On 1/27/2015 7:02 AM, Alexandru Petrescu wrote:
...
Addition: new wifi links on the market since this draft was
experimented, show other new behaviours unkown until now to wired
counterparts.
For example, 802.11ac not only offers more bandwidth, but comes by
default with 2 ready to use independent channels, as two independent
links on 2.4 and 5.x bands.
This is like if each Ethernet cable were actually made of two
independent cables. You'd need to stitch both their ends together by
some software to use the available bandwidth.
If only there were an existing link model for this.
Oh, right - there is. It is called channel bonding.
I'm not saying that MANETs operate using wired-link assumptions, but
rather that wireless links - and everything that encompasses - have been
part of the Internet architecture for decades.
It's far too early to discuss whether this doc could *become* a useful
WG doc; in it's current form, there's simply no hint of that yet. IMO,
the authors should take another shot at it, and then we can see if
there's really anything useful to pursue here. The current doc does NOT
make that case yet.
Joe
_______________________________________________
Int-area mailing list
[email protected]
https://www.ietf.org/mailman/listinfo/int-area
