RE: [DNA-BOF] Using L2 to provide Instantaneous Movement DetectionandNeighborhood Discovery

[NJEDJOU Eric FTRD/DMR/REN <eric.njedjou@francetelecom.com>]

Bernard Aboba wrote  
 
>> The most important thing to take away from all this is that L2 >"hints" are
> generally fallible and that an implementation MUST be robust in the >face
> of misleading hints.

Bernard, i do not agree at all that L2 hints are generally fallible. We should not be having the same interpretation of what a hint is. If you consider GPRS for instance, and the activation of a PDP context (that provides the end node with an IP and underlying QoS parameters), the extraction of the "ACTIVATE PDP CONTEXT ACCEPT" message that has arrived from the network on the node's GPRS interface stack is a definitive indication that the node has IP-attached to the GPRS network. As such, the "ACTIVATE PDPD CONTEXT ACCEPT" message is not a fallible hint. It is indeed a very strong one as are most of hints in GPRS atachment procedures. It therefore can be relied on by the IP layer.

best regards

Eric Njedjou