> At 08:36 PM 10/17/00 +1100, Hesham Soliman (EPA) wrote:
> >=> I'm not sure if I understand what you mean by the foreign agent.
> >Perhaps you have another entity in mind ? Because the current MIPv6
> draft 
> >does not have foreign agents.
> And therefore doesn't have a soft hand off option unless one can literally
> 
> talk with both routers during the hand off. 
> 
        => As far as I'm aware the term "soft handoff" is related to
        certain radio technologies that allow data to be received
        from multiple access points. This in no way implies receiving
        two separate traffic streams. It is an L2 feature that is used 
        for handoffs and also as a mechanism for improving the 
        received signal quality. 

        IMO having foreign agents is not relevant for achieving soft 
        handoffs. In fact the soft handoffs concept (as used today)
        is completely transparent to MIP (v4 and v6) or any L3 
        function as it happens on L2.

> The Hierarchical Mobility draft 
> suggests having an external system offer one of its addresses a second 
> care-of address to provide this; to me, that's a foreign agent in
> disguise, 
> and would be useful in that case. Sorry I was quick on the language.
> 
        => No problems. I obviously agree with comment on the 
        usefulness of the MAP function. 
        We avoided using the term FA in the draft because
        it is associated with the current understanding of FAs in IPv4. 
        There certainly are differences. 

        - FAs are mandatory on each link in MIPv4. MAPs are not.

        - FAs in MIPv4 perform other tasks like authentication 
          which is a AAA function and need not be tied to mobility
          functions.

        -  A MN can choose MAPs in the hierarchy 
           based on various information in the MAP option.
           You can't do that with FAs.

        - MAPs can exist on any level of a hierarchy including
          first hop routers and are completely independant from
          each other.

        You can probably see more differences in the draft.

        What we tried to do is go away from the FA concept because
        we don't believe FAs ( as introduced in V4) are needed in IPv6.
        Perhaps some of the functions in FAs can be useful. But it 
        seems unnecessary to copy and paste the functionality into IPv6. 

        Hesham
--------------------------------------------------------------------
IETF IPng Working Group Mailing List
IPng Home Page:                      http://playground.sun.com/ipng
FTP archive:                      ftp://playground.sun.com/pub/ipng
Direct all administrative requests to [EMAIL PROTECTED]
--------------------------------------------------------------------

Reply via email to