Home > Uncategorized > LBR, PBR, Flooding and Random comparison

LBR, PBR, Flooding and Random comparison

November 21st, 2010 Leave a comment Go to comments

This graph shows the delivery ratios from the start of the simulation for MIT-OCT dataset. It would appear that in fact, LBR performs poorly against the random algorithm, where the P is the probability of passing on the message (1 copy) to the met node. Interestingly though, when P = 1, the delivery ratio increases. The graphs below show the delivery cost for the different algorithms, spread across two plots, so that the very high costs of Random and Prophet do not dwarf the lower costs of LBR and PBR. These three plots together show that whilst the delivery ratio of  prophet and random are high, the cost is very high.

Delivery costs for three routing algorithms on the MIT-OCT dataset, spread over two plots for clarity.

Perhaps a combined metric for comparison is  cost over delivery ratio, this would perhaps give a better idea of general quality of the algorithm.

Categories: Uncategorized
  1. No comments yet.
  1. No trackbacks yet.