RE: stds-80220-requirements: Comment on Functional requirements d document.
On Wed, Jul 16, Jim Tomcik writes ...
Jim> I get nervous about the quality of the spec when I hear "MAC and PHY
Jim> are inextricably woven together" from some participants. For the
Jim> purposes of clean specification, we should attempt to separate the
Jim> functionality (regardless of how it actually gets built). This will
Jim> reduce down-stream problems with the spec.
Folks,
It seems to me that the coupling of MAC and PHY is being discussed in
several different contexts here. It's important to distinguish between
them.
With regard to the air interface specification, there's probably general
agreement that it is important to have a modular design with well defined
interfaces between the various functional blocks and/or layers.
With regard to the design of the MAC and PHY, it's been proposed that the
two should be jointly designed and that the highest efficiencies are
achieved when the resource allocation mechanism (MAC) is tailored both to
the traffic and to the resources (PHY). An implication of this argument
is that different PHYs may be best served by different MACs even within the
scope of 802.20.
There was at least one earlier message that discussed a third context,
namely coupling of MAC and PHY in a particular implementation of the
ultimate 802.20 specification.
As the group discusses the first two types of coupling, I would encourage
everyone to distinguish between them: they're independent. The third
type of coupling is an implementation detail and outside of the group's
purview.
Marc