OpenBSD Journal

Improve large wireless networks with hostapd(8)

Contributed by phessler on from the two-legs-bad-wireless-good dept.

Lets say you have a bunch of wireless access points in Host AP mode. Lets further say they are all part of the same Extended Service Set (ESS, or in laymens terms "network"). Lets further say that the average client tends to associate with different APs during their session. If so, have we something nifty for you.

Reyk Floeter is back with a new daemon, hostapd(8). This new daemon uses IEEE 802.11f to connect wireless Access Points for tracking client associations and movements in large wireless networks, providing roaming improvements. hostapd(8) supports wireless interfaces using 802.11 kernel stack, like ral(4), ath(4), ural(4) and others. It currently does not support the wi(4) driver, since it misses some required functionality.

Future plans include implementing all of 802.11f, and some additional status messages between hostaps.

(Comments are closed)


  1. By Anonymous Coward (139.78.112.166) on

    If this does what I think it does....WOW!!! Very cool guys. I'm amazed at your accomplishments. One thing after another after another. Do you guys sleep? I saw when this was first checked in to the CVS tree and the name intrigued me, but I had no idea it would be something so cool. It seems each release includes some awesome new functionality and 3.8 won't fall short.

  2. By Jason Houx (216.201.34.104) on

    Guys! Amazing! More cool toys to play with. My company keeps buying small Soekris boxes to play with. You guys just keep me busy playing with all your great functions with them! Thank you!

  3. By Venture37 (80.3.64.11) venture37 hotmail on www.geeklan.co.uk

    its a man page from the future: History: The hostapd program first appeared on the 21th Chaos Communication Congress (http://www.ccc.de/congress/2004/) and later in OpenBSD 3.8.

    1. By phessler (64.173.147.27) on

      3.8 means its the first release that its avaliable in. it will not exist in 3.7.

      this is the standard way to mark it.

  4. By Anonymous Coward (213.118.35.44) on

    Lets? Surely you mean "let's".

  5. By Justin (64.81.50.82) on

    I see they have plans to implement all of 802.11f. I agree that the
    roaming broadcasts are a good thing, but can anyone please enlighten me
    as to what other practical features 802.11f has? As far as I know, zero.
    This is the primary reason why the standard is being scrapped for
    something more practical.

    1. By reyk (213.23.239.214) on

      hostapd is not limited to IEEE 802.11f. It's a generic name, isn't it? I plan to implement some useful and simple functionality to improve monitoring and management of OpenBSD-based wireless networks. It don't plan to implement any IEEE 802.11i/WPA stuff.

      In addition to a slightly improved roaming capability, IEEE 802.11f is very useful for realtime monitoring of station movements in the ESS. The protocol will be somewhat enhanced with additional and optional information about wireless events on the Host AP side.

      The are several reasons for the missing success of IEEE 802.11f. On the one hand, it does have some major design faults (these people didn't know about common operating system concepts) which has been implemented in a compatible but totally different way in hostapd. On the other hand, most of the vendors moved to sell centralized WLAN appliances using the IETF's CAPWAP approach with dumb APs and a "Remote-MAC" (It's always good to sell complex and expensive voodoo boxes...). Personally, I prefer decentralized and redundant network concepts ;).

Credits

Copyright © - Daniel Hartmeier. All rights reserved. Articles and comments are copyright their respective authors, submission implies license to publish on this web site. Contents of the archive prior to as well as images and HTML templates were copied from the fabulous original deadly.org with Jose's and Jim's kind permission. This journal runs as CGI with httpd(8) on OpenBSD, the source code is BSD licensed. undeadly \Un*dead"ly\, a. Not subject to death; immortal. [Obs.]