area scenario with a single manufacturer that may or may not have to work with other manufacturers in the future (pound your head against the wall a couple of times now, it will all become clearer or when you wake up, you probably won’t care). I’ll cover cross manufacturing compatibility someday when I get a chance to actually test how it works. For example, I haven’t tested a Cambium ePMP client to see if it will talk to a Rocket AP in 802.11 compatibility mode yet. Theoretically it’s supposed to work (and theoretically Hillary Clinton’s 30,000 emails were about Yoga and not the billions of dollars she got from influencing peddling. Just wait, I’ll get T-Tommy Wheeler in here soon enough.) but how far, how fast, how well, and how many gotchas is not something I have personal experience with. I do know Bitlomat is working on their AP solution that is supposed to work with Ubiquiti clients but I have no idea of when that is coming out either. Supposedly it’s going to require new firmware for the clients, something along the lines of what SkyPilot did. So let’s just go back to what we do know is on the edge for each manufacturer in the PTMP area, Cambium has 802.11n and some boys working on 802.11ac far, far away, Ubiquiti has a whole 802.11ac product line that just passed one FCC certification and now theoretically finally supports 5.15-5.25GHz (firmware still hasn’t shipped as of today so keep checking their website, no DFS yet and lower power in UNI-3 than we are all expecting), and Mimosa has 802.11ac backhauls but isn’t shipping PTMP yet. Since 802.11n is the past, at least for me where all my clients are going to be 2 miles or less (802.11ac will go farther but it’s needed for my deployment strategy), I’m still focusing on Galactus, Destroyer of Wired Worlds, Worlds, Worlds, Worlds, . . .
That leaves us with Ubiquiti 802.11ac radios for the moment. Since I had two birthdays and rehabbed my pitching arm from rotator surgery during the time I was waiting for UNI-I and DFS for the Power/NanoBeams, you have to know that there is no guarantee you will see DFS from Ubiquiti before you see PTMP 802.11ac radios from Mimosa or if Cambium will even announce what they are working on or when. So what you put up will have to work with UNI-1 and UNI-3 for the near future. If you are rural, this isn’t an issue. If you are urban, it’s a gamble but you can always fall back to 10MHz channels if needed, which sort of negates the reason for 802.11ac (now slam your head in the refrigerator door and grab another Diet Coke). But it might buy you the time you need for a few more months.
Right now we have to deploy several microcells in a variety of areas and we will be using Rocket AC-Lites with omni-directional antennas. I can already see people cringing but there simply isn’t enough density to justify spending any more than that. Plus the antennas are going to be mounted at heights of 20-35’ so downtilt isn’t an issue either. Considering that these particular microcells are only handling about 10-50 users to start with, they won’t even be sweating. Race cheap and race often, one of my favorite mottos from my motorcycle days gets translated into, get a lot of small AP locations as cheaply as possible and establish yourself in as many areas as possible. When the inevitable interference comes from competitors, you will be much closer to your clients than they will and you will have new technologies to use. It’s also going to be where the split-frequency capabilities of Mimosa and BitLomat are going to help in future-proofing your network (Bitlomat is interesting in that they can use different size channels in each band down to 5MHz).
Now we get back to another of my favorite sayings, speed, distance, quality, pick 2. If you control the distance, that’s huge. People are also loath to change so if you get to the customer first; it’s going to be hard to take you out. This is especially true if you are starting with radios that can do 100Mbps as easily as T-Tommy Wheeler can create programs to give tax dollars to big Obama contributors (I’m still waiting to hear how well $90B in green energy programs worked out for the Energy Department) and Obamaphone fans (it’s coming, bandwidth for “free” under the same idea). On a quick tangent, it’s also curious that a former Solyndra (the infamous poster boy for “Green Energy” Obama investment) VP of Product Engineering, Wayne Miller is now working for Ubiquiti (as Artie Shaw/Wolfgang from Laugh-In would say, “Verryyyy Interesting”).
To understand how the microcell concept works, let’s look at the financials. Also, keep this in mind; Home Depot is your new BFF. Sticking a pole in the air, either on a roof or on the side of a house, costs less than $100 from them. Yes, you could use a Rohn-25 but the cost just went up by a factor or 4 or more. As long as you can get on the roof of the house with an A-Frame ladder or you can lean it against the pole, you can go about 30’. With proper fastening to the house with channel bar and pipe clamps, it’s not coming down. We usually want 4 points of contact against the wall but 3 would be fine if you are anchored really, really well. If you put in a concrete foundation a couple feet in the ground, then 2 might do it but you are going to want 2 at the top. All of this is in lieu of more expensive mounts. I have yet to have one detach from a wall or have the brackets pull out of the channel bar and we have 70+ mile per hour winds, micro-bursts, and Haboobs (wind storms). If the roof is flat, then just use a 10’ non-penetrating mount which is a lot more portable.
