Product Request: High band point to point 802.11a radio

[NewBWiFiGuy]

In all my researching I've found no solution that does 802.11a upperband Bridging. Since upperband 802.11a operates so far away from low & middleband vendors have not been dealing with it just yet. It's also more meant for just briding so the market demand is not as high. Read on...

What I want:

An 802.11a radio that operates in the upperband 5.725Ghz-5.825ghz range with the maximum legal output power (30dB/1watt) that does point to multipoint long range bridging.

Multiple diverse simultaneous bridged channels option: The AP could have ability for the radio to use up to four full diverse channels in the 802.11a upperband at once to achieve 35/70/105/140mbit of actual thoughput - I note 35mbit per channel since that seems to be a more true actual thoughput level for OFDM than the 54mbit that is advertised. A solution like this would allow the WISP operator to create a very high bandwidth backbone link using just one unit.

*Price of 4 channel radio under $400 retail. The AP must be in this price range. It simply isn't affordable for the small WISP/Hobbyist to spend more per bridge. Remember, after a $100 antenna on each end, two $400 APs, $100 worth of cabling - you are looking at $1100 for a bridge! In the telco and enterprise class ISP market $1100 is a drop in the bucket, but to the rest of us that have to build 3 or 4 bridges just to start a small rural WISP it adds up really quick.

I read that at least one 802.11a OEM chipset maker is building a 2 channel lower band operational mode into their product. Basically this will bridge two of the 802.11a channels into one and give you 108mbit (60-70mbit actual?). So now all we need is a 4 channel device that could fully use the 802.11a upperband for bridging. I also saw that the IEEE may be building an extension into their standards for multichannel bridge modes.

Comments? Ideas? Remember folks, the more the 2.4Ghz spectrum is eaten up, the less your going to have for these big bridges. Migration to low cost 5.xGhz bridges needs to happen IMO. Please also see my post (http://forums.80211-planet.com/showthread.php?s=&threadid=1049) about cheaper 5Ghz point to point antennas to encourage the adoption of cheaper solutions for 802.11a bridges.

Thanks

[NewBWiFiGuy]

An Update:

Since my original post I found a few more vendors coming out with 5.8ghz longer range point to point bridge radios, but I've seen no prices so I worry they'll be charging alot fo them.

It would also sure be handy if the FCC up'ed the transmitter power limit for the point to point upperband. Right now your radio/amp can output 1watt/30dB and you can use any gain antenna you wish. This is much more flexible than 2.4ghz rules. Using a 4ft dish you could clear 66dBi legally.

But.. To acheive the long links it would sure be handy if the FCC raised the TX limit to something like 36dB/4watts. That would allow you to hit 64dBi/2511.9watts EIRP with something like a 28dBi 2ft dish.

They could require that these radios only run high power ouput in bridge mode. Law would require that anything with these power levels is used only in a finely aimed point to point system. Radios could auto trim their output power based on auto gain control communication with the other radios on the bridges.

As mentioned in this tutorial (http://www.80211-planet.com/tutorials/article.php/1428941), the FCC..." allows operation up to 4 watts EIRP. This is 1 watt (the earlier limitation) plus 6 dB of gain"... for 802.11b point to point links. So we know the FCC understands that point to point links need a little more flexiblity power wise and since the upperband 802.11a channels were apparently created for bridging solutions, why not allow a little more flexibility for WISPs? If someone is being a bad boy aiming their 50dB of EIRP into someone's livingroom, the FCC can always shut them down or bring them down to earth.

If anyone knows who at the FCC should see this post please forward them the URL to this thread or give me their info.

P.S. I just found in an old issue of a newsletter the rumor: "Several industry mavens suggest that the IEEE may allow the building of 108 Mbps extensions into the 802.11a standard. Atheros Communications and Proxim already offer the 108 Mbps speed by combining two of 802.11a's 54 Mbps channels. Would these improvisations be codified in a formal standard? " - As I mentioned before, I'd like to see a 4 channel upperband radio designed just to be a bridge that squeezes out the most bandwidth it can for a single antenna point to point bridge. Imagine a clean 150mbit bridge to feed your backbone!

