Hi i need more than 300 Mbit/s full duplex over distance ~ 5 km , the best thing i have accomplished with microtic ~ 200 Mbit/s half duplex ( 4 x CM9 and 4 antenas running in one box , P4 - 3Ghz )
I read about Ceragone , Ortogone ..smth else cheaper

Did you try trubo+nstream? I think you can use 8 minipci in bounding mode but you may have sau distrosions! But I think that you need licensed frecuency, you may try 3,5Ghz

dual nstreme

Trango GigaLink 18ghz with 2ft dishes. Will do 300Mbps full duplex (600Mbps total). Cost is about $20,000USD.

Have you tested one of these travis?

Scott

300Mbit/s is meaningless with radios. Rent a fiber if possible or forget the job...

you can buy radios capable of 300Mbit/s today, but you have no headroom for future. And for spectrum licenses, you'd pay more than for fiber not speaking about limitless expandability of fiber and immediate 1Gbps availability [although you don't need 1Gbps right now]...

fiber. fiber. fiber. no other option, sorry. 300Mbps is just plain too much. Either you want to have stable business or you want to fiddle with radios. Forget small things like Mikrotik, Star-os, Dlink etc. Forget it.

I got 71Mbps with 2pc p3 @ 866Mhz, 2xR52, 5Ghz Turbo! Good luck

fiber. fiber. fiber. no other option, sorry. 300Mbps is just plain too much. Either you want to have stable business or you want to fiddle with radios.

...and some places will never ever get fiber, so they have to manage with the second best.

Forget small things like Mikrotik, Star-os, Dlink etc. Forget it.

Funny comparing Mikrotik with DLink!
DLink is plastic fantastic SOHO gear, do you place Mikrotik in the same bucket?

[not a sales pitch, the following is generally true in the industry]
High speed links:
Licensed Microwave - up to 622Mbps FDX net throughput, Gigabit Ethernet interface
70GHz Millimeter Wave - 5km is possible (but not in tropical rainzone), 1Gbps FDX net throughput
MT-based radios, 77Mbps FDX Nstreme2, and you *could* aggregate multiple links as people have demonstrated and suggested, the downside being uses a lot of spectrum, and it needs to be *quiet* spectrum across all those bands to work at full speed.
You can't use FSO at 5km - 4km is the recommended max.
Moto Orthogon is the only OFDM radio AFAIK that offers 300Mbps, but that's HDX and "flat out" (real world may differ), and not cheap.
AFAIK there are some newer 5.8GHz radios out there that do high speeds, but I'm not sure about 300Mbps+

The MT-based radios would be the cheapest, even with a pair each end aggregated. Use high gain, high quality antennas. However, the issues mentioned by others apply.
Conversely the Licensed Microwave the most solid at that distance. They use 128QAM modulation which is why it's spectrally more efficient than other RF options (e.g. Atheros cards, 64QAM max), and being licensed you don't (or shouldn't) get interference. But it does cost a bit more.

Hope that's of interest

Regards

Whats is the cost of a licence?

Is it a per year thing?

For 5 km you really should put a wire on. A fiber optic will cost you less than anything else on that distance. And will save you lots, lots, lots, lots of trouble.
Telling you from my experience. And I am paying fees to the electricity company, just to have the wire hanging on their "proprety".

Whats is the cost of a licence?
Is it a per year thing?

Depends what country you're in, what freq band, what channel width (3.5, 7, 14, 28, 56MHz in ETSI)
And yes, generally, you pay per year, unless you're a huge utility, or the government itself.
I have heard of fees in the US$300 - $2400 per year range depending on what exactly you're leasing.
Some countries are not "deregulated" and the answer can vary from "no - too bad" to "pay us lots of money".

For 5 km you really should put a wire on. A fiber optic will cost you less than anything else on that distance.

Not if it's across the centre of a busy city like London, Paris, New York it won't!
.. or there's a river or motorway in the way ...

Regards

Not if it's across the centre of a busy city like London, Paris, New York it won't!
.. or there's a river or motorway in the way ...

Sorry. Thought it's common sense that....

Anyway in a major city crowd, it's less than probable to have clear spectrum to do it wireless with non-licensed radios.

You could study alternatives like lasers, seen somewhere a post about 600 mbit link with laser modulators.
( Seen it right in one of your post, right now, Stephen. http://forum.mikrotik.com/viewtopic.php ... ser#p59963 )
Or here, it sais it runs for up to 4km : http://www.airlinx.com/products.cfm/product/19-0-0.htm

Anyway in a major city crowd, it's less than probable to have clear spectrum to do it wireless with non-licensed radios.

Exactly ... most users wouldn't want to bet on 4x 40MHz channels being free (and staying clear) in many major cities in the world. Of course for a rural shot, it could work though.

Laser can do 1.25Gbps FDX today, but for 5km, you need a different technology. Microwave or E-band MMW would be my suggestion.

Regards

you're correct in your "big city" statement - but I'd say that big city has huge advantage [compared to small city / rural area] when 5km 300Mbit/s transfer is required... and that advantage would be amount of dark fiber already laid down and waiting for rent

definitely it's easier to achieve 300Mbit/5km in big city than anywhere else. Highly scalable infrastructure exists anywhere in big cities...

definitely it's easier to achieve 300Mbit/5km in big city than anywhere else. Highly scalable infrastructure exists anywhere in big cities...

Don't know about your zone, but here it comes almoust every time with "higly scalable" prices that make you wonder what's your profit then....

Don't know about you guys, but if radios won't work, and some of you consider fiber the idea of dark fiber doesn't sound so bad...

