HighSpeed TCP for Large Congestion Windows

By Sally Floyd. Request for Comments, 2003

1,428 citations recorded by the source corpus.

Subfields: Artificial Intelligence, Computer Networks and Communications

View Paper: https://www.semanticscholar.org/paper/6aba884d517ab35a63e74da36d20207965eb7e0d

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Artificial Intelligence

1. Mung Chiang
2. A. Robert Calderbank
3. Robert N. Shorten
4. Douglas J. Leith
5. Injong Rhee
6. Sanjay Hegde
7. John Bresnahan
8. John C. Doyle

In Computer Networks and Communications

1. Yunhong Gu
2. Robert L. Grossman
3. Mung Chiang
4. Steven H. Low
5. John C. Doyle
6. A. Robert Calderbank
7. Robert N. Shorten
8. Hari Balakrishnan

Citations in the corpus

• Binary increase congestion control (BIC) for fast long-distance networks (2004, methodology): “That is, the loss rate cannot be more than 1 loss event per 2,600,000,000 packets, which is less than the theoretical limit of the network’s bit error rates [1, 2].”
• UDT: UDP-based data transfer for high-speed wide area networks (2007, methodology): “Composable UDT supports a wide variety of control algorithms, including but not limited to, TCP algorithms (e.g., NewReno [24], Vegas [10], FAST [37], Westwood [25], HighSpeed [23], BiC [67], and Scalable [40]), bulk data transfer algorithms (e.g., SABUL [26], RBUDP [33], LambdaStream [66], CHEETAH ”
• A Compound TCP Approach for High-Speed and Long Distance Networks (2006, background): “At time E, the queue begins to build, the CTCP senses this from the increase in the packet delay, and dwnd decreases to try to maintain a stable window.”
• The Globus Striped GridFTP Framework and Server (2005, methodology): “These solutions include improvements to TCP [21, 35, 39], new transport protocols such as XCP [38], XTP [54] and reliable layers on top of UDP [5, 15, 16, 29, 32, 35, 53].”
• FAST TCP: Motivation, Architecture, Algorithms, Performance (2006, methodology): “In contrast, the packet-level designs of HSTCP [15], STCP [32], and FAST TCP are explicitly guided by flow-level goals.”
• Why flow-completion time is the right metric for congestion control (2006, methodology): “There have been many proposals recently, such as HighSpeed TCP [14] and Scalable TCP [15], to improve TCP’s AIMD behavior.”
• H-TCP : TCP for high-speed and long-distance networks (2004, methodology): “(17) This choice of α H i yields a response function similar to that of HS-TCP [7].”
• Layering as Optimization Decomposition: A Mathematical Theory of Network Architectures (2007, methodology): “Several new algorithms have been proposed in the last few years to address the problems of Reno, including TCP Westwood, HSTCP [42], FAST TCP [56], STCP [67],BIC TCP [156], HTCP [90], MaxNet [154], [155], XCP [62], and RCP [33], etc.”
• CUBIC: a new TCP-friendly high-speed TCP variant (2008, methodology): “HighSpeed TCP (HSTCP) [15] uses a generalized AIMD where the linear increase factor and multiplicative decrease factor are adjusted by a convex fucntion of the current congestion window size.”
• Scalable TCP: improving performance in highspeed wide area networks (2003, methodology): “This work builds on the Highspeed TCP proposal [6] and uses the same arguments to achieve good sharing with legacy applications.”
• TCP-Illinois: a loss and delay-based congestion control algorithm for high-speed networks (2006, neutral): “9Henceforth, we mean SACK based HS-TCP when we men­tion HS-TCP without specifying its base.”