Community discussions

MikroTik App
 
nicopretorius
Frequent Visitor
Frequent Visitor
Topic Author
Posts: 77
Joined: Mon Nov 15, 2004 9:49 am

Wi-Max miniPCI support

Mon Apr 24, 2006 12:03 am

Are there any Wi-Max miniPCI adapters such as the (http://www.wavesat.com/products/mini-pci.html) for which support is available or planned?
 
jober
Long time Member
Long time Member
Posts: 692
Joined: Fri May 28, 2004 12:16 pm
Location: Louisiana,USA

Mon Apr 24, 2006 7:20 am

This would be nice!
37.5mdps with 10mhz bandwidth.
 
User avatar
stephenpatrick
Forum Veteran
Forum Veteran
Posts: 703
Joined: Fri Aug 20, 2004 12:26 pm
Location: UK
Contact:

Mon Apr 24, 2006 10:59 am

It's been asked for ... no sign yet.

AFAIK for most hardware if there is good driver support in Linux 2.6, it could be included in RouterOS. No driver = less likely.

Also, that card is 3.5GHz only. So you'd need a license in most countries.
CableFree - Wireless Excellence - Microwave, E-band Radios, Free Space Optics, High performance Radios & Routers
http://www.cablefree.net
 
jober
Long time Member
Long time Member
Posts: 692
Joined: Fri May 28, 2004 12:16 pm
Location: Louisiana,USA

Tue Apr 25, 2006 9:11 am

They have a 5.8 card too.
 
User avatar
stephenpatrick
Forum Veteran
Forum Veteran
Posts: 703
Joined: Fri Aug 20, 2004 12:26 pm
Location: UK
Contact:

Tue Apr 25, 2006 1:46 pm

Hi again,

I didn't see a 5.8GHz miniPCI on the website - the "5.8 GHz Reference Kit" is a split digital/analogue pair of PCI cards - did you get more information from them i.e. more than that on the website?

Personally, I think this is the "way to go" - supporting this and future WiMax cards for other bands - all the WiMAX systems on ther market are "closed systems" from large or startup vendors.

Regards
CableFree - Wireless Excellence - Microwave, E-band Radios, Free Space Optics, High performance Radios & Routers
http://www.cablefree.net
 
mgm@protenus.com
just joined
Posts: 3
Joined: Sat Apr 29, 2006 5:53 pm

Sat Apr 29, 2006 5:55 pm

I would also like to add my "vote" :wink: to supporting wimax cards, both in unliscensed and liscensed spectrum. The combination of mikrotik and wimax cards, could be a real killer.
 
matt
Member Candidate
Member Candidate
Posts: 123
Joined: Thu Jan 27, 2005 9:29 am
Location: Canterbury, New Zealand
Contact:

Sat Apr 29, 2006 6:11 pm

Bring on WiMAX :)
 
variable
Member Candidate
Member Candidate
Posts: 217
Joined: Wed Apr 13, 2005 4:36 am

Sat Apr 29, 2006 10:43 pm

wimax would be the coolest. definatly keep my buying from mikrotik in the future!
 
pdf
Frequent Visitor
Frequent Visitor
Posts: 92
Joined: Sun Feb 12, 2006 11:56 pm

Tue May 02, 2006 2:05 am

The wavesat product looks more like a developer release (i.e. a reference design) which means it's not (I guess) a final product, but a pre-production release in order to do some testing. Anyway I am waiting for wimax on both licensed and unlicensed frequencies, hoping that Mikrotik will support this one together with IPv6 ;)

Thank you!
 
User avatar
aitsecurity
Frequent Visitor
Frequent Visitor
Posts: 84
Joined: Thu Mar 16, 2006 12:28 am
Location: venezuela

Re: Wi-Max miniPCI support

Tue May 02, 2006 4:54 am

Are there any Wi-Max miniPCI adapters such as the (http://www.wavesat.com/products/mini-pci.html) for which support is available or planned?

