Are Microtik supported the gprs

not at this moment

I work in a company that offers equpment to hfc and also a wimax lately. ex. cambridge broadband networks - vectastar3500

I talked with my product manager to offer mikrotik for 5ghz as a last mile to end-user, but we still have to use 2 systems - this CB and MT - If mt would support wimax (3,4-3,6 ghz) with channel width from 1,75 do 14 mhz - would be great - stupid polish law to pay for channel license.

greetings
mcozio

stupid polish law to pay for channel license.

Not only on Polish...the same things for Italy (we are waiting to see which will buy this spectrum, the auction is expected for this summer).
But this could be not bad, becouse, if the spectrum are licensed, you will see looooooow interference on it, so, better througput on longer distances!
This could be very good for a lot of people on Italy, which is digital-divided.

I work in a company that offers equpment to hfc and also a wimax lately. ex. cambridge broadband networks - vectastar3500

I talked with my product manager to offer mikrotik for 5ghz as a last mile to end-user, but we still have to use 2 systems - this CB and MT - If mt would support wimax (3,4-3,6 ghz) with channel width from 1,75 do 14 mhz - would be great - stupid polish law to pay for channel license.

greetings
mcozio

You mean MT to support mini-pci WiMAX cards ?
I think mini pci wimax cards are not released yet ? am i wrong ?

no, MT does not support wimax yet

when mt will support wimax (802.16XX) rfc or just bands 3,4-3,6 with 3,x-1,75-3,5-7-14 step mhz channel setup ) ?

just 3Ghz bands are already supported, use Ubiquity SR3 cards. Wimax standard is not yet supported, I don't know when it will be supported.

channel width from 1,75 to 14 mhz

That's not the first time I've heard that one - some countries issuing freq licenses but stipulating the ETSI channel widths 1.75, 3.5, 7, 14MHz etc, which excludes systems built using "Atheros WiFi extension" gear using 5, 10, 20, 40MHz.

All other arguments apart, regulation is one which plays in the favour of WiMax in certain countries.

Regards

eh, wimax - 2x14mhz = 60mbps, IP traffic 40mbps - cost for link p-t-p about 10-15k $, license for this speed next 7-10k $ per year. In poland, 1 city over 100k citizents if more then 100k

compare with mt:
linkt with ns2 - 54mbps fd, cost about 1-4k depends on ideas

I realy hope MT could come up with a wimax solution.
Wimax equipment today is really to expensive.

Wimax is great though, good nlos capabilitys and great range.

cost for link p-t-p about 10-15k $

These days no-one is suggesting WiMax is the right solution for P2P links - even Moto openly says it isn't suitable.
WiMax is for P2MP distribution, with some features you will never see on a non-WiMax current-generation Atheros/MT-type system - and vice-versa: WiMax can't operate in contended (read: unlicensed) spectrum where there's any interference. WiMax can do long distance and better non-LOS, but Atheros can do higher bandwidth. A case of "Chalk and Cheese".

There IS a place for WiMax. AFAIK the MT guys are busy on other things (ROSV3, PowerPC port of ROS, 802.11n) and there are no open-source WiMax Linux drivers either yet, so the only suggestion is, "vote for it on the Wiki" and be patient.
Or grab a C compliler and start hacking some home-brew drivers together for some of the miniPCI WiMax cards already available....

Regards

I had mentioned this before, but what would happen if we would take 802.11n and pump up the power like Wimax, with huge Sectors like most Wimax towers. I believe we should get same NLOS. Yes, with power comes noise, and with noise comes quality loss ie throughput loss. But, how much can you lose of 300Mbps? If you deliver 8 Mbps to the client, you already beat Wimax. Remeber, Wimax is high speed download, not upload! That is due to the strong tx power of the tower and the weaker Tx power of the CPE. Best I've seen on Motorolla Nextnet is 8Mbps down and 400 Kbps up. The omni hight is arround 2 meters and the unit transmits 2 Watts at the antenna connector. Navini's Wimax radio is less powerful but with an omni almost 2.5 meters high.

