wireless router capable of being an access point client

No. The wireless access point section is always connected to the LAN side of the router.

There are some wireless devices that have two radios inside a single package. They're fairly specialized and used mostly for mesh networks. One radio would be used for the the backhaul to the ISP by bridging, while the other radio would act as an access point for clients. However, I would advise against trying to do this in a single box. The two radios will need to be on seperate channels and with everything in one box, there just isn't enough isolation to prevent mutual interference.

There are several ways to do what I guess you wanna do:

1. Conventional wireless router connected to wireless client bridge. The bridge need only bridge one MAC address so any "game adapter" will work. The other end of the link is a wireless access point or wireless router. The problem with this method is that you might end up with double NAT if you have two routers.
2. Conventional wireless access point conntected to wireless client bridge. This one needs to be a wireless client bridge that can bridge more than one MAC address (due to lack of an NAT router in the system). Something like a WRT54G with DD-WRT firmware in client mode will work. The other end of the link is a conventional wireless router with NAT.
3. Conventional wireless access point connected to a wireless transparent bridge. The other end of the link is an identical wireless transparent bridge. A pair of WAP54G radios in bridge mode will work. If there is a router, it should be at the far end of the link.
4. WDS bridge. A WDS bridge can simultaneously act as an access point and a transparent bridge in a single box. The other end of the link is an identical WDS bridge radio. A pair of WRT54G wireless routers with DD-WRT firmware should work. If there is a router, it will be at the far end of the link. This is nice because it only uses one radio. The down side is that the maximum speed of wireless to wireless traffic through a WDS bridge is reduced by 50%. This is not a problem if you're sharing a broadband connection, but can be a problem with higher speed LANs.

There are other combinations and mutations of the aformentioned that will also work, depending on your budget and creativity.

[snip]

I guess the wireless client bridges that can connect more than one MAC address are the ones that tend to be proprietary (so you would be better off connecting to an access point of the same brand)? Or are those the transparent bridges that need to be the same brand/model on either side?

Anyway, depending on the answer I am now down to looking for a wireless client bridge that can handle multiple MAC addresses or a single MAC address and have the following requirements:

- 802.11b/g, and hopefully also 802.11a

- external antenna connector

Links of Interest:

The WRT54G does have an RP-TNC external antenna connector: The WGA54AG does both 802.11g and 802.11a, but I think the antenna does not come off.

Any other options?

"Spacey Spade" hath wroth:

Good question. I don't have a complete answer. There are many names for a "wireless client ethernet bridge" adapter:

problem is that it's difficult to distinguish between a wireless client that will bridge exactly one MAC address, and one that will do more than one. The data sheets consistently refuse to mention this not-so-minor detail. It also has degenerated into a marketing tool, where it's more profitable to sell an overpriced "game adapter" for each game console, than a single "wireless client ethernet bridge" and an ethernet switch, to do the same thing for much lower cost (and much better game console to game console performance).

I've wanted to prepare a list of ethernet client adapters and the number of MAC addresses they will bridge. So far, I know that the following will bridge more than one MAC address: WET11, WET54G, WAP54G, WRT54G with DD-WRT in client mode, DWL-900AP+, and WAP11 v2.6. Also, any of the devices labelled as a "workgroup bridge". However, that's all I know for sure.

Multiple MAC addresses are not really proprietary. For example, in Ad-Hoc mode, a client can connect to 8 other devices and MAC addresses. This is clearly defined in IEEE802.11-1999. However, in infrastructure mode, which uses an access point, there was no need to connect to more than one MAC address. So, the commodity clients didn't bother to include the feature. Bummer.

If you add 802.11a as a requirement, the number of possible client radios drops dramatically.

I use DD-WRT on various routers with good results. Also look at OpenWRT as a suitable alternative. However, I have only played with the client mode long enough to be sure it works and no more. I was able to use DD-WRT to simultaneously bridge two laptops to a DI-614+ wireless router without difficulties.

The actually connector is usually either a RP-TNC (or RP-SMA on other products). I usually attach a pigtail adapter to a Type N connector and then to the antenna. The pigtail is to prevent breaking the connector on the radio because coax cables useable at 2.4GHz tend to be rather stiff and thick.

