On Thu, 2011-04-28 at 14:15 -0600, Caleb Phillips wrote: > Thanks for your response, Dan... > > >> I've searched the mailing list and bugzilla without avail, for what > >> seems like a very obvious problem. Whenever I do a scan, wide or normal, > >> the scan results all return with an identical RSSI and SNR. This > >> behavior can even be seen in the documentation for 1.5.2: > > > > That's because the same exact tower and radios broadcast the IDs for > > each of these networks. CLEAR is the actual provider, but the other 3 > > are simply "virtual operators" or really "Network Service Providers" > > while Clear provides the towers. So it's no wonder that they all look > > exactly the same except for the ID: because they are all coming from the > > same exact radios at the base station. > > > > If there was more than one WiMAX provider in your area (not just CLEAR > > and 3 others reselling CLEAR under their own name) then you'd see > > different results. > > At my site on the University of Colorado, I have at least 3 distinct > BSes on different buildings and operating at different center > frequencies, so it stands to reason they would have different CINR/RSSI > stats. I also see other operators, such as CLEAR and SPRINT networks in > my area and those show up with the same RSSI/CINR as well. The middleware will pick up one and present just that one -- probably the highest ones. So unless there is an specific BS <->NSP mapping, where each NSP is being serviced by a different basestation, you wouldn't see it. > I'd be happy to reproduce this and share debug information if you let me > know what would be useful to see (I reckon the output of wimaxcu scan > wide isn't helpful). > > >> I'm a researcher at the > >> University of Colorado and am hoping to use the driver in a measurement > >> platform. So, what I'd really like is more detailed scan results (per > >> station frequency, bsid, CINR, RSSI, etc.). I've tried to determine if > >> this information exists somewhere but as far as I can tell, what's in > >> wimaxd.log is probably my best bet (without substantial, and deep > >> modifications to the driver)? > > Any thoughts on my second question, w/r/t how to get more detailed > per-BS performance stats from the scan results? nope, only what the CommonAPI provides, which I agree, is very general.
