SlideShare uma empresa Scribd logo
1 de 1
1. Introduction
REU 2006-Packet Loss Distributions of TCP using
Web100
Zoriel M. Salado, Mentors: Dr. Miguel A. Labrador and Cesar D. Guerrero
2. Methodology
3. Results
4. Conclusions
5. Acknowledgments
Department of Computer Science &
I want to thank to César D. Guerrero and Dr. Miguel A.
Labrador for their orientations and the National Science
Foundation for supporting this project.
TCP congestion control mechanism regulates the rate at
which packets are transmitted from sender to receiver.
When a packet loss occurs, TCP reduces the
transmission rate to avoid congestion. This reduction
does not consider the available bandwidth in the
connection, and so, it does not maximize bandwidth
utilization.
The motivation for this work is to provide a mechanism
to evaluate how often a bandwidth estimation has to be
performed for TCP to improve its congestion control
mechanism.
The objective of this research is to find the time
distribution between packet loss events in TCP
connections. 
A TCP connection was establish and congested to obtain
time intervals between congestion events.
2.1. Tools
We set a Testbed composed with two end hosts
communicated by a router machine based on
Dummynet. Dummynet is a packet shaper that allows to
set different bandwidths, delays, packet loss rates, and
more.
Congestion events were induced by variation in the packet loss rate and by generating cross traffic
using MGEN. To read from the TCP connection the times when a congestion event occurs, we use
Web100. It is a Linux kernel patch that allows to monitor TCP variables.
Specifically, the readvar command gets the
current value of a Web100/TCP variable from
one connection using the connection id,
obtained from a graphical user tool called Gutil
shown on the right.
Using a python script, we timestamp every
congestion event read with readvar and used a
Matlab program to plot the results.
2.2. Testing scenarios
To induce congestion events into the network, we applied Poisson
Cross Traffic to a 75% rate of the pipe’s total capacity. A FTP
connection was established from server-client and a 3GB file
transfer was sent. From that connection, time between congestion
events was measured.
Bandwidth
Cross
Traffic
512 Kbps (ADSL) 377 Kbps
1.5 Mbps (DS1/T1) 1.1 Mbps
50 Mbps (OC-1) 37.5 Mbps
The Cumulative Distribution
Functions (CDF) for the studied
scenarios show that for less than two
second congestion intervals:
•For an ADSL connection, 83% of the
congestion events occurs is less than
1s.
•For a DS1/T1 connection, the same
83% of the congestion events occurs
in less than 0.5 seconds. Bandwidth: 512 Kbps Bandwidth: 1.5 Mbps
Bandwidth: 50 Mbps
•For a OC-1 connection,
we see that 80% and
more of the congestion
events occur in 0.25
seconds. In addition, in
a lower rate of packet
loss, congestion events
are more often than in
higher rates of packet
loss.
•The lower the bandwidth, the less number of
congestion events and the longer the time between
them. A low bandwidth increases the RTT (Round
Trip Time) and reduces the frequency in which TCP
reacts to congestion events.
•In a high bandwidth link the number of congestion
events decreases when the packet loss rate increases.
In this case, the RTT is affected not by the
transmission speed but by the packet loss rate.
•For high loaded links (75% cross traffic and significant
packet loss rates), a bandwidth estimation mechanism
to provide channel information to TCP, has to be
performed in less than 1 second. That is the time
between the majority of the congestion events.
PLR 0.01 0.05 0.10
Cong.
Events
401 608 815
PLR 0.01 0.05 0.10
Cong.
Events
1048 1596 1778
PLR 0.01 0.05 0.10
Cong.
Events
7976 4216 1488

Mais conteúdo relacionado

Mais procurados

International Journal of Engineering Research and Development (IJERD)
International Journal of Engineering Research and Development (IJERD)International Journal of Engineering Research and Development (IJERD)
International Journal of Engineering Research and Development (IJERD)IJERD Editor
 
