Re: [manet-dlep-rg] TCP clients, servers, and discovery (WAS: Re: notes DLEP meeting @ IETF88)

[Henning Rogge <[e-mail]>]

First of course the transmitter/receiver part of the radio should
sleep as much as possible.

I would guess that you would also increase the intervals of all
"Heartbeat/Discovery" mechanisms a bit so the CPU of the radio can go
to sleep mode until the next data of the router arrives or the next
timer event happens.

It might also make sense to have a rate limitation of the metric
updates, to prevent the radio from waking up the router too much when
no traffic is incoming. Maybe a metric update would be delayed until
the radio has to deliver the next IP packet (either heartbeat or radio
traffic) to the router.

Henning Rogge

On Fri, Nov 15, 2013 at 2:27 PM, John Dowdell
<[e-mail]> wrote:
> On 15 November 2013 12:58, Henning Rogge <[e-mail]> wrote:
>> On Fri, Nov 15, 2013 at 1:55 PM, John Dowdell
>> <[e-mail]> wrote:
>>> My question would be that if I wanted to use a low power radio mesh device
>>> (e.g. 802.15.4), I would like the radio to have the fewest cycles to perform
>>> so that the sleep mode (which might be 90% of its life) functions as normal.
>>
>> Do you mean an IEEE 802.15.4 radio attached via ethernet to a local
>> system or an embedded system with integrated radio that use DLEP as a
>> generic API for its software?
>>
>> Henning Rogge
>
> I would want to use them attached via Ethernet, but even if the module
> is embedded and runs over SPI bus or similar, the concept should be
> the same, maybe using a virtual NIC.
>
> John



-- 
We began as wanderers, and we are wanderers still. We have lingered
long enough on the shores of the cosmic ocean. We are ready at last to
set sail for the stars - Carl Sagan