Why recommend an a/b/g AP instead of just an 802.11g AP?

[zjohnr]

I wish I was posting a better formed question than I am, but at the moment I don't know enough to do so.

Some context. I'm trying to help out with the approximately 12 to 18 PC LAN at the local (non-profit) SPCA. Approximately a year ago they got a proposal for adding wireless support to their currently ethernet 10/100 only Windows client/server LAN.

The question I'm trying to answer at the moment is "What were they thinking?"

The proposal only refers to adding 3 Cisco a/b/g Access Points plus 3 corresponding Cisco Power Injectors for the Access Points. I think this is actually a reference to either a Cisco AIR-LAP1131AG Lightweight Access Point or an AIR-AP1131AG Access Point and to the AIR-PWRINJ3 Power over Ethernet (PoE) doo-hickey.

As I try to dig into this a little further, the first question that pops into my head is why would they recommend the slightly more expensive a/b/g AP?? Why not save a few bucks and go with the 802.11g AIR-AP1121G access point?

The second question I had was why on earth would they recommend using Power over Ethernet?

At the moment I'm not coming up with any good reasons to go with the equipment they (appear to have) proposed using. There are no legacy wireless network support concerns. The only "existing equipment" they have is 3 Linksys wireless PCI cards they have already purchased to allow them to add 11g support to a desktop PC.

But maybe I'm missing something?

Figured it couldn't hurt to post here and see if anyone else wanted to speculate. I realize this is off the usual topic, but I think it's still covered under the umbrella of "Small Net Builder". It's still a small net, just not a home network.

-irrational john

 

[thiggins]

With a "greenfield" installation (brand new) and 11b/g only clients, I see no advantage to an abg AP, except for the rep making a bit more $ and possible "future proofing" in case they wanted/needed to move to 5 GHz.

The power injectors for PoE is actually a good thing, since it eliminates the need to locate the APs near AC power. With PoE, both power and data come over the single Ethernet cable.

Now to more basic questions:
- How large an area needs to be covered and what is the physical environment?
Office cubes, open warehouse, indoor/outdoor, what?

- How many wireless clients will be in simultaneous use? Sounds like only 3 desktops. Why wireless for desktops vs. notebooks? Wouldn't running Ethernet be cheaper?

 

[zjohnr]

With a "greenfield" installation (brand new) and 11b/g only clients, I see no advantage to an abg AP, except for the rep making a bit more $ and possible "future proofing" in case they wanted/needed to move to 5 GHz.

After posting I looked up the two APs on newegg.com and came up with another possible rationalization for the a/b/g AP.
CISCO AIR-AP1121G-A-K9 Wireless Access Point at newegg.com
CISCO AIR-AP1131AG-A-K9 Wireless Access Point at newegg.com

Perhaps the AP1131AG is supposed to cover a wider range than the AP1121G?

To be honest, I tried telling the difference from the specs, but I don't have enough familiarity yet with even this basic aspect of wireless to glean the difference between the two. Transmit power and receiver sensitivity don't really resonate with me. And I'm not sure how much trust I should place in the indoor/outdoor operating range numbers I saw.

But I think if there was a difference in coverage range, it probably wasn't a big one. So I'm still wondering if there is some other benefit to using an AP1131AG that I'm just not seeing.

Eventually I'll probably get around to trying to ask the folks who originally suggested this equipment what their thinking was. It's just that I'd prefer to take a stab at coming up with an answer before I hear theirs, if you know what I mean.

The power injectors for PoE is actually a good thing, since it eliminates the need to locate the APs near AC power. With PoE, both power and data come over the single Ethernet cable.

If you say so, Tim. It seems like potentially pointless puffery given the very basic goals that these folks currently have. When I finally figured out what a "power injector" is the most reasonable purpose for it I could think of was it might allow them to position the AP above the ceiling tiles of the room in one of the two buildings where they want to enable wireless connectivity.

But that room is only two plasterboard walls and a hallway from the inside of the closet which holds the ethernet switch for that building. So if they put an AP in that closet and plugged its AC adapter into a power strip the computer which might use it would be maybe 15 to 25 feet from it. So I'm not immediately convinced that the extra positioning flexibility PoE might provide them would really be a significant benefit to them.

But maybe PoE will make more sense after I spend more time looking at this. And maybe the people who put together the original "proposal" that I'm looking at just threw the PoE injectors in as a "What the heck? Maybe they'll be useful later?" type of thing. (More on that below).

