Mimosa high gain CPE, and lower price


#1

I would ask to Mimosa two things I believe too much important for every WISP.

Essentially, is really not useful to have a CPE with only 20dB gain antenna. It’s really useful into micropop scenario? Yes, of course, but there’s another problem: the horizontal width is 30 degrees, so inevitably in a way or other, your total throughput could be lower by interferences, or into best case, you will have problem with A5’s micropop nearest AP in SRS TDMA config.

I believe that the best option could be a single 20 dB CPE with only 10-12 horizontal degrees beamwidth. More to cover long distance customers, a 25dB integrated CPE with similar or less horizontal beamwidth. Best should be 5-7 degrees.

The last but less important is the prices. On A5’s AP I don’t consider a great option to spent about 800 euro on omnidirectional AP that cover only 600 meters in diameter. I can consider this price, for a A5c, but for me regular A5 price should be a 400 euro.

You need to compare with major competitor like Cambium. Today I can spend about 1000 euro to buy a ePMP 2000 radio+90 degrees sector+Smart Beamforming Antenna with all brackets, power supply and cables. To buy a similar A5c I need to spend about 900 euro only for radio, then I need to choose from other vendor what antenna to buy. About 500 euro for KP Perf antenna, or almost 160 euro for MTI or RF Elements. But those antennas cover only 65 degrees not 90, so I need to buy more AP to cover 360 degrees.

Last is the C5 price. I consider really too much to spend 100 euro for a 20 dB antenna 2x2:2 MIMO. And C5c seems to be even more in prices when it need to buy an external antenna from other vendors.

Even in this case it needed to compare with market: eForce 180 cost about 75 euro with insufficient 18 dB Integrated antenna, but eForce 200 cost only 105 euro with a perfect 25 dB Integrated and rock solid antenna.

I think Mimosa that have a great future, should consider to design a better hardware configurations for AP and CPE, and lowering prices to become a real breakthrough into the Wisp market.


#2

I also sing up for the same petition.


#3

In the 5Ghz band it means that beam width is in direct relation to size. If you need a very small beam, you talk a very big antenna, and high gain. It’s physically impossible to make a nice small antenna with high gain in a narrow beam. Just physics…

I do think Mimosa indeed should develop a bigger antenna for bigger distances. Many wisps work in the 1 to 10km range and a 24-26dBm antenna would be a real high seller.
Not for the micropop development, but for the macropop its needed.

Would macropop (2-10km?) be something for Mimosa? YES. An A5c combined with either a beamforming 4x4 sector or 2 x 2x2 H+V pol. sectors would throw an amount of signal combined with a 24-26dBm antenna that could do that job.
And it would do that job much, much better than what is in the market now…

The capacity of an Mimosa in the combination of new software, new chipset and high cpu power makes it a winner compared to exiting WISP solutions.

But now we come to pricing…

To pay nearly a thousand Euro/Dollar for a high, very high capacity AP is maybe more then the competition. But the AP is therefore also much better then any competition.
And a 1000 €/$ AP is an investment for the many clients on that network. So per client it ain’t that bad…

But now the C5 client units…
I think indeed they should come down in price. For new networks in macropop or rural regions it has to compete with product that are sometimes only 25% of the price of a Mimosa.
And if it comes to migrating… it runs in big, BIG papers.

Imagine you have a network of 1000 clients served by 30 AP’s. 30-35 clients per AP. Spectrum over usage hence max speeds to offer to clients in the range of 10-20Mbps, if you have done a lot of fine tuning…

Now, these 30 AP’s can be changed by some 20 Mimosa AP’s. So an 20 x 1000 = 20.000€/$ investment.
That’s bearable and you’d increase the network such that now you can offer your clients 20-40/50Mbps of speeds. And less AP’s so a bit more spectrum freed… Anyway; a 200% to 300% increase of capacity.

