Figure 3 depicts the measured BPagainst the input network load for FF and the MFK-based algorithms, namely, WUFR-MFK, FR-MFK, and ENT-MFK. Here, FR-MFK performs significantly better than the baseline FF,starting to block later (at about 30% load rather than 20%), and exhibiting lower overall blocking, at least until about 40% load, when its BP converges to that of FF. Even better is the performance exhibited by WUFR-MFK, which blocks significantly less, all the way to 60% load, even when plotted in logarithmic scale.