Wireless Service Providers → Green Field WISP

On 24G from the shots I've seen on the AF24 units (I don't have an AF but I've seen two other WISPs using them) - I'd stay less than 3Km.

For what we do use: PtP at 5.8 I try to keep under 10Km. We deploy a 22db DualPol panel at 10KM and easily get over 100Mbps real traffic across that link. Total link cost is less than $1,100. That cost/speed ratio fits my business case and network design perfectly.

On PtMP We keep 2.4 at 6Km or less and 5.x at 4Km or less. Yeah you can do a lot further but it just gets more likely that you're going to have an issue. I'd rather Not pick up a customer, than pick up one that's going to cost me even one truck roll a year. We get almost no calls per month for service issues. Like a couple a month on a bad month. I have customer's with us 8+ years and We've never heard from them. Ever. That's where your money is made (or lost).

In this area over the past 10 years I've seen 8 other WISPs start and 5 of them failed. IMHO the ones that succeeded were conservative and realistic with their business plans, organically funded and have good engineering, (or at least they have common sense). Although I think I'm preaching to the preacher on that point; you sound like you know that end of the business already - just carry that sensibility forward and you'll probably be quite successful.

On the tower side, with very few exceptions out of 41, we've built our own. Tower rent here is pretty high though so it depends on your market. E.g. at 40M, 1 vertical meter costs ~$225 per month, per face, plus a per antenna charge of ~$75 per month. So three sectors and two backhauls at ~ 40M is roughly $675 + $450 + 5x $75 = a kidney and a lung / month. It's easy to ROI a ~40M tower over 3 to 5 years with those numbers.

O3b has 130ms of latency.......... Fiber gets a 300ms ping into Europe on a good day. Big advantage right there. The world has gone smart phone and tablet. I was reading about this "about to happen" tablet/smart device boom 3 years ago. Not only has it now happened, its a past event and the daily norm. The next big thing is cloud everything. This is already well underway. Latency is crucial.

Are you CHIP certified?

I agree with you. If you search old forum posts using VSAT, MEO, LEO keywords, you will find out that I have been the a consistent defender of satellite backhaul for remote and overseas WISPs. However, when fiber backhaul becomes available, it is almost impossible to ignore the lower cost and higher bandwidth.

P.S. I offer completely free advice through public posts and email. Absolutely no certification whatsoever and NOT offering myself as a "consultant" either. You are smart enough to do this yourself because of your own experience so far. Go to a Mikrotik conference to meet others like you.

Now this has me interested. Everywhere i go i study the towers around me. In this country there seems to be about 7 different types of towers that have been erected by the cell companies.

I came across a real interesting tower only last Friday. I was at an airport waiting for a parcel to come in on a local flight (it was a Ruckus radio running very very late for delivery thanks to our local distributor) and while waiting found a tower that looked to be local built but real solid. Its a nice design and i wish i could take photos of it but its at a police station. You don't take photos at police stations around here

I am very interested in fabricating my own towers. But i want something substantial and with a good ladder system. Thinking to build out of angle iron. Expensive but stiffness provided. I've been Googling for plans - nothing so far. I love those Carbon fiber towers, but as John says and i agree, there may be doubts on how stiff they are. Its more practical to build our own if we can find a good design. I guess the other way is just to look at a cell tower - these guys have it figured out.If we could get a reliable 3Km i'd be happy with that. I was already thinking 2Km but hoping it turned out to be 2 MILES (wishful thinking)....... I think if you can turn down the modcod and not drive it too hard and accept half of the published throughput then it might approach reality. I am not sure - i am just an armchair critic at this point. I've read just about everything i can find on these AF24's. I think they may be onto something with these.

I'm a bit of a skeptic though. I've seen too many failed promises in the RF/IT world. I've always been intrigued by how the real world financials usually shoot down my tech fantasies.

I think i am going to give this a go. You know it cant hurt to have a 24GHz pair running, even if we do go RedLine (but you know, Trango's data sheets just talk to me - what to do?) If we have a licensed link up (if we go that way, but it sure feels comfortable to know that a little slice of the airwaves around here would be mine - it may only be physiological, but when starting out on a new venture any little thing i can find to help me feel solid/strong as i face issue after issue getting it going is good) we could always use the extra BW of the AF24 for something. So many ideas on that. It will be an interesting learning curve. If it works out well we either end up with an extra link up or its so good we dont even bother to go for higher end gear. The aiming sounds intimidating though. Still, aiming is something we do a lot of around here.

