One thing that very seriously bothers me about this paper is the sentence:
"Multipath appears when obstacles exist between the Base Station and Subscriber location."
Actually multipath probably happens 100% of the time - but the unintended signals are not strong enough to capture the intended signal. Multipath can in fact exist without LOS obstructions, and indeed happens in MANY LOS cases where there are any number of surrounding or even far off objects
In the case of these studies, which can not be determined clearly because all of the visual and terrain data are not available about items to the sides or rear of the units, there is a stated dependency on multipath. If this is the case, if you were to have only a blockage within LOS and NO side or back reflective elements there is nothing to get MP from, and this would likely not work (knife-edging, etc. not withstanding.)
Anyone know how such things work in NLOS AND non-MP scenarios??
parsed original message:
"Tim, your understanding of OFDM is way off. Extraordinarily stability is achieved with OFDM - far more than with our FHSS or our DSSS. We have examples of links having only a single packet error over several weeks. I am not basing these off theory, but off actual field results both beta and commercial deployments. Here is a link to some such tests done using our 3.5GHz OFDM PMP which has been shipping for 18 months and is in use in major carrier deployments around the world. This paper includes detailed photo examples of the links.
http://www.alvarion-usa.com/RunTime/Materials/KnowledgePoolFiles/alv_OFDM%20 wp.pdf"
========================================================
/s/ Jim
There are many things that motivate people. The trick is to figure out what motivates who you are dealing with at the moment.
"lack of (the right) information is a dangerous thing" /// "a confused mind says no"
B.A.R.F. UHF Repeater - 443.750 - San Jose PL 100 - Vaca PL 127.3
What I did in my spare time: http://books.mcgraw-hill.com/cgi-bin/pbg/0071410708.html
========================================================
-- general wireless list, a bawug thing <http://www.bawug.org/> [un]subscribe: http://lists.bawug.org/mailman/listinfo/wireless
