Transcript TCP Schemes Investigation in Wired and Wireless Hybrid Networks

TCP Schemes Investigation in
Wired and Wireless Hybrid
Networks
By Weiwei Hu and Wei Zha
Dec. 8, 2004
1
Outline







Introduction
Random Loss Model
TCP-Jersey
TCP-Westwood
Performance Analysis
Conclusion
References
2
Introduction

Transmission Control Protocol (TCP)





reliable
end-to-end
window based
designed for the wired networks (no random packet losses): not fit for
wireless network
Application




WWW - HTTP
File transfer - FTP
E-Mail - SMTP
Remote access - Telnet
3
Introduction (Cont.)

Wireless networks: bursty and high channel error
rates.



Throughput decreases greatly.
Extending TCP as used over the wired links to the
wireless links may bring inefficient usage of
bandwidth.
The TCP protocol is still used to transfer data over
the wireless link.
4
Introduction (cont.)

A lot of attention is currently being paid to the
design of a better TCP scheme over wireless links.

Difficulty in modeling the TCP protocol analytically

Many of these studies are simulation based.

Hard to obtain insight into the effects of particular
parameters on the behavior of TCP using simulations
with specific settings
5
Random Loss Model

Independent identically distributed (i.i.d )

The packet transmission time of the packets on both lossless and
lossy link are exponentially distributed.


The congestion avoidance phrase is modeled using probability.
Correlated



Hiding the losses from the transport layer and moving the problem
to link layer
Applicable on very low bandwidth wireless links
cannot be used to model the fast-transmit and fast-recovery phrases
6
TCP-Jersey

Congestion warning marking function (RED
and ECN)

Routers can take part in controlling TCP transmission rate: active
queue management (AQM)

Random early detection (RED): AQM scheme implementation

Explicit congestion notification (ECN): an extension to RED
RED and ECN are sensitive to parameter settings
Cause unsatisfactory TCP performance

7
TCP-Jersey (Cont.)



TCP-Jersey simplifies the marking function:
use only one threshold (thresh)
Marking all the packets with probability 1 if the
average queue length exceeds the predefined
threshold.
Calculating the average queue length using a
larger queue weight than the original ECN.
8
TCP-Jersey (Cont.)


Available Bandwidth Estimation
Estimate the available bandwidth (Rn) for the TCP
connection.
Ln: size of packet n;
RTT  Rn1  Ln
R

 ABE estimator:
n
(t  t )  RTT
n

n 1
t n-1: previous packet arrival time.
Calculate the optimum congestion window (ownd) size
once every RTT.
ownd n 
RTT  Rn
seg _ size
9
TCP-Westwood



A sender-side modification of the TCP congestion
window algorithm
improves upon the performance of TCP in wired and
mixed network
uses an end-to-end bandwidth measurement.



The ACK-based measurement procedure
filtering the ACK reception rate
The effects of delayed and cumulative ACKs on bandwidth
measurement
10
TCP-Westwood (Cont.)


Main idea: exploit TCP acknowledgement packets
to derive rather complicated measurements.
A source perform an end-to-end estimate of the
bandwidth available along a TCP connection by
measuring and averaging the rate of returning
ACKs.
dk
bk 
t k  t k 1
2
1


t k  t k 1
bk  bk 1
bk 
b k 1 
2
2
1
1
t k  t k 1
t k  t k 1
11
TCP-Westwood (Cont.)
Three types of filters actually used in the simulation:


(a  1)
b  a b k 1 
[bk  bk 1 ]
 Original filter:
2
current_bwe_ = current_bwe * fr_alpha_ + sample_bwe * (1 – fr_alpha_)

Filter 1:
current_bwe_ = current_bwe_ * .9047 + (sample_bwe + last_bwe_sample_) * .0476

Filter 2: (LPF)
current_bwe_ = cureent_bwe_ * .93548 + (sample_bwe+ last_bwe_sample_) * .03225
12
TCP-Westwood (Cont.)

TCP-Westwood controls the window using end-to-end rate
transmission to continuously estimate the packet rate of the
connection at the TCP sender.

It can reset the window to match available bandwidth,
which makes it more robust to the random loss in wireless
link.

TCP-Westwood has many superior features which can
handle losses caused by link errors or wireless channel
conditions more efficiently than most of previous existing
TCP schemes.
13
Performance Analysis

We chose TCP-Westwood as a starting point because of its
similarity to TCP-Jersey.

We tried to simulate TCP-Jersey in ns2. However, due to
the opaque description for detailed mechanism in TCPJersey and lack of references, the implementation seems
infeasible so far.


Problem: The congestion warning part (how to define the proper
thresh of average queue length).
Thus, we simulate TCP-Westwood and all the other
conventional TCP schemes in NS-2 network simulator.
14
Simulation Environment
2 Mb, 1 ms
20Mb, 35ms
S



BS
D
The source connects to a wired based station via a 20 Mb/s
error free link with 35 ms one-way delay.
The base station is linked to the destination, a wireless
mobile node, via a 2 Mb/s lossy channel with 1 ms delay.
A single TCP connection running a long-live FTP
application delivers data from the source to the destination.
15
Bandwidth Estimation

An effective way to calculate the congestion
window size

Control data traffic efficiently.

