Re: DoCoMo Implementation Issues with CTP

> The intended reason for including PAR's IP address in CTAR is
> to allow NAR to construct CTREQ and send it to PAR.
> 
okay.


> When PCTD takes place, there is no reason to do CTREQ.
>
Right.


> (Also, note that you could not match contexts using PAR's IP
> address since that could yield many contexts corresponding to
> multiple MNs.)
> 
I was also under impression that both (PAR IP & MN IP) 
from CTAR & PCTD must match before verifying the token.
Otherwise how can nAR determine from 
which pAR a MN is comming from ? 


> The token calculation does not include PAR's address. It
> only includes MN's previous address (note: when V is 10 in
> CTAR, the token should include both the IP addresses).
> Since the MN includes both its addresses and NAR has
> contexts for both of those addresses, they can be matched.
> The token calculation on NAR is also done in a similar way
> as in the MN; PAR's address is not a parameter in computing
> the token.
> 
> Is this better ?
> 
Yes.


> MN includes the PAR's address in CTAR. NAR should use that
> address in the Dest field of CTREQ.
> Perhaps we should state that the MN should only include the
> IPv4 address of PAR (NAR).
> 
It helps.

regards,
Raghu