Enhanced FAST TCP by Solving Rerouting Problem
Delay-based congestion control algorithms inability to recognize increased RTT related to rerouting from increased RTT related to congestion is their most serious problem which has serious effect on their throughput. FAST TCP is one of delay-based TCP variants that although outperforms other TCP variants in high bandwidth-delay product networks, but suffers from several problems that inhere in its procedure to estimate trip delay. The most serious of these problems is rerouting. When rerouting occurs and round-trip time (RTT) of the new path is longer than RTT of the old path, the throughput of FAST TCP decreases sharply. Because FAST misinterprets the increased RTT as result of the network congestion and consequently decreases its own window size. This paper solves this problem by considering the relationship between sending rate and observed RTT. The simulation results show the effectiveness of proposed solution to solve rerouting problem while simultaneously preserves FAST TCP prominent primitive features.
Open access licenses
Open Access is by licensing the content with a Creative Commons (CC) license.
This work is licensed under a Creative Commons Attribution-NonCommercial-ShareAlike 4.0 International License.