Now we want to go to the maximum possible… we want network sync and offer 100+Mbps…:

1000 clients times 110€/ (you need the power adapter too!) makes an investment of 110.000€ !
What do we get for this 110k? Well, network wide sync. and thus maximal frequency reuse (Only within YOUR network, the competition is still using it and probably will steal the spectrum you just freed…).
What is the advantage of this all? We could now offer 50 to 100Mbps to our clients…and maybe more…

Is this nice? Yes it is. Can we sell it? No, not for much more money. So is it affordable? No, not in the present market.
Does the client want it? Hmm, some nerd will. But most clients are absolutely happy with 10 to 20Mbps. They might see a need for 30 to 50Mbps so they can watch HD streaming TV. But 100/200Mbps??? What for?
Is the client willing to pay a lot more for what he already has (20Mbps). In my market (Spain, Southern Europe) No.

The telco and political world the mainstream talks are about high speed internet and that people are crying for that. “Everybody 100Mbps!”…Reality is that even in rich Europe the bulk of the clients don’t want to pay for it…
30 to 40€/ per month is really the maximum that connection might cost. If VAT taxes are taken away the operator needs to do it all for 25 to 35 €/ per month. That’s ok for the big holdings that are ruling the cellular and fixed cable networks, but for most of the private owned of medium sized WISPS this is barely enough to survive.

In other words, a WISP that spend the money on the AP’s only we still increase the network capacity enough to make a lot of customers happy. He stands out against the competition and he can maintain almost similar rates as our present network. We just transfer existing clients to the new AP…

But if we go for the extra offering of Mimosa network sync with tdma and their possible 200+Mbps offerings we need to invest almost 6 times as much money while we have no tools to earn that back (ROI) in reasonable time…
We cannot even put a new product in the market with higher price tag for the higher capacity. We are already squeezing the max income from the market. Throwing loads of money at your network in the hope you could get a handfull internet speed freaks onboard willing to pay the top price is unrealistic dreaming…
And even worse is, that to offer that high speeds, you need to make the investment for ALL clients, most are not willing to pay for something they are already happy with. Even although you try to tell them it gets better…

Most business need to justify their investment towards their money loaners too. How to sell a 6 times higher investment without any argument it will give you a better ROI then just the investment in the AP’s only?

Many other business just have limited funds to use.
My business for instance could spend this year enough money to upgrade all AP’s into Mimosa.
But to get then all clients into C5’s would cost me the next 5 years to get that money.

So unless the prices of C5’s come down, I am not going to invest into them. We will buy the A5’s and work in interop mode until that day in the future we really need to go higher or we have the money available.

Down sides are off course we have less tools to save spectrum (we still do a little since we need less Mimosa AP’s compared to other branches).
Biggest downside for Mimosa is they will not see high numbers of C5’s going over the table. Maybe their production lines are not up to it but its proven already (ubnt) that its better economics of selling many millions of relative cheap CPE’s than try to earn more per product in a high priced basic product.

So my advice would be; Increase the price of an AP by some percentage, but decrease the price of a C5 (C5c) considerably and come up with an C5LH (Long Haul) of 24dBi.

If they are not doing this, its only a matter of time some manufacturer (UBNT) comes with some sort of same technology as Mimosa for half the money and put Mimosa back in the niche were companies as Mikrotik and eCambium are now…


#4

I would like to hear some official position regarding price matter.

Can you show us your official think ? @Chris


#5

I would like to add that similar problem is with Cambium because for example eForce 200 costs 110,00 euro. Same things or worst than Mimosa C5. But Mimosa C5 need the PoE adapter, so you are forced to buy a PoE customer Modem/Router or G2.

How much G2 ? Simply around 50 euro. So total cost is 150,00 euro per customer. And if you would like to sell VoIP. Weel you need to change modem/router and actually no one support PoE on its WAN port including 2 FXS ports. Changing idea, simply buy the PoE for around 11,00 euro. So in worst case, you have 110,00 euro per customers excluding modem/router VoIP. Really a bad things.

