[Tinyos-host-mote-wg] Re: TinyOS 2.0 host-mote working group

[Philip Levis]

On Apr 11, 2005, at 10:26 AM, Gilman Tolle wrote:

> The only way for a host to discover whether a mote is attached is to
> send a data message to a (hopefully) unused AM type, and listen for
> the ACK.
>

Yes. Problematic.


> All messages passing over the serial link are framed in the radio's
> link header. Therefore, communication between a PC and a mote is
> impossible without a priori knowledge of the radio installed on the
> mote.

Yes, problematic.

>
> The mote radio and the serial link have mismatched speeds, which
> causes packet loss under heavy contention for the serial link.
>

This is just a basic problem on any device with two network interfaces. 
I'm not sure if a protocol should deal with this, although providing 
some kind of support to detect it's happening is probably a good idea.


> Some motes may send multiple different types of packets over the
> serial link. For example, consider a TOSBase that sends and receives
> 802.15.4 data packets as different from a hypothetical 15.4 packet
> sniffing application that sends 15.4 beacons, data packets, and acks.
>

I think this is just a similar problem to the one above.

> The types of messages that should flow from a host to a mote are
> different from those that travel between motes. For example, a message
> that configures the addressing information or changes the radio state
> of the attached mote.

Why do there need to be operations that are only invokable from the 
serial link? This would mean that the only way to invoke them is to 
plug in the mote. Why not just make these AM handlers? If you care 
about corrupted or problematic radio packets, then you can wire to a 
UART-only AM interface.

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'