CUBIC: a new TCP-friendly high-speed TCP variant

  title={CUBIC: a new TCP-friendly high-speed TCP variant},
  author={Sangtae Ha and Injong Rhee and Lisong Xu},
  journal={ACM SIGOPS Oper. Syst. Rev.},
CUBIC is a congestion control protocol for TCP (transmission control protocol) and the current default TCP algorithm in Linux. [] Key Result This paper documents its design, implementation, performance and evolution as the default TCP algorithm of Linux.

Figures and Tables from this paper

“A survey on TCP CUBIC variant regarding performance”

In this Paper, CUBIC performance evaluation is done by Network Simulator-2 experiments and performance analysis of other TCP versions is also discussed.

Understanding synchronization in TCP Cubic

It is demonstrated that several factors contribute to the appearance of synchronization in TCP Cubic: the rate of growth of the congestion window when a Cubic source reaches the capacity of the network and its relation to the RTT of the connection, the way the congestion Cubic tracks the ideal cubic curve in the kernel and the competition among the Cubic sources.

Stateful-TCP—A New Approach to Accelerate TCP Slow-Start

This work proposes a different approach called Stateful-TCP, where path bandwidth estimated in a previous flow is used to instantly ramp up the transmission rate of the subsequent flow to the same destination, enabling TCP to efficiently utilize the available path bandwidth right from the beginning.

Should paced TCP Reno replace CUBIC in Linux?

  • Santosh Chavan
  • Computer Science
    2016 8th International Conference on Communication Systems and Networks (COMSNETS)
  • 2016
This work shows that in a many flows regime, with or without selective acknowledgements (SACK) support, paced Reno performs at the most, or better than CUBIC and Compound TCP over a single bottleneck link with Drop-Tail buffers, and suggests that paced Reno can possibly replace CUBic in the Linux OS.

A comprehensive study of modern and high speed TCP-variant in linux kernel: TCP CUBIC

The CUBIC design architecture with the pseudo code of the algorithm, TCP support in Linux kernel and implementation, and the performance of CUB IC is evaluated both in wired and wireless environment under the parameters of goodput and intra-protocol fairness.

TCP Internal Buffers Optimization for Fast Long-Distance Links

  • A. BaiocchiS. MascoloF. Vacirca
  • Business, Computer Science
    Proceedings IEEE INFOCOM 2006. 25TH IEEE International Conference on Computer Communications
  • 2006
This work aims to investigate, by using both analytical models and simulation results, optimal sizing of the retransmission and out-of-order buffers in case of modified TCP congestion control settings and to highlight differences between NewReno TCP and SACK TCP packet loss recovery schemes in terms of TCP internal buffers requirements.

Evaluation of TCP Based Congestion Control Algorithms Over High-Speed Networks

This work aims to explore the following research questions: how tuning affects the performance of TCP and over 10G networks, and compare TCP variants over a high-loss back-to-back environment.

TCP congestion control algorithm for heterogeneous Internet

CLTCP: An Adaptive TCP Congestion Control Algorithm Based on Congestion Level

This letter proposes a novel scheme, i.e., the congestion-level-based TCP (CLTCP), which could perform effectively in both high-BDP and lossy networks and does not use the delay variation but the congestion level as a signal to control the number of virtual parallel flows in a TCP connection.

QTCP: An Approach for Exploring Inter and Intra Protocols Fairness

A new Quick Transport Control Protocol (QTCP) has been designed to overcome the congestion issue and the results based on NS-2 simulator shows that QTCP provides better performance compared to other transport control protocols.



TCP-Illinois: a loss and delay-based congestion control algorithm for high-speed networks

This work introduces a new congestion control algorithm, called TCP-Illinois, which has many desirable properties for implementation in (very) high-speed networks and extends recently developed stochastic matrix models of TCP to accommodate window size backoff probabilities that are proportional to arrival rates when the network is congested.

Scalable TCP: improving performance in highspeed wide area networks

The preliminary results gathered suggest that the deployment of Scalable TCP would have negligible impact on existing network traffic at the same time as improving bulk transfer performance in highspeed wide area networks.

Binary increase congestion control (BIC) for fast long-distance networks

This work presents a new congestion control scheme that alleviates RTT unfairness while supporting TCP friendliness and bandwidth scalability, and uses two window size control policies called additive increase and binary search increase.

TCP westwood: Bandwidth estimation for enhanced transport over wireless links

TP Westwood is a sender-side modification of the TCP congestion window algorithm that improves upon the performance of TCP Reno in wired as well as wireless networks and is extremely effective in mixed wired and wireless networks where throughput improvements of up to 550% are observed.

Evaluation of Advanced TCP Stacks on Fast Long-Distance Production Networks

This paper evaluated new advanced TCP stacks that do not require multiple parallel streams while giving good performances on high speed long-distance network paths and compared the New Reno TCP with the following stacks.

H-TCP : TCP for high-speed and long-distance networks

The new protocol, H-TCP, is shown to fair when deployed in homogeneous networks, to be friendly when competing with conventional TCP sources, to rapidly respond to bandwidth as it becomes available, and to utilise link bandwidth in an efficient manner.

FAST TCP: Motivation, Architecture, Algorithms, Performance

FAST TCP is described, a new TCP congestion control algorithm for high-speed long-latency networks, from design to implementation, and its equilibrium and stability properties are characterized.

TCP Veno: TCP enhancement for transmission over wireless access networks

  • C. FuS. Liew
  • Computer Science, Business
    IEEE J. Sel. Areas Commun.
  • 2003
This work proposes and study a novel end-to-end congestion control mechanism called Veno that is simple and effective for dealing with random packet loss in wireless access networks and can achieve significant throughput improvements without adversely affecting other concurrent TCP connections, including other concurrent Reno connections.

Experiences in Design and Implementation of a High Performance Transport Protocol

The rationale behind UDT is presented: how UDT integrates these schemes to support high performance data transfer, why these schemes are used, and what the main issues are in the design and implementation of this high performance transport protocol.

Equation-based congestion control for unicast applications

A mechanism for equation-based congestion control for unicast traffic that refrains from reducing the sending rate in half in response to a single packet drop, and uses both simulations and experiments over the Internet to explore performance.