Thomas Kunz
Systems and Computer Engineering
12
Conclusions
nOLSR_R1: simple change, with little additional overhead,
but already significant improvement
(mostly due to change to routing
table calculation)
nOLSR_R2 and OLSR_R3 seem to be perfect, paper contains proof that both MPR selection criteria allow to
reliably detect optimal
path
nOLSR_R2 selects fewer MPRs, less overhead, therefore preferred
nTransmission range has important impact: higher range means denser network but not necessarily higher overhead
or error rate