Performance of Robust AQM controllers in multibottleneck scenarios

Recently, novel robust AQM congestion control strategies for time delay systems have been proposed. These are based on the use of robust reduction methods and H??? strategies resulting in memory and memoryless control laws respectively have been proposed. Their design was carried out on a single bottleneck topology for guaranteeing robust local stability in the presence of network uncertainties in terms of average round-trip time, load and the link capacity. Here we analyze the performance of such robust AQM control strategies on multibottleneck scenarios by NS-2. We will show that despite their original design on single bottle- neck link, they can still guarantee better performance than other more traditional AQM controllers even in the multibottleneck case.