[Tinyos-help] (no subject)

[Philip Levis]

On Jul 6, 2006, at 11:21 AM, Adam wrote:

> Hi,
>
> Anyone has empirical results to adjust tranmission power of CC2420  
> based on
> LQI -- like some linear equation? I have tried adjustLQI() in
> MultihopLQI.nc, the return value is not stable and oscillate in a  
> large
> range. Actually, I wonder why not use the original LQI in TOS_Msg  
> header?
> Since CC2420 datasheet said: "A correlation value of ~110 indicates a
> maximum quality frame while a value of ~50 is typically the lowest  
> quality
> frames detectable by CC2420."
>
> Anyone successfully get a linear relation between the LQI value and  
> actual
> link quality?

It's not a linear relationship. Kannan recently published a paper  
(EmNets 2006) that scratches the surface of the topic, and should  
soon (in the next week) have a tech report that looks at it more  
deeply. You can find the paper here:

http://csl.stanford.edu/~pal/pubs.html

Phil