EXPERT RESPONSE
This is an excellent research topic (IMHO). WLAN performance analysis has been challenged by the nature of its Layer 1 and Layer 2 behaviors; there is far greater variability in the network path (mobility, routing, negotiated capacity, inherent loss) than in fixed wire. Then getting beyond Layer 2 to application performance offers both a rich area for research but also a valuable contribution waiting to made to the market.
I am unfamiliar with any serious efforts to address all the issues of WLAN performance monitoring. At this point I think some good work has been done looking at how to define performance relative to WLAN.
I would point you at several researchers starting with recent work by Samarth Shah (University of Illinois Urbana-Champaign) available at the agenda page of the Bandwidth Estimation Workshop in San Diego 2003
(http://www.caida.org/outreach/isma/0312/agenda.xml).
His presentation is Available bandwidth estimation in IEEE 802.11-based wireless networks. A number of other presentations are also highly relevant, especially anything aimed at "available bandwidth" and timescales for sampling techniques.
I would also suggest you look into the various works of Ashok Agrawala who has done some brilliant work with regard to positional triangulation (an unobvious key to dealing with the mobility and capacity negotiation aspects of the WLAN).
I suspect there is a lot of other good work being done in this area that has not fully matured yet and I would suggest you look into publications like IEEE Wireless Communications, Journal of Wireless Networks, Mobile Computing and Communications Reviews.
Here's a few more excellent references:
http://www.nwc.com/1408/1408sp3.html?ls=TW_050903_rev
http://www.nwc.com/showitem.jhtml?articleID=17600132
Measurement Approaches to Evaluate Performance Optimizations for Wide-Area Wireless Networks by Rajiv Chakravorty, Julian Chesterfield, Pablo Rodriguez and Suman Banerjee
|
