What is the exact difference between the key update and re-keying? Following the discussions here, I understood that the key update is the process of deriving the new session key from old key. In the case of re-keying, the new key is independent of old key. My questions are as follows:
Can key-update protect future messages from a compromise of an old key in any way? Is the re-keying only solution to achieve this?
In the case of re-keying, is it always necessary to perform the authentication process between communicating entities prior to every re-keying, especially when sensing devices are deployed in a hostile environment? If not, how communicating devices will trust each other?
If I am using ECC based key establishment protocol with a pre-shared master key for resisting the meet-in-the-middle attack, what are the possible options for the key update on resource constrained devices?