Date: Wed, 19 Jun 2002 07:00:53 -0700 From: "Kevin Oberman" <oberman@es.net> To: Christoph Kukulies <kuku@accms33.physik.RWTH-Aachen.DE> Cc: Marco Molteni <molter@tin.it>, freebsd-mobile@FreeBSD.ORG Subject: Re: wireless lans with multiple accesspoints Message-ID: <20020619140053.B1FA35D04@ptavv.es.net> In-Reply-To: Your message of "Wed, 19 Jun 2002 08:41:19 %2B0200." <20020619084119.C27055@gilberto.physik.rwth-aachen.de>
next in thread | previous in thread | raw e-mail | index | archive | help
> Date: Wed, 19 Jun 2002 08:41:19 +0200 > From: Christoph Kukulies <kuku@accms33.physik.RWTH-Aachen.DE> > Sender: owner-freebsd-mobile@FreeBSD.ORG > > On Wed, Jun 19, 2002 at 02:42:09AM +0200, Marco Molteni wrote: > > On Tue, 18 Jun 2002 00:36:39 +0100, Brian Somers <brian@Awfulhak.org> wrote: > > > > > The access points will negotiate with eachother and choose the one with > > > the strongest signal. > > > > I think this is incorrect. The APs don't negotiate anything among them. > > > > See below > > > > > On Mon, 17 Jun 2002 11:19:53 +0200, Christoph Kukulies <kuku@accms33.physik.rwth-aachen.de> wrote: > > > > Assume you have a LAN with several access-points attached. > > > > The reachability areas of these access-points are overlapping. > > > > > > > > There is a DHCP server in the network that supplies IP adresses for the > > > > access-points and the clients, e.g. notebooks with wireless pc cards. > > > > > > > > What happens when you are in the area that is covered by two access-points? > > > > > > > > I mean, which access-point takes over the 'routing'? > > > > I am not sure of what you mean by 'routing', since an AP is a layer 2 > > I chose the wrong term. Routing not in the sense of what we understand as > a router. I just meant 'passing through' the packets. It's a bridge, > of course (kind of). > > In this vein the term 'collision domain' came up. In how far does > a 100 Mb network consisting of several (3COM 3000) switches which are > cascaded (using TP cables, not a matrix cable) still form a collision > domain? Does it really? The use of that term goes way back and it really is no longer appropriate in the era of full-duplex where collision simply don't exist. But it matters here. There are two issues: 1. Can they build a spanning tree? As long as all connections are switched (layer 2) and not routed (layer 3), this should work. 2. Are they in a common collision domain? When one AP detects a collision, do the other APs se it? If so, they are in a collision domain. In more common terminology, are they connected by a hub or a switch? The issue is significant if you want to support mobility. Say you want to enable wireless in a large facility and allow people to walk around with their laptops without losing connectivity. The APs must be in a single collision domain. Otherwise the switch forwarding cache will break things when laptops switch from one AP to another. Officially, Cisco does not support roaming between cells, but it works. I'm not sure about others. R. Kevin Oberman, Network Engineer Energy Sciences Network (ESnet) Ernest O. Lawrence Berkeley National Laboratory (Berkeley Lab) E-mail: oberman@es.net Phone: +1 510 486-8634 To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-mobile" in the body of the message
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?20020619140053.B1FA35D04>