Throw in an AP, an omni-directional antenna, 1 or 2 backhaul radios, a router or switch, a NEMA box, and whammo, for less than then cost of a set of tires and trip for me through a Wendy’s drive-through, we have an AP location generating several hundred dollars a month. This is probably the simplest of APs but it works (but if the husband is makes the decision, you better get the wife to sign off so that when the house starts looking like a NASA communications center, she doesn’t go apoplectic). With the backhaul radios we have coming, this is kind of a no-brainer. And microcells that can be deployed anywhere give you the flexibility to not only get closer to the target client base, they can also be strategically deployed between your area and larger towers behind you. If the idea wasn’t basically sound, Vivint wouldn’t be doing it.
Then again, I’m not sure what Vivint was thinking (tangent time) when they announced a 100Mbps upgrade. That’s pretty ambitious based on current technologies, their current mechanical and electrical design, and the current firmware on Quantenna chipsets that they sort of developed in house. This gets especially interesting if the goal is to still try to have some of the users on indoor only APs. Last I saw, they were looking for up to 24 users per relay AP. It makes for good spreadsheet returns but unless their developer, Huawei, has um, acquired some new firmware through their usual R&D processes which historically has involved blatantly violating the patents and trademarks of most technology companies (yes, Huawei is an arm of the Chinese government who sanction them stealing technology without penalty), I doubt they are going to do much more than mess up 5GHz wherever they go. They certainly aren’t going to make a profit unless it’s piggybacked with their other offerings like solar or alarm. At least they raised their price to about $60 instead of $50 to more accurately reflect their cost. I don’t have a problem with Vivint trying to get into the market, I just have problem with them using Huawei who has been banned from the U.S. cellular industry because they are crooks that have stolen American technology to grow their company. This is especially bad after the most recent Chinese hack of our Federal Employee Database. I’m beating a dead horse but to even think if doing business with Huawei in the United States is a serious public relations mistake with Huawei’s military connections.
Since there hasn’t been much movement in the industry lately, I’m going back to why microcells are a good idea. It’s been argued by some that omni-directional antennas are not the best way to go for multiple reasons such as low-gain, pollutes the spectrum, limited number of users, etc… Some of those arguments go away with the Mimosa A5-360s but until there get here, we are limited to Ubiquiti or KP Performance omni’s. Also, there is a lot spectrum in 5GHz to play with. In my case, I’m moving to 802.11ac so whether I stay with the Ubiquiti 802.11ac products or move onto Mimosa later on, I am not waiting. And with only a single radio on a roof, then I don’t need GPS meaning Ubiquiti is okay.
The decision behind this part of a deployment partly has to do with the fact there isn’t a functioning PTMP 802.11ac GPS system out yet. And no, I haven’t tested AirPrism yet to see how well it works so as an alternate to GPS, it’s still unknown. According to the specs, the adjacent channel rejection should be enough to put 4 radios on a tower. However, I’d like to see a test of AirPrism on a tower versus a GPS enabled product. I’m okay with using RF armor antenna shields to reduce noise, I just put it up on our last tower a couple months ago. With 802.11n Rockets and the original Ubiquiti sector antennas, it’s a necessity, not an option. The real question though is if you use AirPrism, do you need RFArmor or possibly is the combination of RFArmor with AirPrism mean GPS isn’t needed. Then the question is if the new 802.11ac antennas from Ubiquiti are so good that with AirPrism, it doesn’t need RF Armor. Then again, how will it compare to a GPS system like on the Cambium ePMP. Theoretically, it should be much faster. Unfortunately I don’t have the time to go through all this testing and I don’t need it yet.
However, since it will be several month before another option comes to the 802.11ac table, it’s mute. Rocket ACs with AirPrism are your only option. As for the antennas, the one advantage to the newer Ubiquiti AC antennas is they are available in 45 and 60 degree sectors. But now you have 6-8 Rockets on a tower with no GPS and with 802.11ac and 40MHz channels, it’s anyone’s guess how AirPrism is going to work in that environment. I know I’ll be waiting to see that test. And if that test gets done, it’s really only going to be valid with at least 20 users or more on each AP so I don’t see this happening soon, especially since all the DFS channels aren’t going to be available on Rocket AC’s in the forseeable future.