I only see the long run advantage of Wimax will not be NLOS, and/or quality of service as you can do that even on the cheapest APs, but PPS. Unless 802.11n can handle the same PPS as Wimax equipment, then 802.11n based MTs will be able to bear the same number of clients as Wimax - Not even Nstreme or disabling CSMA-CD would help. Have read before in the forum, the best Atheros mini-pci at this stage can handle only 8000 PPS. That is why most Wimax units can handle many clients, ideal 150, low speed 300 and theoretical 600.

Yet let us see the cost of 802.11n. If each can handle 50 clients, then we need 3 radios and 9 antennas with amplifiers which might still workout cheaper than one Wimax unit.

Anyway, there are other technologies like CDMA that MT should consider. Qualcomm states 8km NLOS. 30Km LOS.

Very interesting debate!
and some very good points there.

802.11n, AFAIK most of the performance gain is through MIMO, relying on reflected signals and signal processing to increase throughput. That works well indoors, because there's plenty of walls/ceilings/doors to make signals bounce around. I don't think the modulation rates are any higher than 802.11a/g (64QAM max), correct me if I'm wrong.
Mostly because in unlicensed spectrum with low-gain antennas, the noise floor is too high for 128QAM or more at any meaningful distance.

I've yet to see any suggestion that works in an *outdoor* environment, i.e an array of 3-4 high gain sector antennas on a mast and mostly-clear LOS to the clients, who have single (non-diversity) antennas. I'd like to hear from anyone who has tried that, even a brave experimenter pulling the back off a "pre-N" linksys/belkin/netgear, attaching outdoor sector antennas to the radios, and seeing what happens when you take the laptop outdoors at range.
Beam steering of course improves link margin, and with that, usually bandwidth because you can use higher modulation rates, but that's not giving the 300Mbps "headline figure" the 802.11n'ers claim - that requires reflected signals, not just the beam steering.

There are subtle points in WiMax to do with the way the airside signals are structured, particularly in 802.16e, with SOFDMA (the S is important) which massively helps the near/far problems inherent in outdoor wireless, which I don't think 802.11n ever intended to look at, because it's an indoor standard. In 802.16e, there's a far more flexible split between which tones get used for which subscribers, at differing modulation rates. That's one reason WiMax can do longer range, it's not just about brute force "turn up the TX power". Near and far subscribers get allocated different parts of the RF signal, with adaptive modulation to cope with varying channel conditions.

Price argument is always a factor. However, WiMax is a standard, which operators like, because it means mix/match BS and CPEs. A "tweaked 802.11n" with some version of Nstreme running in a 40MHz channel doubtless would give higher throughput, but major operators have big issues about selecting a proprietary system, even if it's half or a quarter the price.

I don't think this is a "one horse race": far from it: industry experts think you need both. Picture a base site with MT boxes with 802.11n AND WiMax cards, sector antennas connecting both fixed and mobile subscribers, automatically "handing-off" as mobile (laptop/PDA) users move between short-range WiFi and longer-range WiMax. Dual-mode WiFi/WiMax laptops? already happening. Mobile device dropped off WiMax coverage? GPRS/3G it is then. Sure, throw in some proprietary "nstreme-like" sectors on the towers for fixed users as well, so when they get back home or to a hotel, they connect back to in-building WiFi which is then back-hauled from a fixed CPE to the same or another tower at high speed. All without a break. It's not a pipe-dream, it's called "4G": seamless service. It's going to happen, like it or not. I hope the MT folk can cut enough code to do it sooner rather than later.

Regards

I've yet to see any suggestion that works in an *outdoor* environment, i.e an array of 3-4 high gain sector antennas on a mast and mostly-clear LOS to the clients, who have single (non-diversity) antennas. I'd like to hear from anyone who has tried that, even a brave experimenter pulling the back off a "pre-N" linksys/belkin/netgear, attaching outdoor sector antennas to the radios, and seeing what happens when you take the laptop outdoors at range.
Beam steering of course improves link margin, and with that, usually bandwidth because you can use higher modulation rates, but that's not giving the 300Mbps "headline figure" the 802.11n'ers claim - that requires reflected signals, not just the beam steering.

