[Tinyos-host-mote-wg] [Tinyos-2.0wg] Network stack/AM proposal
Philip Levis
pal at eecs.berkeley.edu
Mon May 30 18:00:06 PDT 2005
Background: last telecon, we briefly discussed what the HIL for a
radio is. I raised the question of whether it is AM or something
below AM (e.g., a raw data link packet). Vlado commented that he
thought it should be the latter.
I noted that this separation means that, for AM to be hardware
independent, it must have its own addresses and fields, independent
of the underlying data link layer. E.g., it would preclude some of
the 1.x 802.15.4 techniques, e.g., of using the 15.4 address as the
AM address. The real kicker to this is that it can preclude software
controlled synchronous acknowledgements (e.g., on the mica2), or, in
order to allow synchronous acknowledgements, will require software to
map between data link and AM addresses (e.g., on 802.15.4). Going
down the ARP/RARP road seems like a bad idea.
I argued that Active Messages are the radio HIL. This means that
there's a platform dependent implementation, allowing the kinds of
optimizations that we've seen historically. This does not mean that
there cannot be additional HILs, such as raw data link packets; it
merely means that AM implementations are chip dependent.
Specifically:
Every radio chip has a component with the name of <chip>AM, which
encapsulates AM-level communication, providing a parameterized
AMSend, Receive, as well as Packet, SplitControl and AMPacket. Every
platform that supports AM has a component ActiveMessageC, which wires
to the correct chip component.
This naming allows a platform to have two radios, support AM on both,
but specify one as the "default" AM interface.
This leaves the AM packet structure completely hidden from anything
above the HIL, which is good: only the HIL knows the format, and it
can be different for different implementations. E.g., the 802.15.4
can generate length() from the 15.4 length, while the mica2/CC1000
has an AM length field.
Phil
-------
"We shall not cease from exploration
And the end of all our exploring
Will be to arrive where we started
And know the place for the first time."
- T. S. Eliot, 'Little Gidding'
_______________________________________________
Tinyos-2.0wg mailing list
Tinyos-2.0wg at Mail.Millennium.Berkeley.EDU
https://mail.millennium.berkeley.edu/cgi-bin/mailman/listinfo/tinyos-2.0wg
More information about the Tinyos-host-mote-wg
mailing list