Cambium for example have introduced Elevate. It seems situation is slightly better here:

Elevate license @ 35,00 euro + PowerBeam M5 300 (22dBi antenna) @ 60,00 euro = 95,00 euro
Elevate license @ 35,00 euro + AirGrid M5 HP 5GHz (23dBi antenna) @ 45,00 euro = 80,00 euro
Elevate license @ 35,00 euro + AirGrid M5 HP 5GHz (27dBi antenna) @ 54,00 euro = 89,00 euro

Probably you could have better lower price, if you decide to buy 10 Elevate’s license one time, passing from 35 euro per CPE to about 30 euro per CPE.

Last but not least, in future Elevate Firmware version, it should support Mikrotik. Consider that best selling product here is LHG 5 that have a average cost of 40-45 euro.

Considering such prices, Mimosa surely should redesign their CPE and lower price to compete on market.


#6

Those are all old N devices. Mimosa sells AC.


#7

Do you use AC on customer’s sites? And for what?

AC is a good protocol for PTP not for PtMP where spectrum is crowded and you can’t use more than 20 Mhz channel. If you can set 40 Mhz or 80 Mhz spacing in PtMP, you’re very lucky environment.

In Urban Scanario, you should create tens of MicroPOPs synced, but the problem is always CPE costs.


#8

I dont see your point here. If you want cheap high gain CPE which works on N protocol, look elsewhere. If you want high capacity and stable AC PTP or PTMP solution, buy from mimosa. Its that simple.


#9

Have you read previous consideration ?


#10

Nikolay; Massimiliano; The truth is as always in the middle. Many, many WISP are looking for a way to go into the future. Meaning higher speeds to the clients at reasonable prices.
One of the fist issues is that due the introduction of proprietary run tdma in the last 8-10 years most WISP’s are bound to one product manufacturer.
Most of these manufacturs are basically at the top of what can be squeezed out of ‘n’ without network wide sync in overcrowded spectrum.
‘ac’ can be a solution, but if the network is not going to be fully syncd the interference issue is still there and indeed, Mikrotik, UBNT and eCambium are not there yet of haven’t matured their ‘ac’ enough to have a stable system.
In the meantime competition is fierce nd where everybody is squeezing already the max out of their capacities and spectrum (max 40-50 per AP and 30-40Mbps max) the battle field is at the price front. Income per client has eroded to such extend that most WISPs hardly can afford to make major investments into new stuff, they need to find investors. And investors want a nice future. Something new, something much better then the competition…

So WISPS are looking elsewhere. And if the WISP is looking for something new, it should be a major leap forward. Otherwise it’s not worth the move. Otherwise he won’t find the investors to help him…

UBNT has still issues with ‘ac’ according their forums and once gone for ‘ac’ you’ll find it doesn’t even do ‘interop’ plain 802.11 anymore… you’re tied to them even harder.
Mikrotik has ‘ac’ but that’s it. It seems their Wireless platform lacks real innovations for WISPS and there are hardly improvements and no news in the pipelines. (Maybe next weekend in Milan will surprise us!)
eCambium tries to conquer back what they lost in the past (after all, it was Motorola) and now came with what looks a smart move in the ‘Elevate’ solution. But you’ll pay a license fee, you’ll lose all updates from original builder and guarantees with it…while still working with the ‘old’ manufacturers devices. And what I have been reading in forums its not a ‘smooth’ process. And indeed, we’re still talking ‘n’ protocol.

802.11ac protocol any good? YES. You can run 80Mhz wide channel to offer much higher speeds to the clients and those ‘n’ units still around just use the main 40Mhz channel. The other (yes, even 3rd party CPE’s) that run ‘ac’ and run at 80Mhz but occur interference on one of the two 40Mhz channels stop using the one with interference. This way it becomes possible to use 80Mhz on the AP and those units that can will use it, those that cannot won’t. The nice thing is that you can overlap your 80Mhz with another AP at distance. As long as the signals of one is some 40dB lower than the other you can run two AP’s in the 80Mhz where one has its main channel on the high end and the other on the lower end. Since we all work with directional CPE’s the bulk of the CPE’s will have a considerable (>35-45dB) difference in signal between its main AP and the other he doesn’t want… And if not it will just omit the channel from the not wanted AP to speak on its main channel with the other in 40Mhz…
This is the theory. I tried something in that direction; In my 63 clients AP-14 network in ‘interop’ mode we had a 80Mhz channel set where the secondary channel has a 20Mhz overlap with another AP (tdma). 70% of the CPE’s (all Mikrotik SXT’s) with ‘ac’ connected fine with 80Mhz and 80-100Mbps to these was not an issue. Some I could see now and then to drop back to 40Mhz channel. All the other SXT’s in ‘n’ mode where connected to the primary AP’s channel and could get 50-70Mpbs depending on their conn. rates.