Now to more basic questions:
- How large an area needs to be covered and what is the physical environment?
Office cubes, open warehouse, indoor/outdoor, what?
- How many wireless clients will be in simultaneous use? Sounds like only 3 desktops. Why wireless for desktops vs. notebooks? Wouldn't running Ethernet be cheaper?

They pretty much only want to cover one room in each of two adjacent but separate buildings. The AP in one building would be providing connectivity for up to 3 PCs in a single room. The AP in the other building would primarily need to service one PC in a small animal surgery room. The usage would all be low bandwidth stuff. Internet surfing, running some server based programs, and (small) file editing.

I think they might also like the coverage to extend to a large conference room which is immediately above the surgery. That's more of "Gee, wouldn't that be neat?" potential extra benefit at this point. There is a concrete floor separating the conf room from the surgery below it. Not sure how much that might attenuate any signal. But currently I'm thinking one AP could cover both areas. (In other words, I think they only need 2, not 3 APs.)

The building with the office is most likely a donated residence so the only obstructions would be wood or plasterboard. The other building was built in 2004 and I think it has a steel frame. But the rooms are separated by plasterboard as far as I know.

There currently is no wireless support of any kind. All the computers currently on the LAN are connected via ethernet. They clearly have a lot of Cat 5 (maybe Cat 5e ... I haven't looked) already in place. They want to use wireless so they can easily move a few computers to an area that the existing ethernet cables apparently don't reach.

As for the cost of wireless versus extending the ethernet cabling, here's my best guess. I think it's part of the sometimes upside down way in which resources flow into a non-profit. The assumption is that they can get the Cisco APs as "donations". In that case they would only need to pay a relatively very low administrative processing fee to obtain the APs (and even less for the PoE injectors ... which is why I suspect they might have been included on a "Why not?" basis.)

If they had to pay the full retail price for this stuff I doubt they would go this route. They just don't have the money for it. Likewise, I assume it would also cost them more than they could afford to pay someone to run new ethernet cable to the areas they want to put the computers. No "cable guy" has stepped up and "volunteered" to run ethernet cable for free.

Maybe running the cable is something their building maintenance person could take care of in house. I don't know at this point. I would love to get my hands on some sort of a building (wiring) diagram which showed the layout of the existing network cables. Unfortunately, I don't think one exists. Or at least if a wiring diagram does exist, no one I've talked to so far seems to know how to lay their hands on it.

As for laptops versus notebooks ... again it's cost. The organization currently has two laptops and both are already in use for another purpose. There's no money to buy another laptop.

Apologies for rambling on like this. I never intended to make this post so long when I started it.

-irrational john

 

[thiggins]

If the products are being donated, they are fine and will do what's needed. PoE isn't necessary. It just might come in handy.

If they have to purchase equipment, they could use any inexpensive 802.11g wireless routers, convert them to APs, connect them into the Ethernet LAN and be up and running just fine.

 

[zjohnr]

If the products are being donated, they are fine and will do what's needed. PoE isn't necessary. It just might come in handy.

If they got them it would (technically) be a donation. But they'd have to pay an "administrative fee" to an organization that acts a middle man in the transfer of the equipment from the donating company to the receiving organization so it's not completely without cost.

I guess my question is if there is any reason to pay slightly more for the a/b/g AP1131AG? Or, knowing in advance that only 802.11g will be needed, will the more single minded AP1121G do just as good a job while leaving more money to purchase kibble?

I'll keep your point about the PoE injector in mind. I agree with you that it just might come in handy so it's probably a good idea to include them.

If they have to purchase equipment, they could use any inexpensive 802.11g wireless routers, convert them to APs, connect them into the Ethernet LAN and be up and running just fine.

Actually, I picked up a Trendnet TEW-637AP, which is apparently a rebadged Ralink model RT2561 AP, when newegg was selling them for $25 shipped. I don't know how it would compare to the Cisco AP when doing just 802.11g, but I think it would be adequate for their needs. I was thinking of donating it to them. (A possibly better "deal" since I don't charge an administrative fee. )

I assume the Cisco APs would actually be overkill. Since the Cisco APs currently retail for at least 8 times more than a TEW-637AP, I'd expect a Cisco AP to be able to handle more PCs simultaneously without slowing down.

Is this naive of me?

-irrational john