[NewBWiFiGuy]

On March 4th I saw this announcement on WI-Lan's new point to mulipoint upperband 802.11a radio: http://www.shorecliffcommunications.com/magazine/news.asp?news=1277

Here is the site: http://www.wi-lan.com/solutions/main15.html
I haven't seen pricing yet and their solution says it reaches only 24 miles. Who knows.


There is another option for those wishing to use 2.4ghz radios to build a 5.8ghz backbone link. I notice on www.ydi.com they have this product http://www.ydi.com/products/ca2458.php that will convert your 2.4ghz link to the 5.8ghz range. It also boosts the power to 250mw. Now, as seen in this document http://www.ydi.com/deployinfo/ad-wireless-backbone.php they are showing 100 mile 5.8ghz links. How wonderful this capability would be. But as usual anything cool comes at a price. Their converter amp is $800 per end if I am reading their site correctly. It's really too bad they can't price something like this cheaper. It would be nice to have such a system as an option.

By the way YDI, I appreciate the "deployment info" section of your site. It's handy.

[Uncle Ken]

Slight correction: Wi-Lan's product is 5.8GHz, but it's not 802.11a. There are no high band 802.11a products at this time.

[NewBWiFiGuy]

Ah thanks for the info. I should have looked deeper into their system. I thought it was a 802.11a radio.

By the way everyone, check out the new Guiness World Record for 2.4Ghz link. Guiness world record for a 2.4ghz link (http://www.shorecliffcommunications.com/magazine/news.asp?news=1280) . 192 miles with 6 watts to a baloon. Reminds me of the these
http://www.wired.com/news/wireless/0,1382,56961,00.html

[MJShimek]

Dream on brother! You will never do that in a single hop. Seldom is a commercial microwave link over 30 miles.

[NewBWiFiGuy]

Well I did say I'd do the 80 mile links if I *could*.. :) Right now the best I've seen is 40-50 with 802.11b and some boost. And I'm not sure about range with 802.11a because there are no upperband radios out yet.

What I am asking for is the FCC to promote the 802.11a upperband channels for bridging use and upgrade the power limitation to a level that with a clean dish point to point implementation it could reach 100 miles with clear line of site.

Most of my backbone links are 30 miles. If I had the option, I've got some mountain tops about 80 miles away with towers I'd link to.

Thanks

[NewBWiFiGuy]

New product from Dlink:
DWL-7000AP 802.11a/802.11b/802.11g trimode radio. It feature bridging which seems to be a first for a 802.11a radio. All radio's I've seen still don't operate in the upperband.

In the product specifications (http://www.dlink.com/products/wireless/dwl7000ap/SpecsTable.htm) it lists:
Wireless Frequency Range:
2.4 to 2.4825GHz
5.150GHz to 5.850GHz

Did Dlink just make a mistake here? Or is this really our first upperband bridging radio. If so it could be just what I need for my central wireless hub (mountain top). They advertise "..... With 11 non-overlapping channels" 802.11b/g has only 3 non-overlapping channels, and 802.11a has 12. bands (http://aaronsmith.net/802data/index.html#bands) Count just the 3 + 802.11a lowerband + 802.11a middle band and you get 12. So.....

[Uncle Ken]

I don't want to rain on your parade, but I think you'll find that this access point, like other dual frequency 802.11 products, only bridges at 2.4 GHz. I've found D-Link tech support to be pretty responsive -- why don't you shoot them an email and let us know what they say?

[NewBWiFiGuy]

No worries about raining on my parade, I have faith we'll see a upperband radio this year. I was just hoping this one would do it. You're probably right it only bridges in 2.4ghz band.

I still wonder if it actually does run in 5.8 upperband (http://aaronsmith.net/802data/index.html#bands) as they list.

[Sumaryo]

NewBWiFiGuy: Since my original post I found a few more vendors coming out with 5.8ghz longer range point to point bridge radios, but I've seen no prices so I worry they'll be charging alot fo them.