Congestion Control in Wireless Sensor Networks Using Genetic Algorithm
Congestion Control in Wireless Sensor Networks Using Genetic AlgorithmCongestion Control in Wireless Sensor Networks Using Genetic Algorithm
Congestion Control in Wireless Sensor Networks Using Genetic AlgorithmEditor IJCATR
 
Polynomial time algorithms for multirate anypath
Polynomial time algorithms for multirate anypathPolynomial time algorithms for multirate anypath
Polynomial time algorithms for multirate anypathIMPULSE_TECHNOLOGY
 
Network Application Performance
Network Application PerformanceNetwork Application Performance
Network Application PerformanceShumon Huque
 
DPRoPHET in Delay Tolerant Network
DPRoPHET in Delay Tolerant NetworkDPRoPHET in Delay Tolerant Network
DPRoPHET in Delay Tolerant NetworkPhearin Sok
 
Paper id 35201569
Paper id 35201569Paper id 35201569
Paper id 35201569IJRAT
 
Congestionin Data Networks
Congestionin Data NetworksCongestionin Data Networks
Congestionin Data Networkslily888
 
MANET Experiment - I (Using Network Simulator NetSim -www.tetcos.com)
MANET Experiment - I (Using Network Simulator NetSim -www.tetcos.com)MANET Experiment - I (Using Network Simulator NetSim -www.tetcos.com)
MANET Experiment - I (Using Network Simulator NetSim -www.tetcos.com)Amulya Naik
 
IRJET-A Survey on congestion control with TCP network
IRJET-A Survey on congestion control with TCP networkIRJET-A Survey on congestion control with TCP network
IRJET-A Survey on congestion control with TCP networkIRJET Journal
 
Mncs 16-08-3주-변승규-opportunistic flooding in low-duty-cycle wireless sensor ne...
Mncs 16-08-3주-변승규-opportunistic flooding in low-duty-cycle wireless sensor ne...Mncs 16-08-3주-변승규-opportunistic flooding in low-duty-cycle wireless sensor ne...
Mncs 16-08-3주-변승규-opportunistic flooding in low-duty-cycle wireless sensor ne...Seung-gyu Byeon
 
Analysis of Rate Based Congestion Control Algorithms in Wireless Technologies
Analysis of Rate Based Congestion Control Algorithms in Wireless TechnologiesAnalysis of Rate Based Congestion Control Algorithms in Wireless Technologies
Analysis of Rate Based Congestion Control Algorithms in Wireless TechnologiesIOSR Journals
 
Project
Project Project
Project J M
 
Network Traffic Anomaly Detection Through Bayes Net
Network Traffic Anomaly Detection Through Bayes NetNetwork Traffic Anomaly Detection Through Bayes Net
Network Traffic Anomaly Detection Through Bayes NetGyan Prakash
 
Adaptive congestion control protocol (accp) for wireless sensor networks
Adaptive congestion control protocol (accp) for wireless sensor networksAdaptive congestion control protocol (accp) for wireless sensor networks
Adaptive congestion control protocol (accp) for wireless sensor networksijwmn
 

Mais procurados (20)

International Journal of Engineering Research and Development (IJERD)
International Journal of Engineering Research and Development (IJERD)International Journal of Engineering Research and Development (IJERD)
International Journal of Engineering Research and Development (IJERD)
 
network devices, types of delay
network devices, types of delaynetwork devices, types of delay
network devices, types of delay
 
Congestion Control in Wireless Sensor Networks Using Genetic Algorithm
Congestion Control in Wireless Sensor Networks Using Genetic AlgorithmCongestion Control in Wireless Sensor Networks Using Genetic Algorithm
Congestion Control in Wireless Sensor Networks Using Genetic Algorithm
 
i2ct_submission_105
i2ct_submission_105i2ct_submission_105
i2ct_submission_105
 
Polynomial time algorithms for multirate anypath
Polynomial time algorithms for multirate anypathPolynomial time algorithms for multirate anypath
Polynomial time algorithms for multirate anypath
 
Network Application Performance
Network Application PerformanceNetwork Application Performance
Network Application Performance
 
