Conclusions
	n OLSR_R1: simple change, with little additional overhead, but already significant improvement (mostly due to change to routing table calculation) n OLSR_R2 and OLSR_R3 seem to be perfect, paper contains proof that both MPR selection criteria allow to reliably detect optimal path n OLSR_R2 selects fewer MPRs, less overhead, therefore preferred n Transmission range has important impact: higher range means denser network but not necessarily higher overhead or error rate