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

By Sangtae Ha, Injong Rhee, Lisong Xu. OPSR, 2008

2,692 citations recorded by the source corpus.

Subfields: Computer Networks and Communications, Electrical and Electronic Engineering

View Paper: https://www.semanticscholar.org/paper/31238dbf9945bf934663637b2013ff2cc8971c4b

How researchers see this paper

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

In Computer Networks and Communications

In Electrical and Electronic Engineering

Citations in the corpus

Presto: Edge-based Load Balancing for Fast Datacenter Networks (2015, methodology): “We ran the default TCP implementation in the Linux kernel (TCP CUBIC [27]) and set parameters tcp_sack , tcp_fack , tcp_low_latency to 1.”
Deadline-aware datacenter tcp (D2TCP) (2012, background): “High-speed TCP [7], CUBIC [22], and XCP [15] all successfully improve the performance of TCP in high bandwidthdelay-product networks.”
Multipath QUIC: Design and Evaluation (2017, methodology): “Both single-path TCP (in the Linux kernel) and QUIC (in the quic-go and the Chromium 3 implementations) use CUBIC [19].”
Understanding Operational 5G: A First Measurement Study on Its Coverage, Performance and Energy Consumption (2020, methodology): “However, the traditional loss/delay based TCP algorithms suffer from extremely low bandwidth utilization— only 21.1%, 31.9%, 12.1%, 14.3%, for Reno, Cubic, Vegas, and Veno, respectively!”
Adaptive forwarding in named data networking (2012, background): “, TCP Vegas [17], TCP CUBIC [32] and TCP NewReno [34].”
Data center TCP (DCTCP) (2010, background): “Much research has been devoted, with success, to improving TCP performance on paths with high bandwidth-delay product, including High-speed TCP [8], CUBIC [30] and FAST [33].”
Sharing the Data Center Network (2011, methodology): “We derive the distributed control loop in Class 1 from TCP-Reno though any other ﬂow-oriented protocol [4, 1, 29, 32] can be used, so long as it can extend to provide weighted allocations, as in MulTCP or MPAT [11, 39].”
MPTCP Is Not Pareto-Optimal: Performance Issues and a Possible Solution (2013, background): “[26] S. Ha,I.Rhee, and L.Xu, CUBIC: A new TCP-friendly high-speed TCP variant, Oper.”
A Deep Reinforcement Learning Perspective on Internet Congestion Control (2019, methodology): “For each test, we compare the results of our model against TCP CUBIC (the current standard for congestion control), and PCC Vivace (Dong et al. 2018, a robust, state-of-the-art algorithm).”
Design, Implementation and Evaluation of Congestion Control for Multipath TCP (2011, neutral): “Newer versions of TCP [24, 9] have”
UDT: UDP-based data transfer for high-speed wide area networks (2007, methodology): “Network researchers have proposed quite a few solutions to this problem, most of which are new TCP congestion control algorithms [8, 23, 25, 37, 40, 53, 55, 67] and application-level This paper is partly based upon five conference papers published on the proceedings of PFLDNet workshop 2003 and 2004”
PCC Vivace: Online-Learning Congestion Control (2018, background): “Traditional algorithms [6, 15, 23] fail to satisfy the first two requirements; their performance can be as high as 10× away from the optimal under non-congestion packet loss [11].”