Next: Alternate Path Length
Up: Experimental Evaluation
Previous: Experimental Evaluation
In Experiment 1, DCN and OTN were expanded two-fold10at every run. The pure-random model was used for generating edges with
edge probabilities 0.2311 and 0.13 for the OTN and DCN respectively. The average
restoration time from all primary paths to all secondary paths in each
randomly generated OTN was measured for each of the four activation
schemes. Figure 7 shows the performance of each
of the four
activation schemes for different sizes of the OTN and the size of the
DCN (scaled down ten-fold) at which the measurement was performed.
As seen in the figure, the proposed serial and parallel OBS-based
schemes outperform their non-OBS counterparts.
Restoration time v/s OTN and DCN size
(Edge probability: DCN = 0.13, OTN = 0.23).