Contrary to what it said, 5.8 GHz bridging products had been offered by many companys in the past, including western multiplex (now under proxim brand), solectek and so on. It was not an 802.11 devices though. However, they were priced such that out of most people reach. The release of ne AR5111-AR5211-AR2111 combo 802.11a/b chipset by Atheros make it possible to make a low cost 5.8 GHz devices.
Here some of the spec:
Specifications:
Frequencies:
802.11b: 2.412 - 2.472 (5 MHz step, channels 1-13), 2484 (channel 14)
2.512 - 2.732 (20 MHz step, channels 15-26)
802.11a: 5.150 - 5.350 (5 MHz step)
5.470 - 5.825 (5 MHz step)
Main Chipset Atheros AR5111, AR5211, AR2111

[NewBWiFiGuy]

Sumaryo: Thanks so much for taking the time to reply! You made my evening.

Your right, I saw the 5.8Ghz radios for quite a while, but they were often a bit slower than I wanted or the prices were far out of reach for the bridging model I'm aiming for. Most were all proprietary too. I like standards, they bring me comfort ;-)

This chipset looks great.. I hope your right and some upperband lowcost bridging radios show up soon. I read on altheros.com about the 90Mbps thoughput and compression they've come up with. Boy would that make a nice backbone/backhaul connection.

[ccrum]

'm curious why you couldn't just use an upper band AP with a dish and using some MAC filtering to only allow the other AP? What would be the problem with this other than the easily breakable security? If the antennas are good enough, most people wouldn't be able to detect the dignal at the ground level anyway. Just a thought.

[ccrum]

By the way, I talked with the Smart Bridges people last week and they plan to make a dual mode AP later this year that will have an 802.11g AP and a 5 gig (although not .11a but still high bandwidth) bridge built into it. They say it will be out by 4Q, but who knows?

[ccrum]

Any way I can get a show of hands on how many people would be interested in having a high band 5.8 gig "A" bridge? I have a need for several hundred myself, but my vendor won't consider producing them unless I can order on the order of 1000. These will be affordable...less than $500. Any takers?

[NewBWiFiGuy]

Hello Ccrum! :)

Well, I cannot handle a big order, but I know over the next year I would use at least 20. The radio has to meet some requirements however:

* It must offer multipoint bridging that brings complete subnetworks together. In other words, my backbone design is one big network. It will behave just like everyone is on a single network switch. I have each radio tied into a SOHO Dlink SWITCH. it has some generally advanced switching so stray packets don't go where they aren't needed.

* I have 1 mountain top that can be seen for 80 miles around. With 802.11a upperband, high gain antennas, and maximum legal power output I expect I can get 50-60 miles if I am lucky. Bascially, the mountain top has either one multichannel radio or 4 single channel radios that must support connects from the satellite radios located around the valley. My current problem is that Linksys radios support only 4 connected remote APs, and the Dlink supports more but they told me they recommend only 5 or 6. Since this mountain top location is my main "super hub" it needs to be able to connect to many many remote radios. I could handle just 15 or so, but 30+ would enable me to really make my business work!

* Power output: For $500 I would expect the radio to at least have 24dBm/250mw of power output, but 30dBm/1watt would be much more desirable! With 5.8Ghz amps still very expensive, I'd rather have the radio capable of the power itself. Or if it features low power output I'd need the radio to be cheaper and possibly come with a lower cost amplifier option to mount up with the Antenna. Point is, I can't make the distance without going to the maximum legal power level, and I can't install lots of units if they are insanely expensive.

Optional:
* Would be handy if the radio would use more than 1 upperband channel. Would be wonderful if it would use 4 bridged channels to acheive a very high amount of bandwidth.

* Usefull if the unit was a mast mountable so you could put it right after the antenna and simply run ethernet cable down to your switch. Weatherproof casing shouldn't add more than $50 to a device.

* I'd prefer it the units were 802.11a upperband standard.


You might want to fire off an email to electro-comm.com and see if they are interested. They've got some pretty helpful people there. Maybe if the unit was commercially viable they, or other distributors, would carry it. I'm still surprised it's taking so long for manufactures to come out with a good upperband bridging solution. 2.4Ghz just isn't viable anymore for large backbones. 5.8Ghz is cleaner and allows for much higher EIRP in point to point applications.

[ccrum]

What linksys and d-link are you using? I can find some 5.8 amps but the mfgs all tell me 1W at 5.8 gig is very difficult. Most of them who are going to be making these are only going to be around 250 mW.