Wimax, i am thinking is the name for marketing,

wimax 802.16 need the same L.O.S, and use license bands :-( only for very big WISP, who pay the regulatory ,etc the very big PSTN ,etc

now, i will buy wimax, when no need L.O.S in wimaxforum.org, you see, is the same modulation OFDM, only change the bands.

no way, for me.

who here make a NLOS with wimax ?? who test, i am thinking wimax is WIFI with other name.

Best Regards
 
pdf
Frequent Visitor
Frequent Visitor
Posts: 92
Joined: Sun Feb 12, 2006 11:56 pm

Re: Wi-Max miniPCI support

Tue May 02, 2006 11:53 am

Are there any Wi-Max miniPCI adapters such as the (http://www.wavesat.com/products/mini-pci.html) for which support is available or planned?

Wimax, i am thinking is the name for marketing,

wimax 802.16 need the same L.O.S, and use license bands :-( only for very big WISP, who pay the regulatory ,etc the very big PSTN ,etc

now, i will buy wimax, when no need L.O.S in wimaxforum.org, you see, is the same modulation OFDM, only change the bands.

no way, for me.

who here make a NLOS with wimax ?? who test, i am thinking wimax is WIFI with other name.

Best Regards
probably you are right, WiMax is the last hyped product: I don't know if it will have better/same performances than what we use today. But I don't agree with the fact I need the WiMax standard (802.16) for many reasons (mainly commercial reasons) on both licensed and unlicensed frequencies.... ;)

Thank you
 
pacman
Frequent Visitor
Frequent Visitor
Posts: 80
Joined: Sat Jan 29, 2005 11:46 am
Location: Gloucestershire, UK
Contact:

Tue May 02, 2006 10:47 pm

Although the basic modulation scheme is the same, OFDM, I think it uses many more sub-carriers than 802.11a etc and as such could well offer better NLOS performance. Time will tell of course!
Loop Scorpio Ltd
providers of rural broadband
 
jarosoup
Long time Member
Long time Member
Posts: 600
Joined: Sun Aug 22, 2004 9:02 am

Tue May 02, 2006 10:50 pm

We have tested some WiMAX gear from Airspan (ASMAX) operating in the 3.5GHz band with a single 8 or 10dB omni (don't recall which it was). We had diversity disabled (attenuated) and the transmit power on the primary was 33dBm. We were getting about 1km NLOS in an urban environment (inside buildings), with throughput ranging from 4-10Mb. The NLOS performance did vary quite a bit depending on where we were. I know that there's no way we could get these kinds of links with 2.4Ghz or 5GHz bands as we have tried at this same location (using CM9 cards, no amps) without much success.

It would be interesting to see Mikrotik support this new card and see how it does :)
 
User avatar
aitsecurity
Frequent Visitor
Frequent Visitor
Posts: 84
Joined: Thu Mar 16, 2006 12:28 am
Location: venezuela

Wed May 03, 2006 3:32 am

We have tested some WiMAX gear from Airspan (ASMAX) operating in the 3.5GHz band with a single 8 or 10dB omni (don't recall which it was). We had diversity disabled (attenuated) and the transmit power on the primary was 33dBm. We were getting about 1km NLOS in an urban environment (inside buildings), with throughput ranging from 4-10Mb. The NLOS performance did vary quite a bit depending on where we were. I know that there's no way we could get these kinds of links with 2.4Ghz or 5GHz bands as we have tried at this same location (using CM9 cards, no amps) without much success.

It would be interesting to see Mikrotik support this new card and see how it does :)
mnmm now, the price of the product Airspan, compare vs 802.11x standars , is very expensive?

i live in Venezuela i am sure only the very big PSTN make regulatory , the normal Wisp, is very hard because the little wisp no have founds for pay license bands, ufffffffffffffffffffffff very expensive. ??

talk me about the price of airspan for the base station and for CPE.

thanks very much
 
jarosoup
Long time Member
Long time Member
Posts: 600
Joined: Sun Aug 22, 2004 9:02 am

Wed May 03, 2006 8:19 pm

The base station is priced somewhere around $20K (USD), the indoor CPE is $400-500 (it looks like an Apple Airport Base Station), and the outdoor CPE is somewhere around $600 (this looks a lot like their WIPLL stuff). Since I don't work for Airspan and haven't purchased any of their WiMAX products, these are not official prices, but estimated costs from what we have heard - your costs or quotes may vary.

