Hi,
Thanks for your reply.
>>
> It partially depends upon the nature of the traffic. Simple cases
> wherein the data is encapsulated via tcp AND ip allow you to infer which
> interface sent the traffic and additionally allow you to infer
> temporality-tracking differences between hosts by matching up timestamps
> and sequence numbers. Other packet-formatting combinations may provide
> fewer data-points for you to utilize while engaging in the measurement
> exercises you indicated interest in. I'd be interested to hear if
> anyone's successfully made use of protocol-independent techniques.
Mostly I am concerned with TCP but a protocol-independent technique is
definitely preferred! Anybody?
Thank you,
Andrej