DPRoPHET in Delay Tolerant Network
DPRoPHET in Delay Tolerant NetworkDPRoPHET in Delay Tolerant Network
DPRoPHET in Delay Tolerant Network
 
Paper id 35201569
Paper id 35201569Paper id 35201569
Paper id 35201569
 
Ch3 5 v1
Ch3 5 v1Ch3 5 v1
Ch3 5 v1
 
Congestionin Data Networks
Congestionin Data NetworksCongestionin Data Networks
Congestionin Data Networks
 
MANET Experiment - I (Using Network Simulator NetSim -www.tetcos.com)
MANET Experiment - I (Using Network Simulator NetSim -www.tetcos.com)MANET Experiment - I (Using Network Simulator NetSim -www.tetcos.com)
MANET Experiment - I (Using Network Simulator NetSim -www.tetcos.com)
 
IRJET-A Survey on congestion control with TCP network
IRJET-A Survey on congestion control with TCP networkIRJET-A Survey on congestion control with TCP network
IRJET-A Survey on congestion control with TCP network
 
Dos unit 2
Dos unit 2Dos unit 2
Dos unit 2
 
Delay , Loss & Throughput
Delay , Loss & ThroughputDelay , Loss & Throughput
Delay , Loss & Throughput
 
Mncs 16-08-3주-변승규-opportunistic flooding in low-duty-cycle wireless sensor ne...
Mncs 16-08-3주-변승규-opportunistic flooding in low-duty-cycle wireless sensor ne...Mncs 16-08-3주-변승규-opportunistic flooding in low-duty-cycle wireless sensor ne...
Mncs 16-08-3주-변승규-opportunistic flooding in low-duty-cycle wireless sensor ne...
 
Analysis of Rate Based Congestion Control Algorithms in Wireless Technologies
Analysis of Rate Based Congestion Control Algorithms in Wireless TechnologiesAnalysis of Rate Based Congestion Control Algorithms in Wireless Technologies
Analysis of Rate Based Congestion Control Algorithms in Wireless Technologies
 
Ch1 delays, loss, and throughput l5
Ch1 delays, loss, and throughput l5Ch1 delays, loss, and throughput l5
Ch1 delays, loss, and throughput l5
 
Project
Project Project
Project
 
Network Traffic Anomaly Detection Through Bayes Net
Network Traffic Anomaly Detection Through Bayes NetNetwork Traffic Anomaly Detection Through Bayes Net
Network Traffic Anomaly Detection Through Bayes Net
 
Adaptive congestion control protocol (accp) for wireless sensor networks
Adaptive congestion control protocol (accp) for wireless sensor networksAdaptive congestion control protocol (accp) for wireless sensor networks
Adaptive congestion control protocol (accp) for wireless sensor networks
 

Destaque

Estimating Packet Loss Rate in the Access Through Application-Level Measurements
Estimating Packet Loss Rate in the Access Through Application-Level MeasurementsEstimating Packet Loss Rate in the Access Through Application-Level Measurements
Estimating Packet Loss Rate in the Access Through Application-Level MeasurementsSimone Basso
 
Cqi ecno rscp rssi concepts
Cqi ecno rscp rssi conceptsCqi ecno rscp rssi concepts
Cqi ecno rscp rssi conceptsMuhammad Ahsan
 
Interprocess communication
Interprocess communicationInterprocess communication
Interprocess communicationSushil Singh
 
WCDMA optimization & Drive test analysis
WCDMA optimization & Drive test analysisWCDMA optimization & Drive test analysis
WCDMA optimization & Drive test analysisTABREZ KHAN
 
RSCP RSSI EC/NO CQI
RSCP RSSI EC/NO CQIRSCP RSSI EC/NO CQI
RSCP RSSI EC/NO CQIFaraz Husain
 

Destaque (6)

Estimating Packet Loss Rate in the Access Through Application-Level Measurements
Estimating Packet Loss Rate in the Access Through Application-Level MeasurementsEstimating Packet Loss Rate in the Access Through Application-Level Measurements
Estimating Packet Loss Rate in the Access Through Application-Level Measurements
 