OpenWRT. |

| If you don't want to use alternative firmware, methinks any of the following will work (not 100% sure). WET54G: | | | | Etc. Check the spec sheets on the various vendors of "access point" devices as these usually also include a "client mode". The problem is whether the client mode will do more than one MAC address. The spec sheet usually doesn't say.

It might be handy to see what's in the web based configurations: DLink:

[snip]

If that wasn't complicated enough, I've added:

- decent internal processor, flash and, um... non-flash memory

- effective isotropic radiated power that is close to the legal 100mW

For example, I saw a list at a certain retailer, which showed "transmitted power" dbm number for only 1/3 of their 70+ models of access points. I'm not sure what transmitted power means, but none of the models went over 20dbm, which is the legal limit (=100mW). Some had only 12dbm, which is awful (=16mW). You could put a high gain antenna on the cheapie 12dbm model, and come out the same as a 20dbm model.

A little knowlege is a dangerous thing.

Spacey

"Spacey Spade" hath wroth:

Careful. You may get what you ask for.

Sorry. I only deal in indecent processors. However, they do have proper descendents.

Could you be a bit more specific as to what you consider "decent".

Please watch the terms. Isotropic means equal radiation in all directions. I don't think that's what you want. Methinks you might mean EIRP (effective incident radiated power) which is the xmit power times the antenna gain. In the US, the FCC limit for point to multipoint is 1 watt output (+30dBm) for the transmitter, and a maximum of 6dBi (4 times) antenna gain. There are very few radios that transmit at this high level. It's usually done with an expensive power amplifier. It's also not a great idea as the high power xmitter tends to cover a much larger area than it can receive, and therefore is no better than a wide area jammer.

Are you in the US or somewhere else where the xmitter limits are lower? RDNS shows you're in Alabama USofA.

You can get xmitter power specifications from the FCC type certification reports.

Yeah, something like that. Rule de thumb: 6dBi increase equals 2x the range. 12dBi increase equals 4x the range. 4 times the power equals 2x the range. 16 times the power equals 4x the range.

- Increasing the TX power does nothing for the range unless the other end of the link also increases the power by the same amount. - Antenna gain increases range in both directions. - TX power increases range only in one direction.

Too much knowledge is also dangerous. That's where I have to explain things to those with too little.

Spare me, ok. I've had to have a LOT of patience in teaching people to use things like Windows Explorer. It's really hard for some people that (1) Have never been into anything technical, and have always gotten by on memorization rather than understanding the concepts (2) Haven't exercised their minds in decades because they've had cushy surroundings. Sometimes, they just don't have the IQ, but the intelligent aformentioned really drive me crazy. But then I think... well, that's why I'm special! And then everythings ok. But I continue to froth at the mouth for the fun of it.

Correct them over here:

Actually, I would want the isotropic version, so that I can compare access points regardless of their antenna. I plan to use my own antenna.

I'm just here until my daughter lets me go off wherever I want. This is not my location of choice. I think I may have misread the limit somewhere.

[snip]

So if I put a very high gain antenna on a weakling, I should be matched up with most any consumer access point out there in terms of the effective incident radiative power. ;-)

On 6 Aug 2006 23:00:18 -0700, "Spacey Spade" wrote in :

Do it yourself -- that's the whole point of a wiki.

I'm not qualified. The last few days is the extent of my venture into this body of knowledge.

On 6 Aug 2006 19:37:02 -0700, "Spacey Spade" wrote: : [snip] : > Anyway, depending on the answer I am now down to looking for a wireless : > client bridge that can handle multiple MAC addresses or a single MAC : > address and have the following requirements: : >

: > - 802.11b/g, and hopefully also 802.11a : > - external antenna connector : : If that wasn't complicated enough, I've added: : : - decent internal processor, flash and, um... non-flash memory : - effective isotropic radiated power that is close to the legal 100mW : : For example, I saw a list at a certain retailer, which showed : "transmitted power" dbm number for only 1/3 of their 70+ models of : access points. I'm not sure what transmitted power means, but none of : the models went over 20dbm, which is the legal limit (=100mW). Some : had only 12dbm, which is awful (=16mW). You could put a high gain : antenna on the cheapie 12dbm model, and come out the same as a 20dbm : model. : : A little knowlege is a dangerous thing. "Drink deep or taste not the Pierian spring."

More to it...