Take the 2-minute tour ×
Server Fault is a question and answer site for professional system and network administrators. It's 100% free, no registration required.

After browsing serverfault and superuser for a while, I have noticed many posts about wireless router connection limits such as this one and this one. The sane limit for a router (without dropping connections excessively) seems to be around 40 (at most). Is there an explanation for why this limit exists?

share|improve this question

closed as not constructive by Sven, mdpc, John Gardeniers, Khaled, Ward Jan 1 '13 at 9:35

As it currently stands, this question is not a good fit for our Q&A format. We expect answers to be supported by facts, references, or expertise, but this question will likely solicit debate, arguments, polling, or extended discussion. If you feel that this question can be improved and possibly reopened, visit the help center for guidance.If this question can be reworded to fit the rules in the help center, please edit the question.

1  
Can the downvoter please explain why this was downvoted? –  Snakes and Coffee Dec 31 '12 at 23:50
    
I didn't down vote you but I'm going to assume because you're question isn't appropriate for serverfault. It should've been posted on superuser.com. –  slm Jan 1 '13 at 0:14
    
I considered doing so, but felt that trying to build a 200+ client wireless system falls under enterprise. –  Snakes and Coffee Jan 1 '13 at 0:40

2 Answers 2

With the most capable (and yet affordable) wireless access points I know of, you'll still only get 30-50 clients per AP. Remember a client is a single device, but on average, a user has 2.5 devices.

The general reasoning behind this is that most APs have a single radio, and there's a reasonable limit to the amount of traffic a single radio can handle. Much like there's a limit to the amount of traffic a road can handle, before it becomes congested.

There are some wireless APs (or rather, arrays) which can handle 200-300 clients, but unless they're all sat in the same place, you'd be better off with a network of APs, each with 20-50 clients each, and with a controller to handle per-access point hand-off, and roaming.

share|improve this answer

The comparison to 14.4k modems isn't apt. In the case of an access server with a bunch of modems attached, each client has a dedicated connection and its own share of bandwidth. In the case of wireless all clients are sharing the same radio frequency space. As the client density on a given AP increases, the likelihood of contention for this available resource also increases. The 40-some-odd client is a guideline and can vary based on client activity (among other things).

This issue is independent of what actually tended to be an issue with access servers, which is oversubscription of uplink bandwidth. Several hundred 14.4 users being directed out of a single T1 meant an often untenable degree of resource contention but that contention wasn't between the client and the access server but rather between the access server and the rest of the world (vs the wireless case where contention occurs between client and access point).

share|improve this answer
    
The question is specifically what limits the connection capacity. I am aware that "resource contention" causes this, but what resource? –  Snakes and Coffee Dec 31 '12 at 23:39
    
As I said: "In the case of wireless all clients are sharing the same radio frequency space." Within a given frequency on a single AP a single client can be transmitting at any given time. Each station listens for an opening to transmit. If two stations transmit simultaneously, both will back off for a random period before retransmitting. The more stations, the higher the odds the transmit will be delayed or a retransmit will be attempted. If too much time passes or too many retransmits occur, the packet may be dropped. –  rnxrx Dec 31 '12 at 23:49
    
Remember that there is overhead incurred in maintaining the client/server connection so at some point this overhead consumes more and more of the total bandwidth available, eventually consuming all the bandwidth so that there is none left for actual data. –  slm Jan 1 '13 at 0:49

Not the answer you're looking for? Browse other questions tagged or ask your own question.