IPv6 Tunneling (reverse tunnel need)

I'm currently working on a research project and ran into a 
question regarding tunneling/decapsulation in IPv6.

<------> [ A ] ===== tunnel from A to B =====> [ B ] <------->  

Packets tunneled from A to B are only decapsulated at node B if
B has setup a tunnel-device in direction from B to A (reverse tunnel).

Is the need for the tunnel-device on node B an implementation
issue to handle tunneling, or does it come from something 
else (i.e. standard requirement) ?

Further, I am wondering why the possibility to use the ANY_ADDRESS
as tunnel-end-point isn't present in IPv6 ?
The IPv4 implementation allows the usage of "0.0.0.0" as remote_address for
tunnel-end-points.


Any help would be greatly appreciated.

Regards,
Felix Fischer
-- 
-------------------------------------------------------------------
Institute of Communication Networks and Computer Engineering (IKR)
University of Stuttgart, Germany
-------------------------------------------------------------------