Now ....its Ubnt..... is there any stock of the AF24's to be had?

That is one of the major cons to Ubnt. Stock. ...is basically C Band Tx. Its very good and solid in rain. I see a max of a 2dB hit in heaviest rain conditions. I would feel comfortable with 5.8. No worries at all if we have the LOS needed.

Now would be a great time to get some AF24's up for testing - its our rainy season now

For sure. It has an UPTIME problem though at least where we are. Sat is alive and well and its not going anywhere. Its really a matter of horses for courses. We have done and are doing fine on VSAT services. I have a job to do where i need to setup a DR situation for a good client. I have a 5IF at HQ for this client and am putting another 5IF at the clients DR location 4.8 Klms away. The 5IF's will be setup as Geo-redundant - linked via fiber to replicate databases every night. However i just cant rely on that fiber link between the HUBs. So i am going to to put up a radio PtP. And that got me into all of this. (and at 4.8 Km AF24's are not going to cut it - so if i put a relay in the middle it would be fine, but this adds an extra point of failure on a DR link). Back to the high end again. Thing is, i would really prefer ONE type of radio to work with, so there is no easy answer to any of this - unless $$$ are no problem. I've learnt the hard way, when it comes to this type of work, money solves all problems.

I only need the one PtP to do this job. But i get access to 4 good and 1 prime location at this clients premises around town. This sparked the thought that we could use those spots to put up PtP links and make a ring. From the PtP's we could shoot PtMP and do wireless links. Ahhh, human nature.... one little idea has turned into an avalanche. We have discussed internally for the last moth to 6 weeks throwing this whole thing around. We have decided to look a bit more seriously.

So now it all comes down to the right equipment - and so we go through the loop again. I posted questions here because there are some real hard core radio guys on here. If anyone can pick holes in my plans its you guys. Dealing with a consultant will be like a walk in the park after you guys I wouldn't be so sure of that But yep - lots of hands on. The screwy things we have seen over the years!!!! This job is high stress and loooong hours. We never leave the network - if i'm not in the field in some far corner of the country i am sitting at my desk monitoring the network. Having said that, we love our work.

I'd love to have a chat with you one day. Everyone i have mentioned my O3b idea to shuns the whole idea as stupid. I dont see it that way. It would be nice to talk to someone with a positive attitude about it.

What is your target market? Residential or Business? Take a look at SAF radios too, very solid and they offer 24 ghz links

Definitely, we will chat about past, present and future of satellite connectivity! I am Lutful (dot) Khan on Gmail and LinkedIn - really easy to get in touch.

You will probably find following comments interesting:

From Jan 2010: sadly ~50% of the world's population are not reachable by an ISP hooked up directly to fiber. google is funding O3B's meo mesh network . O3B stands for the "Other 3 Billion" to emphasize that reality.

It is possible to design Ku/Ka-band LEO/meo orbit mesh networks that multicasts at 1Gb/s rate (to millions of ISP caching servers) with 50-100ms latency.
»Re: How does the internet work and who runs it?

From 2007: Even the latest multiple spot-beam incarnations (like Wildblue) will surely bog down as customer numbers increase and it costs them ~$200M to add another geo satellites.

In this proposal, the LEO satellite constellation will mesh in space using high-capacity inter-satellite links using steerable antennas unlike current LEO systems.

Dozens of terrestrial gateways that deploy multiple steerable dish antennas will connect to multiple satellites simultaneously to complete the internet backbone.

Rural/remote WISP and ISPs who will be the final "subscribers" of this system will use scaled-down versions of the terrestrial gateways to feed their network which could be wireless or DSL or whatever.

There will be some exceptions for hooking up schools, NGOs, etc but there will be no direct-to-home subscribers to bog down this system.

»Scotty beamed up for $495 and future impact on WISPs

Alignment wasn't that difficult according to the guys that hung them. Tighter than a simple panel but being plumb and on the right inclination are key I think. The one link that I saw a throughput test on was about 2.2Km and doing ~680Mbps real. I'm sure further is possible at lower rates.Sorry - to be more accurate; we've erected our own. Only actual design and build we do is when mounting to walls etc. Otherwise just erecting pre-engineered towers.

BTW :I probably missed it but where is 'Here' ?