As for interference, keep in mind that the best way to prevent interference is to control the environment. Also remember that the low-cost provider should easily capture 30% or more of a market. If you are in an area with cable providers, to reduce the interference, just undercut them and take them out on price. It’s kind of impossible for Time Warner or Comcast to interfere and also set up hot-spots if they don’t have customers to base them out of. Dish and DirectTV also have indoor 5GHz products now for their video streaming so that’s something else you have to watch out for. It also helps if you provide the indoor AP also that you manage.
While we are on this hot-spot idea by cable and DSL providers (grab a handrail, tangent time again), it’s a huge pet-peeve of mine because the range of these indoor APs is about 30’ outside the house. Not only does it add to the interference level because they crank the power way up, but it’s a great way to meet new and interesting strangers who want to hang out in front of your house. The best ones are the cars that park on your street and sit and play on their laptops or phones. Yea, that’s a great idea, especially in areas with lots of kids. Since nobody wants a bunch of unknown people standing on your front sidewalk, I would suggest you call your cable provider and make sure they turn that feature off.
It’s not as easy as it sounds though, they not only don’t let you do that yourself by logging into your router (it’s the only thing they don’t let you change), they tell you it takes 4 hours for it to take effect even if you do call them. The bigger problem is that Comcast never turns it off, even after they tell you they did. My suggestion is to not believe the 4-hour story and make them turn it off why you are on the phone with them. If they ask you why you want it turned off, and they will, tell them you buy their product, you aren’t there to promote their product. Or just tell them you are sending them a bill for $25 per month for Router/AP service rental. If you really want to have fun, tell them that to encourage even more users to use their system, you are going to set up an outdoor AP on your house and let the entire neighborhood use it but you are going to take off your encyryption to promote it and use the same SSID.
Moving on with another of my pet peeves, the recent brilliant decision by the FCC to fully support unlimited and illegal file sharing (they are really saying we can’t throttle anyone for downloading illegal content but your carrier can disconnect you if one of your users downloads too much) means more work, more costs, more chance for a lawsuit, and less profit. This is definitely a case of where bureaucrats have to be bought and paid for because nobody can possibly be this intellectually dishonest. So my question then is this, what if we weren’t WISPS any more, what if we are network services who allow users to connect to our network using a secured infrastructure? Or what if we don’t sell unlimited internet, but we sell a secured internet browsing experience? What if you use private IP addresses, is that a private network? If the user knows he is signing up because of the security of a web filter like a Barracuda which will filter torrents, hacked web-sites, and block viruses going both way, does that violate Net Neutrality or does it legislate that common sense internet usage isn’t possible?
I’m sure Google never told T-Tommy Wheeler that not everyone is capable of protecting themselves while using Internet since it didn’t fit into their plan of taxation and control. And since Google is now abandoning their space-based satellite internet and abandoning their fiber expansion, maybe they will try to quit pulling T-Tommy Wheeler’s strings to put WISPS and smaller ISP’s out of business. I’m sure WISPS weren’t in Google’s target vision like they were in CenturyLink’s, but the reality is that Google has a monopoly in direct violation of the Sherman Anti-Trust Act with their search engine, they have a bunch of bought and paid for politicians, and they have a corrupt Justice Department led by a corrupt White House administration that are perfectly willing to let them keep that monopoly. Combine that with the fact they wanted to extend their search engine monopoly into a full bandwidth play monopoly and you would have had a huge problem. Fortunately for the ISP and WISP industry, they realized the technologies they wanted to back didn’t meet their financial goals (Elon Musk still hasn’t figured out the speed, distance, quality equation yet so he is still in the satellite game. Heck, he hasn’t even figured out the speed, distance, cost, and profitability yet for electric cars either but I like the vision).
As usual, this article started off in one direction and veered onto several different side roads. With the lack of new PTMP 802.11ac unlicensed products coming out in the market for a few months, it was a good time to dump some stuff. It’s also a good time to keep pushing the idea that WISPS need to come down from the hills and to keep looking at the different ways to do it. Knowing what the environment is going to be like is important. Knowing the best way to do it today and the best way to do it tomorrow is also important. Today, assuming you can live without DFS and you are ready for 802.11ac, then Ubiquiti is the way to do it if they can at least get 5.1GHz out relatively soon. It’s passed FCC cert and I have stickers, usually a good sign, but a month later still nothing so my fingers are crossed. If it doesn’t come out soon, then I’d say stay with 802.11n for now.