Take a look here: http://80211n.wifinetnews.com/archives/ ... speed.html

Nice one, MIPLAND - well spotted. Glenn is a very respected author.
Looks like the main outdoor improvement for 802.11n will be the MAC layer improvements.

And, the one he doesn't mention, the "beam steering" possible with an antenna array, improving the signal at the client (fixed or mobile). The 802.11n standard and chipsets allow for this mode, i.e use it in beamforming rather than MIMO modes. The link budget improves, so you get better signal at the clients (and uplink too), so higher modulation rates, hence higher throughput.
Now there's something AFAIK no-one has suggested in public for 802.11n:
Consider a 90degree sector. Put up 3 antennas for the sector [RF gurus please start thinking now, comments please], say each is a 45 or 60 degree beamwidth, with overlapping lobes to cover the whole 90 degrees.
The 802.11n card has each radio connected to a separate antenna, and radios are used in beamforming mode.
The signals are "steered" towards the subscriber, using some clever (probably already defined in the standard) algorithm. The uplink benefits similarly, as the BS receivers act together to increase the effective receive antenna gain.
Range improvement? for sure. Higher throughput per client? for sure.
Lower interference on uplink? also true (the beamforming "nulls" out signals coming in from "irrelevant" directions, increasing SNR)
-> "Increase cost of the BS justified?" 3x the number of sector antennas, and a more expensive atheros card. Under $1k per sector increased cost? For a busy tower, or one reaching capacity, probably costs in easily.
The best bit? You could start that sector with 1 antenna at close to a conventional 1-radio cost (only the smarter radio to pay for, which does have the MAC advantages/improvements), and upgrade later by plugging the other antennas in.
Or: petition some bright antenna manufacturer to build 3 antennas into a single housing, with 3 N connectors on the back.
I bet at least one of them is going to be very happy to do that.

To show we're not OT: Note the 802.16e WiMax crowd are all doing this: smart antenna arrays on the BS, and even diversity on subscriber units, all to improve link budget and range.

A few links below for background reading.

Food for thought ...

Regards

http://en.wikipedia.org/wiki/Smart_Antennas (mentions 802.11n specifically)
http://en.wikipedia.org/wiki/Beamforming
http://en.wikipedia.org/wiki/Multiple-i ... unications
(under categories of MIMO, it's the Precoding and Diversity coding that can be used outdoors.
Spatial Multiplexing modes, which give the huge bandwidth rise, probably can't be- they rely on indoor reflections.
Note that Moto/Orthogon use the Diversity coding on their high speed Spectra P2P links. It does work.)

Thx to all of you for gr8 inputs and Stephenpatrick for corrections. Appreciated.

Or: petition some bright antenna manufacturer to build 3 antennas into a single housing, with 3 N connectors on the back.
I bet at least one of them is going to be very happy to do that.

or another rather creative way to approach this would be to put up two 60 deg horizontal polarity sectors (spaced to cover 90 degrees) and one 90 degree vertical polairty sector, and then use a dual polarity panel CPE and attach one connector to vertical and another to horizontal, you get almost the same beam steering ability on the horizontal, plus the ability to jump to vertical polarity on a moments notice to help avoid intermitant interference on the horizontal channel, not to mention there will doubtless be times where due to some odd refraction off a roof or passing truck or other enviromental factor, you actually get a stronger/cleaner signal on the vertical from the horizontal transmitter (I've seen odd spikes while aiming dual polarity dishes before where a wrench or something throws a weird refraction and the H-Pol link just saw the V-Pol link signal jump 25db for a brief moment)

This is something we have wanted to try with diversity for some time (using just one dual polarity sector, and to try it both with and without a dual polarity CPE), but MT never gave us diversity support.

Navini BS has 1.5km NLOS indoor penetration range. And a total of 20Mb/s throughput.( Air frame,not ethernet level).

Think about VoIP, broadband and Streaming media on a single Navini BS for 20 customers. I dont think it survives too much.

and you pay it 250,000 USD !!!