What's the real scope of hardware openness?

John Seghers jseghers at cequint.com
Tue Aug 7 17:49:24 CEST 2007


Luca Dionisi wrote:
> 
> Many NEOs -> many batteries.
> Few NEOs -> few power need.
> Scaleable, isn't it?
> 
> I know I'm simplifying.  I'm not saying it's trivial.

The main problem here is that the WiFi hardware was not designed for the
ultra-low-power modes required to extend the "standby" time.  Note how most
phones today can go for days without recharging while waiting for a call,
but get only a few hours of talk time.

GSM and CDMA radio systems are designed to use an absolute minimum of power
when in standby. They broadcast location occasionally, but most of the time
they are just listening--and listening only to the "housekeeping" channel.

In a mesh network using WiFi, however, the nodes need to be active any time
they are being used for data transfer.  Even topological negotiation is
going to require data transfer.

> E.g. the previously mentioned link says:
> Netsukuku is designed to handle an unlimited number of nodes with minimal
> CPU and memory resources. 

Unfortunately this doesn't say anything about power requirements.

- John





More information about the community mailing list