[NewBWiFiGuy]

linksys WAP54g and Dlink DWL-2000AP.

Bummer, I didn't realize it was any more or less difficult to make a 5.8ghz amp. Well at 250mW one could still get some distance in point to multipoint. You're limited to about a 12dBi directional antenna. Then at your remote point you could use a much higher gain dish to make the reach. 250mw on a 24dBi gain antenna would net 53dBm EIRP.

[Sumaryo]

Guy,

Just a little consideration.

All 802.11 devices, wheter it's a plain 802.11 (2.4 GHz DSSS 2 Mbps / 2.4 GHz FHSS), 802.11b or 802.11a /g are so called "indoor" protocol. The reason is with the way they are achieving multiple access, i.e. CSMA/CA (carrier sense multiple access / carrier avoidance). This is basically a listen before transmit scheme, that assumes all radio involved able to hear one each other. However, if we use this protocol outside (outdoor), the phenomenon called "hidden node" - where one or more radio can not hear what the other radio was doing is really killing the system. The use of RTS/CTS mechanism might reduce the adverse effect of "hidden node", but will not eliminate it at all.

The question is, why are you so eager to use such system for mission critical application (long range backbone/carrier)? If the reason is "cost", better you shop arround for more proper radio/techniques/protocol.

[oshea85]

The Cisco 1400 Series has been announced!

http://www.cisco.com/en/US/products/hw/wireless/ps5279/index.html

[ccrum]

Any Idea on the price?

[ghency]

5.8 GHz, Bridge to 100 spots, cost-effective, and it is polling so you don't have to worry about "hidden node".

Stocking now at fine web outlets.

[ccrum]

Still don't see where this is a bridgeable unit. It is simply an AP.

[mvario]

Huh? The Cisco 1400 is a bridge, not an AP

Other 802.11a bridges:
http://www.airaya.com/
http://www.radiolan.com/

not quite 802.11a:
http://www.proxim.com/products/all/tsunami_mp11/mp11a.html

[ccrum]

I was refering to the proxim unit.

[NewBWiFiGuy]

ghency,

While the Proxim units look very nice, the price is far from cost effective. I still need a good low cost 802.11a upperband bridging radio for my large hub sites.

[ghency]

As a fellow WISP, if you want to establish a dialog to share information, please email me at ghbond@yahoo.com. I have spent considerable time in R&D over the last two months.

The Proxim MP.11 is a wireless bridge designed for WISP use, or other distribution of bandwidth. It comes in a "b" version, which I am currently testing at my shop in Michigan. I have 1 base and 6 residential subscriber units. The equipment has been flawless so far.

It also comes in a new "a" version. Both types have Worp, which is Proxim's version of Karlnet. If that means nothing to you, go to the Karlnet website and learn about 'hidden node" and packet inefficiency.

There are other interesting products I found. Solectek is higher in price, but will have an "a" product out soon. They also make a 200mW AP, as does Zcomax.

smartBridges (to the future) makes an outdoor-ready line of products, but it is confined to "b" right now.

For picocells, the only Wi-Fi equipment that meets my requirments is D-Link. They have a flexible AP which can bridge, and they meet 55 degrees C temperature.

I am starting to look at Buffalo, and have some thoughts on future development.

As far as "a" equipment, it will stay pricey until the market volume increases. Almost all the manufacturers are just releasing a, or have plans to do so soon. It's people like us who will drive the market for "a", until next year when all equipment should move towards tri-band.

gh

[ccrum]

OH...I see now. You were talking about the Tsunami MP.11a not the Orinoco .11a. There is a big difference. However, this unit is far fom cost effective. I don't know about where you are providing service, but my subscibers won't pay more than $150 for the CPE. I'm certainly not going to subsidize it either. There is nothing about these radios that makes them this costly. It's rediculous to price them this high.

[NewBWiFiGuy]

Ccrum,

I agree, there is nothing in these radios to boost the price to their crazy levels. I'm surprised they don't go for the lower price sell more units model.

[ghency]

CCRUM and NBWFG, what are you using for CPE equipment right now?

