FAST TCP over optical burst switched networks: Modeling and stability analysis

Abstract

Abstract-FAST TCP is important for promoting dataintensive applications since it can cleverly react to both packet loss and delay for detecting network congestion. This paper provides a continuous time model and extensive stability analysis of FAST TCP congestion-control mechanism in bufferless Optical Burst Switched Networks (OBS). The paper first shows that random burst contentions are essential to stabilize the network, but cause throughput degradation in FAST TCP flows when a burst with all the packets from a single round is dropped. Second, it shows that FAST TCP is vulnerable to burst delay and fails to detect network congestion due to the little variation of round-trip time, thus unstable. Finally it shows that introducing extra delays by implementing burst retransmission stabilizes FAST TCP over OBS. The paper proves that FAST TCP is not stable over barebone OBS. However, it is locally, exponentially, and asymptotically stable over OBS with burst retransmission.

An improvement over TCP Vegas to enhance its performance in optical burst switching networks

“…Based on the result of the research in [4], burst contention can cause throughput degradation on the performance of Fast TCP Vegas. The first goal of this research is to investigate Fast TCP over OBS networks.…”

“…Researchers have done much studying on this issue to improve the performance of optical networks running under protocols like TCP. A large number of previous enhancements have modified TCP or OBS to adapt them to optical networks [1][2][3][4][5][6][7][8][9][10][11].…”

“…Among these techniques, OBS is the one that can attain an efficient and dynamic bandwidth allocation for handling a large amount of traffic on the Internet [2][3]. Control packets and data bursts are fundamental components of OBS networks [4]. Packets are aggregated into collections called data bursts in edge nodes.…”

“…Being bufferless and deploying a one-way reservation can lead to a problem called data burst contention. When two bursts are competing on one resource (i.e., a wavelength) to reserve it, one of them will be dropped [1][2][3][4]. This packet drop can mislead protocols in detecting congestion status in the network, especially protocols such as TCP, which assumes that every packet loss is because of congestion in the network.…”

“…In these networks, due to random data burst contention, packet loss does not necessarily reflect a congested condition. This situation can lead to a significant reduction in the performance of the networks [3][4][5][6].…”