Pricing of commercial WiMax products is quite high, and as you have said, will prevent all but the largest ISPs and telcos from getting into this space. This is why I think it would be very important for a company like Mikrotik to take interest early and get some developers working on supporting these new WiMax radios. They could make a killing on both hardware and software if they had a WiMax solution. It would also open the doors for many WISPs looking to get into this space.
 
patrik.backhuvud
just joined
Posts: 6
Joined: Sun Feb 13, 2005 12:42 pm

Wed May 03, 2006 8:52 pm

The base station is priced somewhere around $20K (USD), the indoor CPE is $400-500 (it looks like an Apple Airport Base Station), and the outdoor CPE is somewhere around $600 (this looks a lot like their WIPLL stuff). Since I don't work for Airspan and haven't purchased any of their WiMAX products, these are not official prices, but estimated costs from what we have heard - your costs or quotes may vary.

Pricing of commercial WiMax products is quite high, and as you have said, will prevent all but the largest ISPs and telcos from getting into this space. This is why I think it would be very important for a company like Mikrotik to take interest early and get some developers working on supporting these new WiMax radios. They could make a killing on both hardware and software if they had a WiMax solution. It would also open the doors for many WISPs looking to get into this space.
Expensive???

How many 802.11 basestations would be required to make indoor penetration with a radius of 1 Km?? And what would that cost?
 
jarosoup
Long time Member
Long time Member
Posts: 600
Joined: Sun Aug 22, 2004 9:02 am

Thu May 04, 2006 12:27 am

I'm not sure if I understand your question...how many WiMax base stations? As I said before, we had varied results within 1 km inside and behind buildings with a single base station. As with most wireless deployments, each location will be somewhat unique, so results could be much better or much worse from site to site depending on many factors.
 
User avatar
aitsecurity
Frequent Visitor
Frequent Visitor
Posts: 84
Joined: Thu Mar 16, 2006 12:28 am
Location: venezuela

Thu May 04, 2006 4:27 am

I'm not sure if I understand your question...how many WiMax base stations? As I said before, we had varied results within 1 km inside and behind buildings with a single base station. As with most wireless deployments, each location will be somewhat unique, so results could be much better or much worse from site to site depending on many factors.
good question, i have other question

what is new about security? in wimax, is more secure, or have the same problem in wifi example mac spoofing, ,etc
 
patrik.backhuvud
just joined
Posts: 6
Joined: Sun Feb 13, 2005 12:42 pm

Thu May 04, 2006 2:54 pm

I'm not sure if I understand your question...how many WiMax base stations? As I said before, we had varied results within 1 km inside and behind buildings with a single base station. As with most wireless deployments, each location will be somewhat unique, so results could be much better or much worse from site to site depending on many factors.
to make 100% (or as near as possible) indoor coverage you need less WiMax basestations than you need 802.11 basestations. At what range did you get full coverage with WiMax from Airspan? And what was the model of basestation you used?

And for Airspan beeing expensive, let say you need 10 basestations with WiMax to cover an area while you need 200 basestations with 802.11 to cover the same area. What would be most expensive?
 
User avatar
ghmorris
Member Candidate
Member Candidate
Posts: 284
Joined: Fri May 28, 2004 12:14 pm
Location: Minden, Ontario, Canada
Contact:

Thu May 04, 2006 4:20 pm

802.16 is designed to use very predictable licensed spectrum where the WISP can predict with some certainty that their radio packets will reach the user.

TDMA without ARQ causes all kinds of problems with throughput, latency etc when used in a noisy enviroment. That is one of the key reasons WiMAX is initially deployed in licensed spectrum.

