[olsr-dev] Question about LQ theory

BILLOUDET Gwenaël (spam-protected)
Fri May 6 14:56:11 CEST 2005


I discovered olsrd few days ago. The Link Quality Extension interested me the
most. I have a question about the theory of this extension.

The link_quality is determined using the HELLO_LQ messages which are sent by
broadcast radio mode.
The problem : in 802.11's broadcast radio mode, the packets are sent with a low
rate (1Mb/s) and with a modulation which allows packets to be sent very far.
Thus, a node A may receive a broadcasted packet from a node B whereas A may NOT
receive an unicasted packet from B.

Thus, it is easy to see that the calculated value of the link_quality is wrong.

My idea would be to create a new format of packet which are send to neighbors in
unicast mode. A node can look in his neighbors list to get the IP adress of his
neighbors. The node send these packets regularly and the neighbors can
determine the link_quality properly.

Is my remark correct ? If not, tell me why.


Département Télécommunications, Services et Usages
5e année

More information about the Olsr-dev mailing list