We use DHCPv6 to assign just one IP address to the CPE. This is because otherwise our routers do not know where to route the /48 that is also passed along with DHCPv6-PD. The routers are stupid I know, but it is what we got. So we simply implemented a variant of static routes for 2001:db8:x::/48 to 2001:db8::x/128. The DHCP server knows to give you matching /48 and /128. Apart from operational simplicity, we also do not want our routers to keep track of a million ND cache entries. Our system pushes that down to the CPE. In the network we only have one ND cache entry per customer. The Android tethering guy seems to think that tethering should be a bridge. But it should of course be routed. The phone in tethering mode should be getting exactly what we do - one /128 on the uplink interface and a ton of address space it can use internally and sub delegate to tethering clients. What exactly is the argument against supporting a sane environment like that? As a side note, NAT is not the only solution if someone should try to block tethering. I would propose a VPN tunnel. You can then have as much address space you want from the VPN. This is extra easy if you are not locked into the belief that tethering should be a bridge instead of routed. Regards, Baldur