Bugzilla – Bug 1888
COST231 - corrections
Last modified: 2014-03-26 18:48:44 UTC
Created attachment 1806 [details] Patch that addresses bug issues. The formula that calculates COST231 path loss that is implemented in GetLoss() function of src/propagation/model/cost231-propagation-loss-model.cc is not fully in accordance with corresponding COST231 algorithm ( http://en.wikipedia.org/wiki/COST_Hata_model) Following points have mismatch: - units of distance should be in km and not in meters - units of frequency should be in MHz instead of GHz - it is used log with some corretcting multiplier instead of using just log10 that is according to COST231 algorithm, this produces some errors. In attachmente is patch that adresses mentioned issues.
+1
In the docs, please refer to: http://www.lx.it.pt/cost231/final_report.htm Ch. 4, eq. 4.4.3 ì, pg. 135 instead of using wikipedia. Thanks
Hi Tommaso, thanks for reply. I refered to wiki because that was the reference in original ns3 code. I took a look into the reference for COST231 that you gave, and my conclusion is that there are still those issues in the ns3 code: - frequency used in COST231 calculation should be in MHz, not in GHz as it is in current ns3 code - distance used in COST231 should be in km, not in m - and on page 134, footnote, of the document that you gave it is said that "1) "log" means "log10"". Should I add the reference to the code and generate a new patch? Thanks Biljana
Hi, the Wiki page reports exactly the same formula as the COST chapter 4. I like more the original source purely because the wiki page can be changed, the final COST report can't :) Feel free to link both. Probably it's easier for most people to go to the wiki page. There's no need to re-generate the patch. I trust you and, after all, it's just a comment. T.
changeset: 10684:156ee4e1f4aa