yes, as we have a PAT rule that makes it difficult to track down what did what,
the scenario here is that we have a virus detected from without the organization coming from our public, incedental use wireless (users are identified but anyone can get a day use username and pw) it provides a port, IP, and time.
so yes, we are interested in the translates: but as you suggest just collecting guest user NAT translates would be better (as this is the only PAT'ed range).
how could I collect such a thing? Suggestions?