Expecting TCP-Westwood can reach a good
estimation for bandwidth so that the lossy link
can be utilized as much as possible.
16
Bandwidth Estimation (Cont.)
• When the simulation time is greater than 7s, the TCP Westwood reaches the
max bandwidth utilization.
17
Bandwidth Estimation (Cont.)
• The slow start threshold matches very well with congestion window settings.
18
Goodput





An explicit characteristic for TCP performance.
The effective amount of data delivered through the
network.
Run the simulation for TCP -Tahoe, -Reno, New
Reno, -Sack, and -Westwood respectively.
Random link error rate varies from 0.001% to
10%.
The simulation time is 60s.
19
Goodput Performance (Cont.)
• When the lossy rate
increases, TCP Westwood
outperforms other TCP
schemes explicitly,
especially when error rate
increases to 0.001.
Tahoe
Reno
New Reno
Sack
Westwood
1600
1400
1200
Goodput (Kbps)
• The first point at link error
rate 0.00001, all TCP
schemes are closely match
to each other, which means
that they perform the same
at nearly no loss rate
condition.
1800
1000
800
600
400
200
-5
10
-4
10
-3
10
Wireless Link Error Rate
-2
10
-1
10
20
Goodput Performance (Cont.)
5
Tahoe
Reno
New Reno
Sack
Westwood
4.5
• The buffer size is kept
as a constant, B = 60pkts.
• TCP-Westwood is a
slightly better than other
schemes.
3.5
Goodput (Mbps)
• TCP-Westwood varies
almost linearly to track
the bandwidth variations
of the bottleneck.
4
3
2.5
2
1.5
1
0.5
0
0
0.5
1
1.5
2
2.5
3
Bottleneck bandwidth (Mbps)
3.5
4
4.5
5
21
Packet Loss
3rd dup Ack triggles a
retransmission of
packet #115000:
Fast Recovery
22
Packet Loss (cont.)
The sending rate is
slowed down:
Congestion
Avoidance phase
23
Packet Loss (cont.)
Reenter the
Slow Start phase
24
Conclusion

Investigate the performance of the different TCP algorithm
in a wired and wireless hybrid networks.

Provide an analytical model for TCP-Jersey and TCPWestwood.

First, we started with TCP-Westwood to perform the
simulation.

Haven’t reproduced the TCP-Jersey so far in results of
obscure description and lack of references.

Limitation:


Congestion Warning: restrict the threshold of marking algorithm.
Cooperate ECN to mark the packets.
25
Conclusion (cont.)

Different TCP algorithms, namely, Tahoe, Reno, NewReno, Sack, and Westwood are studied in the given
simulation environment. According to the experiment
results, TCP-Westwood operates more robust than other
existing TCP schemes, which shows the benefits of the
bandwidth estimation in wireless or hybrid networks.
26
References
1)
2)
3)
4)
5)
6)
7)
Kai Xu, Ye Tian, and Nirwan Ansari, “TCP-Jersey for Wireless IP Communications”, IEEE
Journal on Selected areas in Communications, vol. 22, no.4, May 2004.
Gerla M., Sanadidi M.Y., Ren Wang Zanella, A. Casetti C., and Mascolo S., “TCP
Westwood: Bandwidth Estimation for Enhanced Transport over Wireless Links”, IEEE
Global Telecommunications Conference 2001, vol. 3, pp. 25-29, Nov. 200.
David D. Clark, and Wenjia Fang, “Explicit Allocation of Best-Effort Packet Delivery
Service”, IEEE/ACM Trans. Networking, vol. 6, pp.362-373.
Farooq Anjum and Leandros Tassiulas, “Comparative study of various TCP versions over a
wireless link with correlated losses”, IEEE/ACM Transaction on Networking, vol. 11, no. 3,
June 2003.
A. Kumar and J. Holltzmann, “Comparitive performance analysis of versions of TCP in
local network with a lossy link – Part II: Rayleigh Fading Mobile Radio Link”, Rutgers
Univ., New Brunswick, NJ, Tech. Rep. WINLAB-TR 129, Oct. 1996.
M.May, J. Bolot, C. Diot, and B. Lyles, “Reasons not to deploy FED,” Proc. Int. Workshop
Quality-of-Service (IWQoS), 1999, pp. 260-262.
A. Chochalingam, M. Zozi, and R. R. Rao, “Performance of TCP on wireless fading links
with memory,” in Proc. IEEE Int. Conf. Communications, June 1998, pp. 595-600.
27
References (Cont.)
8)
9)
10)
11)
12)
13)
14)
15)
16)
V. Jacobson, “Congestion avoidance and control,” ACM Computer Communications
Review, vol. 4, pp. 314-319, Auguest 1988.
C. Casetti, M. Gerla, S. Lee, S. Mascolo, and M. Sanadidi, “TCP with fast recovery”,
MILCOM 2000, Los Angeles, CA, Oct. 2000.
A. Zanella, G. Procissi, M. Gerla, M. Y. Snanaddi, “TCP Westwood: analytic model
and performance evaluation”, Technical Report, URL:
http://www.cs.ucla.edu/csd/phus/pubs.html.
ns-2 network simulator. LBL, URL: http://www.msh.cs.berkley.edu/ns.
TCP Westwood modules for ns-2. URL: http://www1.tcl.polito.it/casetti/tcowestwood.
K. Fall and S. Floyd, “Simulation-based comparisons of Tahoe, Reno, and SACK
TCP,” ACM Computer Communication Review, 16:5-21, 1996.
C. Casetti and M. Meo, “A new approach to model the stationary havavior of TCP
connections”, In Proc. of IEEE INFOCOME, pp. 367-375, Mar. 2000.
J. Mo, R. La, V. Anantharam, and J. Walrand, “Analysis and comparision of TCP
Reno and Vegas,” In Proc. of IEEE INFOCOM, pp. 1556-1563, Mar. 1999.
J. Padhye, V. Firoiu, D. Towsley, and J. Kurose, “Modeling TCP throughput: a simple
model and its empirical validation,” ACM Computer Communication Review,
28:303-314.
28