802.16 that will be suitable for a randomly noisy unlicensed environment will need ARQ and more flexibility with scheduling inside the MAC. That in turn will cause problems with QoS which removes one of the key benifits of WiMAX in the first place.

Read here for a first class discussion of the issues. It does come from a vendor, but regardless its by far the best analysis I've ever seen.
http://lists.wispa.org/pipermail/wirele ... 24928.html

George
 
jarosoup
Long time Member
Long time Member
Posts: 600
Joined: Sun Aug 22, 2004 9:02 am

Thu May 04, 2006 8:35 pm

aitsecurity,

I think the security of a WiMax base station will really be determined by the vendors who provide the equipment. There are some registering functions for CPE and base station, but I honestly don't see reason why you couldn't spoof a MAC if you had a CPE that allowed it and enough information about the network. Additional authentication measures could be added to the CPE in order to authenticate it on the network itself (or the computer/network behind it) to prevent alien CPEs on the network. Perhaps there is someone on the forum that could answer this better than I. Our time was limited and we didn't get to explore everything.

patrik.backhuvud,

The base was called AS.MAX. As I said before, every location is different - the results we had from this test may mean nothing to a location you are interested in trying. Not trying to be difficult, but I don't want to mislead you either. Some places we tested (outside, behind buildings) worked great and some of the same distance and similar blocking didn't work so well. Indoor results were the same as far as variability. we tested out to about 1.5km. Just because we can get a full link 1km out at one location doesn't mean that a 1km radius will have that level everywhere. Buildings, bounces, construction density and materials, or even interference if using unlicensed all come into play. All these comments are based on NLOS. Obviously, LOS is a lot more predictable and one can guess-timate or calculate how a link should behave. We were really only interested in NLOS links as so that is all we tested. The key is being able to bounce off of another structure - similar to doing NLOS with OFMD. Overall, we were impressed with the results. We hope to revisit this test using diversity as well as switching from omnis to sectors.

As for pricing, in your example, it would be the same if WiMax base = $20k and 802.11 base = $1000. This wouldn't account for the cost of CPE though. But again, there is no one "static" formula for NLOS coverage - there are many factors that come into play and ever locate is different.

george,

Thanks for that link - interesting read. It's unfortunate that the 802.11 folks don't sound very cooperative in coming up with a joint interference solution. Luckily, we were able to test in licensed frequency. I have no doubt that our results would have been bad at best had this been in the unlicensed 5GHz band. The area we tested is pretty much saturated with 2.4, 5.2, and 5.8.
 
User avatar
aitsecurity
Frequent Visitor
Frequent Visitor
Posts: 84
Joined: Thu Mar 16, 2006 12:28 am
Location: venezuela

Fri May 05, 2006 5:38 pm

I'm not sure if I understand your question...how many WiMax base stations? As I said before, we had varied results within 1 km inside and behind buildings with a single base station. As with most wireless deployments, each location will be somewhat unique, so results could be much better or much worse from site to site depending on many factors.
to make 100% (or as near as possible) indoor coverage you need less WiMax basestations than you need 802.11 basestations. At what range did you get full coverage with WiMax from Airspan? And what was the model of basestation you used?

And for Airspan beeing expensive, let say you need 10 basestations with WiMax to cover an area while you need 200 basestations with 802.11 to cover the same area. What would be most expensive?

mmm give a idea.

for example in Venezuela, the first techonology for mobile cellular phone, was AMPS, later NMAPS, now digital CDMA, TDMA, and GSM


when the operators of mobile phone, use AMPS and NMAPS, (analogue), need high tower , and put only one and other , the density of celular tower is minimal.


now i can see, when use digital CDMA, GSM, or TDMA, put more tower and the tower is very short.


now why?, simple, the operators tell me, in the old system AMPS, or NAMPS, have bad concept, because think about the cost for put a cell (tower and radios ,etc), now put more cell and put more short the towers, why, simple, the operators can give "LESS" coberture and the cell can management more QoS , and the operators mobile can get more users.

i am thinking the wimax, no is the solutions, for CPE, or residential CPE, no now, i will prefer put more AP 802.11a or 802.11b.

is more work, but? is the same concepts of cell or mobile cellular phone.