Cqi ecno rscp rssi concepts
Cqi ecno rscp rssi conceptsCqi ecno rscp rssi concepts
Cqi ecno rscp rssi concepts
 
Cqi basics
Cqi basicsCqi basics
Cqi basics
 
Interprocess communication
Interprocess communicationInterprocess communication
Interprocess communication
 
WCDMA optimization & Drive test analysis
WCDMA optimization & Drive test analysisWCDMA optimization & Drive test analysis
WCDMA optimization & Drive test analysis
 
RSCP RSSI EC/NO CQI
RSCP RSSI EC/NO CQIRSCP RSSI EC/NO CQI
RSCP RSSI EC/NO CQI
 

Semelhante a Packet Loss Distributions of TCP using Web100

A Packet Drop Guesser Module for Congestion Control Protocols for High speed ...
A Packet Drop Guesser Module for Congestion Control Protocols for High speed ...A Packet Drop Guesser Module for Congestion Control Protocols for High speed ...
A Packet Drop Guesser Module for Congestion Control Protocols for High speed ...ijcseit
 
A packet drop guesser module for congestion Control protocols for high speed ...
A packet drop guesser module for congestion Control protocols for high speed ...A packet drop guesser module for congestion Control protocols for high speed ...
A packet drop guesser module for congestion Control protocols for high speed ...ijcseit
 
File_Transfer_Protocol_Design
File_Transfer_Protocol_DesignFile_Transfer_Protocol_Design
File_Transfer_Protocol_DesignVishal Vasudev
 
Comparative Analysis of Different TCP Variants in Mobile Ad-Hoc Network
Comparative Analysis of Different TCP Variants in Mobile Ad-Hoc Network Comparative Analysis of Different TCP Variants in Mobile Ad-Hoc Network
Comparative Analysis of Different TCP Variants in Mobile Ad-Hoc Network partha pratim deb
 
MANET, Unicast Routing Protocol, Multicast Routing Protocol.
MANET, Unicast Routing Protocol, Multicast Routing Protocol.MANET, Unicast Routing Protocol, Multicast Routing Protocol.
MANET, Unicast Routing Protocol, Multicast Routing Protocol.Editor IJMTER
 
Improving Performance of TCP in Wireless Environment using TCP-P
Improving Performance of TCP in Wireless Environment using TCP-PImproving Performance of TCP in Wireless Environment using TCP-P
Improving Performance of TCP in Wireless Environment using TCP-PIDES Editor
 
Iaetsd an effective approach to eliminate tcp incast
Iaetsd an effective approach to eliminate tcp incastIaetsd an effective approach to eliminate tcp incast
Iaetsd an effective approach to eliminate tcp incastIaetsd Iaetsd
 
Congestion control algorithms.pptx
Congestion control algorithms.pptxCongestion control algorithms.pptx
Congestion control algorithms.pptxujjwalmatoliya
 
A THROUGHPUT ANALYSIS OF TCP IN ADHOC NETWORKS
A THROUGHPUT ANALYSIS OF TCP IN ADHOC NETWORKSA THROUGHPUT ANALYSIS OF TCP IN ADHOC NETWORKS
A THROUGHPUT ANALYSIS OF TCP IN ADHOC NETWORKScsandit
 
A throughput analysis of tcp in adhoc networks
A throughput analysis of tcp in adhoc networksA throughput analysis of tcp in adhoc networks
A throughput analysis of tcp in adhoc networkscsandit
 
Efficient and Fair Bandwidth Allocation AQM Scheme for Wireless Networks
Efficient and Fair Bandwidth Allocation AQM Scheme for Wireless NetworksEfficient and Fair Bandwidth Allocation AQM Scheme for Wireless Networks
Efficient and Fair Bandwidth Allocation AQM Scheme for Wireless NetworksCSCJournals
 