Mimosa just has an ‘interop’ mode costing you nothing if you don’t want to move 100% to them and still get already a major improvement in you WISP network against relative reasonable price (Only the AP’s).
It works twice to three times as good as a Mikrotik or Ubnt tdma network and probably once as good as eCambium.

But it the industry is it all about money. Both for the WISP as for the manufacturer. Many WISP’s cannot not justify the high investment in the full change-out of its CPE’s towards C5’s just to get the very high capacity network Mimosa seems to be able to do. You can stick with the ‘interop’ mode but then can’t pick the biggest Mimosa fruit of all; network wide sync with spectacular frequency re-use.

Mimosa on the other hand needs to grow fast to come out of its position of a nice niche product providers. They say they step in the leap left between fixed cable and cell operator with either their high prices (which aren’t that high in Europe it seems…) or their high throughput promises which in reality are only rarely obtainable.
Sure there is a market to compete with fibre in the US (don’t see it here in my country) but at a world scale this is marginal. The big bucks are to be earned in a worldwide revolutionary system that is affordable. And like I showed, “affordable” is just where the shoe doesn’t fit yet… Changing out a full WISP network’s clients into the many C5’s with their power adaptors (or their very good G2’s) is very hard to justify.

My point is if the Mimosa C5 could be dropped in price by some 50% (That’s still more then a SXT or UBNT equivalent. We need a 48V adaptor and/or the wifi in the process) that would boost the sales a lot. I for one would then go for them. And my decision making won’t be very difference from many others that look with a eye to what Mimosa is doing…
And in the process it would new startups or ‘beating fibre’ even more easy! Imho this could made them a overall winner.

But don’t wait too long… Others see what Mimosa is doing and copy cats are fast these days…


#11

Off topic, but the Cambium cnPilot R200P/R201P routers do exactly this. Not sure the POE type though.

Rich


#12

R200P/R201P don’t support PoE port


#13

I beg to differ: http://www.cambiumnetworks.com/products/wifi/cnpilot-home-small-business/


#14

Sorry. Yes it support PoE on P versions.

It can power PMP and ePMP CPE. These are Power specification of Cambium Power Supply:

http://community.cambiumnetworks.com/t5/ePMP-Installation/ePMP-PoE-Powering-Primer/td-p/49944

I believe Voltage output is 30V with this configuration:

ePMP Connectorized
Compatible with
Cambium PoE pinouts (V+ = 7 & 8, Return = 4 & 5)
10 – 30 V

ePMP Integrated
Compatible with
Cambium PoE pinouts (V+ = 7 & 8, Return = 4 & 5)
10 – 30 V

So I don’t believe it’s good for Mimosa C5 and C5c or not ?


#15

Agreed. The pinouts can be compensated (I think Tycon do a cable to swap pins, or manually swapping wires in the RJ45 crimp), but the voltage is a problem.


#16

I don’t understand why manufacturer doesn’t agree to use PoE 802.3af or
another standard to avoid that problem to their customers.

Massimiliano Iavazzo
Responsabile tecnico
Telecomunicazioni srl con socio unico
sede legale: Centro Direzionale di Napoli is A7 - 80143 - Napoli NA
sede operativa: via G. Verdi 48/A - 81031 - Aversa CE
P.IVA: 07136161218
t: +39.081.19309465
f: +39.081.19758072
www.connecting-peoples.com
supporto@connecting-peoples.com
Rispetta l’ambiente. Stampa questa email soltanto se è estremamente
necessario.