Best Regards
 
simonkizi
newbie
Posts: 45
Joined: Mon Jan 30, 2006 10:38 pm

Fri May 05, 2006 7:23 pm

We had a meeting with Personnel from an Italian Company called Navini. They install real Wimax systems. I believe it is 802.20.

30 Km NLOS at only 70mw 3.5 Ghz
80mbps throughput
Cost per tower equipment in South Africa arround $100,000

Cheers
 
User avatar
stephenpatrick
Forum Veteran
Forum Veteran
Posts: 703
Joined: Fri Aug 20, 2004 12:26 pm
Location: UK
Contact:

Wed Jan 31, 2007 1:27 pm

Very interesting Martin -
Thanks for your input. One note:
mikrotik is not opensource or linux based
Current MT 2.9 based on Linux 2.4, the new 3.0 is based on Linux 2.6 from what I was told. But correct, it is not open source.
I think the cost of the SDK, lack of an opensource Wavesat driver, and internal priorities at MT are the limiting factors.

From a previous comment
30 Km NLOS at only 70mw 3.5 Ghz
80mbps throughput
Not at the same time ! Be careful of Intel Hype and vendor fibs. At max distance, or non-LOS, the bandwidth comes down to a tiny fraction of that.

Regards
 
hebeda
newbie
Posts: 36
Joined: Sat Jun 17, 2006 8:12 pm
Location: Leipzig, Germany
Contact:

Fri Feb 09, 2007 8:55 pm

wimax will start to be VERY interesting with new 5.8ghz WAVESAT cards, massproduction will start middle of this year i got told by a manufacturer first hand.

but problem with the software support remains.

i have learned that the WAVESAT linux software development kit doesnt include any basestation or client drivers.

so basically whoever buys the SDK has to develop the driver on his own.

i have gotten one offer so far for a WAVESAT SDK incl. clientmode linux driver, but the problem with the Basestation mode remains.

Driver development cost some serious money ... and talented programmers are very rare in this field.


My personal conclusion is: with 5.8GHZ wimax and working linux drivers, the classic 802.11abg wireless Networking will die slowly within the next 1-2 years....

i doubt mikrotik will do the investment anytime soon to implement wavesat drivers into their OS, due to the very high cost of development ...
 
User avatar
normis
MikroTik Support
MikroTik Support
Posts: 24496
Joined: Fri May 28, 2004 11:04 am
Location: Riga, Latvia

Mon Feb 12, 2007 8:52 am

we will work on wimax when there are drivers and documentation available.
 
0ldman
Forum Guru
Forum Guru
Posts: 1446
Joined: Thu Jul 27, 2006 5:01 am

Sat May 12, 2007 8:24 am

Not to sound too terribly negative, but I'll believe it when I see it. Every big company has PR reports to talk about how they're the best and supporting the GNU community, small businesses, etc...

Very rarely is it more than a PR gimmick.

I do have my questions about 802.11n, especially in rural areas like mine, multipath will be an issue.

On that same note, will they really outperform what we have now?

At this point, I put more faith in extended range, as I've seen it work, than 802.11n in long range shots.

As someone said on this board, XR is usefull in the situation that some problem occurs. Rather than drop connection, you just slow down. Customers are a lot more forgiving for a slower connection than a dropped connection.

My test link would have worked today if ROS supported XR. It would not have been fast enough for my needs, but I would have slept better being able to see the AP @ .250Kbps, rather than having a -90dB link smirking at me with the occasional link.

With the additional speed promised by 802.11n, I just can't get over the fact that signalling is that much more complicated, I can't see it having signifigant range over a well designed 802.11g network.

Who is online

Users browsing this forum: McSee and 43 guests