Understanding and Improving TCP Performance Over Networks with Minimum Rate Guarantees

1
Understanding and Improving TCP Performance Over
Networks with Minimum Rate Guarantees

• Feng, Kandlur, Saha, and Shin
• Transactions on Networking
• April 1999

2
Abstract
3

• Tolerant playback applications
• Can tolerate some loss of fidelity
• Client buffering, streaming mode
• Such as PointCast, RealAudio, VDOnet,
• Quality of playback varies from excellent to
  intolerable, depending on the network load
• Intolerant applications
• Require an absolutely faithful playback and
  cannot tolerate loss of fidelity

4

• Real-time applications
• needs the data in each packet by a certain time
  and, if the data has not arrived by then, the
  data is essentially worthless
• Elastic applications
• always wait for data to arrive without time limit
• Such as FTP, TELNET

5

• Predictability in network services
• To guarantee a minimum level of throughput at all
  time
• To allow for higher throughput during periods of
  light load
• Good for
• Tolerant playback applications
• Elastic applications
• And others

6

• The objective of this paper
• To enhance network services within the framework
  of the IETFs controlled-load service ?
  architecture, but with minimum enhancements on
  the network infrastructure, especially the
  routers.?INTSERV offers two types of
  services (1) Guaranteed service (2)
  Controlled-load service
  (better-than-best-effort service)

7

• TCP Slow Start Congestion Avoidance
• When (snd_cwnd lt ssthresh) snd_cwnd snd_cwnd
  1 (SS)
• When (snd_cwnd gt ssthresh) snd_cwnd snd_cwnd
  1 / snd_cwnd (CA)

8

• TCP Slow Start Congestion Avoidance
• When (snd_cwnd lt ssthresh) snd_cwnd snd_cwnd
  1 (SS)
• When (snd_cwnd gt ssthresh) snd_cwnd snd_cwnd
  1 / snd_cwnd (CA)

9
(No Transcript)
10
(No Transcript)
11
(No Transcript)
12
(No Transcript)
13
(No Transcript)
14
(No Transcript)
15
(No Transcript)
16
(No Transcript)
17
(No Transcript)
18
(No Transcript)
19
(No Transcript)
20
(No Transcript)
21
(No Transcript)
22
(No Transcript)
23
(No Transcript)
24
(No Transcript)
25
(No Transcript)
26
(No Transcript)
27
(No Transcript)
28
(No Transcript)
29
(No Transcript)
30
(No Transcript)
31
(No Transcript)
32
(No Transcript)
33
(No Transcript)