In real life, if you don't boost to the maximum power your access point, you will have no connection at all in ad-hoc mesh networks at 2.4Ghz.
Moreover, the other interfering nodes will not run the proposed algorithmn (because are not from your network), so if you lower the txpower you will isolate that node. Again, this proposal is cross-layer. You are binding the routing protocol to a specific radio equipment. Is this what you want ? We want to get rid of ISO/OSI so quickly ? Please before starting to propose new ideas, everybody should have a look to the state of the art, like OLSR, BABEL, BATMAN, the Freifunk network in Berlin etc etc .... we don't need to start from 0 again :) Saverio 2010/4/12 Korvin <[email protected]>: > Hello again. > > Could someone tell me, whether is it possible to tune AP transmit power/range > individually for each node? I know that modern wlan cards support this > feature. > > My idea is to tune node's range in order to optimize Netsukuku covering. > > In normal conditions, redundancy is good, but when amount of APs grows, they > begin interfering with each other, causing degradation of overall speed. If > we'll try to manage the tx power, things would be better. > > So, each node knows amount of other devices in range. If there are no devices > around, or it's count is small, node transmits signal at full power, > maximizing the covering. > > When amount of in-range devices grows, node may slightly decrease it's range > and try to detect actual range to other devices. By knowing alternate routes > to peers we may detect whether peer is fully dependant on our node, or it has > another links to the world. > > Then we may decide, whether to decrease the range forcing remote peer to use > other links, or to leave the things untouched. > > In case of overpopulated area, there is high possibility that node may have > many routes to the world, so this would not be a problem. > > So: > > 1) if reducing of signal strengh would increase the overall quality of service > (because of reduced interference) and out-of-range devices may switch to other > nodes, then it is a good idea. > > 2) If reducing of signal strengh leads to loss of connectivity with large > amount of nodes or even netsplit is occured, then it is a bad idea, so this > node is important to network and actually needed to be fully powered. > > Actual algorithm is need to be discussed, but I think it may be some kind of > minimax. Of course this operation need to be tied to actual network topology > and range detection should be done using 'non invasive' method. > > _______________________________________________ > Netsukuku mailing list > [email protected] > http://lists.dyne.org/mailman/listinfo/netsukuku > _______________________________________________ Netsukuku mailing list [email protected] http://lists.dyne.org/mailman/listinfo/netsukuku