The MP.11 is more than a radio, it's a computer hosting a radio. I don't plan on having MP.11 as a CPE except for high-level subscribers. I want a level of distribution under the MP.11. This is where the D-Link equipment comes into play.

The setup I have right now is a linksys router (WAN) tied to a D-Link AP in bridge mode(back-haul link), bridging up to my MP.11 base station (intermediate distribution point), and switched to a SMC wireless router (Base-located AP).

From there, I have a MP.11 subscriber unit going to a router going to a desktop (heavy-use business subscriber setup). This is the unit I am typing this reply on.

I have another MP.11 subscriber unit tied to an D-Link access point in AP mode (First hop in residential, or multi-location business).

I'd love to have equipment under $150, but have yet to find a useable solution. I saw an AP ad for $9 today. Maybe if we start to cooperate on our designs and our efforts to drive manufacturers, we can get the products we need at the prices we want. If not, we may have to build our own. I am considering both options right now.

[ccrum]

I have a direct tie into several of the manufactures in Tiawan. I already have one buildng 5.8 bridges for me for just over $200. Our configuration seems much simpler than yours, but our application is slightly different. We service apartment complexes and install enough AP's outdoors to serve everyone indoors. There is no CPE equipment except their wifi adapter in thier computer. Right now we set up a central hub (AP in bridge mode). Then our AP "sites" are an AP serving the clients tied to a bridge which goes back to the main hub. Currently we are doing all in 2.4 becuse we can't find cheap enough 5.8 radios. This mfg in Tiawan will make these exclusively for us...we may market them ourselves later. Since we do all our authentication through a gateway/radius combination, we don't need super sophisticated AP's or bridges. We are also all RF engineers who have been designing wireless networks for many years. We know how to minimize interference using the correct antennas (which we design ourselves, and market by the way) and proper frequency planning and coordination. We "isolate" each AP from the other spatially rather than trying to do it by bruit force. We will begin serving private residences soon as well, but our configuration will be even simpler. One tower with 6 AP's (3 in 2.4 and 3 in the 5 gig range). We will have two levels of service, residential and high end business. The business customers will buy a 5.8 bridge which will link back to our 5.8 AP running in bridge mode at the tower. Residential will be serviced with 2.4...one channel for each sector. If the area requires more than one tower, or we hit our capacity limits on the 3 AP's then we will build/lease another tower in an appropriate location and run the same config from there....sort of like cell splitting. Residential customers will pay us $150 for a 2.4 bridge and installation...basically covers the cost of the equipment...we'll eat the labor. TOwers will also be linked via 5.8 bridges so we have all the WAN backhaul in one place...also saves us from having to buy mulitple gateways. Our RADIUS server billing, email, etc. all sit at a NOC with triple redundant high cap links to the internet.

[ghency]

CCRUM

Your configuration is basically the same as mine, except I am using an MP.11 to get out to the AP's. I found no cheap access point that could support a polling mode and packet compression.

I would like to look at your antennae. What is the site?

I am a EE, my partner is a PhD in comm, and we also are stressing efficient use of channel and RF deployment.

Your apartment model is interesting. We are doing mostly 3 story or less and coming down from the attic.

You did not say what area you are operating in, but in MI we have a lot of trees. This limits the effectiveness of central distribution. We are currently looking at how to create threads out into neighborhoods which weave in between the trees.

It is good to see other engineers out there on the right track. I recently stumbled accross www.broadbandcentral.us, which also has put a focus on antenna design.

It is interesting that you believe residential subscribers will pay $150 for installation. My partner believes that any installation cost for residential customers will create a serious impediment to doing business.

I would like to schedule a conference call with you to share business and technical ideas. Please send me an email if you are agreeable.

gh

[ccrum]

You can check us out at Dot11 Networks (http://www.dot11net.com) and the antennas are available trough us at our antenna page (http://www.dot11net.com/antennas.asp) . Email me direct at ccrum@dot11net.com. The antennas are slotted waveguide antennas with extrememly high gain. They list for much lower than "the best" you can get right now from YDI and others. There isn't an antenna on the market which matches our performace and gain and especially for the price. The list price is on the site...mulitple purchases obviously brings it down. If you want to try one, send me a line on e-mail and we'll work it out.