On the other hand our fellow is looking for an unexpensive solution (not cheap).

Any of you know how much for the lease of 5km (minimum if it was a straigth line) would be against owning the equipment and paying for the frequency?

reliability vs price?

just my toughts..

In our area we have no access to utility poles. Even if I wanted to run a fiber 3 blocks (1000 feet) I would not be allowed to do it - even if I was willing to pay. That has been my experience because I asked once. I git shot down so fast I haven't even revisited it yet to date.

Does anyone know of any rules in the US / FL / FL Power & Light / Brevard County that would govern the usage of utility poles to run fiber between point A and B?

I think i was told that only a licensed utility company could utilize the poles and only with an arrangement with FPL... I have heard other places are easy to deal with... I wish I could run fiber here, I would in many circumstances.

Thanks,
Scott

I'm not in the us, but here, in most cases you can rent a pole for about 1.5 usd without taxes (19%), from the electricity company.
That is every month, you pay 1.5 usd. For 40 m between poles. In most cases, they will allow you to use them, unless they are really full of other wires. Second you must pay a tax to the road administrator. That i don't know wich amount it has, it's a per /km tax, not that big, just annoying.

So, 5000 m / 40m ~ 187.5 ( 220 ) usd a month, just the tax. Ain't that much. But if you have to cross other roads, or have an intersection in the way, well, that's a different story. And you are making a fixed time period contract. Not undetermined. They "could" ask you to take your fiber off after 1 year.

Have you looked at Canobeam or other laser solutions? they can push gig speeds.

Yes, fiber is the way to go if its available. So, supply and demand dictate that often forgotten concept that bandwidth costs money. Also, identify the profile of utilization. Does he need the majority of bandwidth in one direction, or is the traffic symmetric? I can only wonder why he would need 300Mbps full duplex which aggregates to 600Mbps. Or is it 150 symmetric, 300 aggregate, or 'up to 300 at times in one direction'?

5km is no problem, squeezing more out of multiple radios, and using lots of proprietary mechanisms will only get you so far. If the customer really needs that bandwidth, there should be a linear relation between need for speed and ability to pay. If they cant pay, and they really need the speed, something is wrong with somebody's business model if they wont pay for what it really takes to do this high end task.

Otherwise, get the most out of a MT platform with NStreme-dual, the narrowest bandwidth/gain/beamwidth combination possible to get 30+dB SNR at about -60dBm at least, but not too far into the 50s. This could produce quite a bit of throughput, and my experience tells me, they likely wont need the 300. But if they must backup those multi-terrabyte servers each night before the morning shift, calculate how many hours at the best speed you can give them, and you would be surprised at what users can agree to. Gee, backup half the data every other night, you get the idea. Somehow dividing the load, and graduating to higher speeds when you can prove the concept can work for a given easy to deliver bandwidth. Otherwise, good ole sneakernet would be faster. I have already calculated and proven that.

If they are blasting 10,000 VOIP calls, nailing up 300Mbps (I have seen it), run for your life before they come to you and say 'I'll take whatever you can give me', and they nail up a hard earned 88Mbps 24x7. If its a private network, so be it, if its carrying Internet traffic, you have a lot of calculating to do. Thats why it costs money for bandwidth, because even money cant break the laws of physics. Light has a lot of 'spectrum', thats why this is fiber and laser technology territory.

Moral of the story, you get (or sell) what the market calls for. You should be able to calculate a ballpark per MB cost to you, and that in itself might reveal your solution. Might be easier to tell the customer to move 5km closer for all the trouble you have and will encounter. If fiber and licensed radios are out of range, his 300Mbps is out of range. So I would ask him what his budget is, and work back from there. You literally could have a spreadsheet, or rate card with something like, 5,10,45,100Mbps pricing. The 100Mbps, or whatever you can get as your max, is the baseline. If he still wants 300, multiply it by more than the incremental multiplier to cover extra integration work, and multiple sets of equipment to deploy.

If you try to sell him more than the radios really can comfortably deliver, you wont be comfortable when he comes a knockin on your door to tell you it isnt what he expected. So, keeping in mind the notion that this isnt really RF territory (except for those dreaded expensive links), give him something he can count on and I believe you can land that customer. You might even entice the customer with the 'future upgrades' that should be coming, just not sure when, but suggest he try what you can reasonably deliver with the best 'maxed out' config you have. If he doesnt have 300Mbps now, he doesnt know if he needs it either.

That keeps you in the comfort zone using MT configs, and keeps the reputation of wireless clean. Nothing worse than hearing later 'I tried wireless and it didnt work'. Get him happy with what works well, and keep workin it from there....

Never ever had a wireless link beat me yet. Stay away from unrealistic expectations. Dont solve problems that either dont exist or cant be reasonably accomplished. If you cant get there from here, it cant be built or sold. Identify 'suitablility for the purpose'.
Thats my take on it. Sorry for the verbosity, but couldnt resist...

300Mbit/s is meaningless with radios. Rent a fiber if possible or forget the job...

you can buy radios capable of 300Mbit/s today, but you have no headroom for future. And for spectrum licenses, you'd pay more than for fiber not speaking about limitless expandability of fiber and immediate 1Gbps availability [although you don't need 1Gbps right now]...

fiber. fiber. fiber. no other option, sorry. 300Mbps is just plain too much. Either you want to have stable business or you want to fiddle with radios. Forget small things like Mikrotik, Star-os, Dlink etc. Forget it.

We have two 1.25gbps radios, both Bridgewave. One at 60GHz (58.5/62.5) and one at 80GHz (72.5/82.5). The payback period over fibre was less than 24 months.