PhyRxDrop vs RxError

[Masood Abbasi]

Hi All,

I am using an implementation of STDMA done in ns3.18. It uses broadcast mechanism to transmit without carrier sensing by its StdmaNetDevice inherited from WifiNetDevice.

1. Can any one tell me that for calculating collisions is PhyRxDrop  appropriate or RxError (I am getting different values for these two traces?)

In following thread in this group i came to know that collision probability is calculated using RxError and RxOk traces

https://groups.google.com/forum/#!topic/ns-3-users/7byjPHGIcmo

2. Also with RxEnd trace the manual says its generated when packet is successfully received, so could it be with error(collisions) or not?.

Also i get same values /count for RxEnd and RxOk traces? 

3. What trace will provide information on corrupt packets (non decode able) due to low SNR ,fading  etc.?

Thanks

[Masood Abbasi]

Should we divide RxError or phyRxDrop by number of nodes to count actual collision as traces is generated for all nodes. Assuming all nodes hear each other?