Using the TTL, we figured out that the censor kicks in at the kth hop. This kth box belonged to Airtel.
If a box was censoring after Airtel, we would have received a clean response (ICMP timeout) at hop k as well. Of course, the TTL itself can change during the run, but that wouldn't happen for so many cases :)
If a box was censoring after Airtel, we would have received a clean response (ICMP timeout) at hop k as well. Of course, the TTL itself can change during the run, but that wouldn't happen for so many cases :)