PERFORMANCE EVALUATION OF SELECTED E2E TCP CONGESTION CONTROL MECHANISM OVER ...
PERFORMANCE EVALUATION OF SELECTED E2E TCP CONGESTION CONTROL MECHANISM OVER ...PERFORMANCE EVALUATION OF SELECTED E2E TCP CONGESTION CONTROL MECHANISM OVER ...
PERFORMANCE EVALUATION OF SELECTED E2E TCP CONGESTION CONTROL MECHANISM OVER ...ijwmn
 
Predictable Packet Lossand Proportional Buffer Scaling Mechanism
Predictable Packet Lossand Proportional Buffer Scaling MechanismPredictable Packet Lossand Proportional Buffer Scaling Mechanism
Predictable Packet Lossand Proportional Buffer Scaling MechanismIDES Editor
 
Icfcc conference real
Icfcc conference realIcfcc conference real
Icfcc conference realUM
 
TCP INCAST AVOIDANCE BASED ON CONNECTION SERIALIZATION IN DATA CENTER NETWORKS
TCP INCAST AVOIDANCE BASED ON CONNECTION SERIALIZATION IN DATA CENTER NETWORKSTCP INCAST AVOIDANCE BASED ON CONNECTION SERIALIZATION IN DATA CENTER NETWORKS
TCP INCAST AVOIDANCE BASED ON CONNECTION SERIALIZATION IN DATA CENTER NETWORKSIJCNCJournal
 
A Review on Congestion Control Approaches for Real-Time Streaming Application...
A Review on Congestion Control Approaches for Real-Time Streaming Application...A Review on Congestion Control Approaches for Real-Time Streaming Application...
A Review on Congestion Control Approaches for Real-Time Streaming Application...IJCSIS Research Publications
 
Token Based Packet Loss Control Mechanism for Networks
Token Based Packet Loss Control Mechanism for NetworksToken Based Packet Loss Control Mechanism for Networks
Token Based Packet Loss Control Mechanism for NetworksIJMER
 

Semelhante a Packet Loss Distributions of TCP using Web100 (20)

A Packet Drop Guesser Module for Congestion Control Protocols for High speed ...
A Packet Drop Guesser Module for Congestion Control Protocols for High speed ...A Packet Drop Guesser Module for Congestion Control Protocols for High speed ...
A Packet Drop Guesser Module for Congestion Control Protocols for High speed ...
 
A packet drop guesser module for congestion Control protocols for high speed ...
A packet drop guesser module for congestion Control protocols for high speed ...A packet drop guesser module for congestion Control protocols for high speed ...
A packet drop guesser module for congestion Control protocols for high speed ...
 
File_Transfer_Protocol_Design
File_Transfer_Protocol_DesignFile_Transfer_Protocol_Design
File_Transfer_Protocol_Design
 
Ba25315321
Ba25315321Ba25315321
Ba25315321
 
Comparative Analysis of Different TCP Variants in Mobile Ad-Hoc Network
Comparative Analysis of Different TCP Variants in Mobile Ad-Hoc Network Comparative Analysis of Different TCP Variants in Mobile Ad-Hoc Network
Comparative Analysis of Different TCP Variants in Mobile Ad-Hoc Network
 
MANET, Unicast Routing Protocol, Multicast Routing Protocol.
MANET, Unicast Routing Protocol, Multicast Routing Protocol.MANET, Unicast Routing Protocol, Multicast Routing Protocol.
MANET, Unicast Routing Protocol, Multicast Routing Protocol.
 
Improving Performance of TCP in Wireless Environment using TCP-P
Improving Performance of TCP in Wireless Environment using TCP-PImproving Performance of TCP in Wireless Environment using TCP-P
Improving Performance of TCP in Wireless Environment using TCP-P
 
Congestion control mechanism using network border protocol
Congestion control mechanism using network border protocolCongestion control mechanism using network border protocol
Congestion control mechanism using network border protocol
 
Iaetsd an effective approach to eliminate tcp incast
Iaetsd an effective approach to eliminate tcp incastIaetsd an effective approach to eliminate tcp incast
Iaetsd an effective approach to eliminate tcp incast
 
Congestion control algorithms.pptx
Congestion control algorithms.pptxCongestion control algorithms.pptx
Congestion control algorithms.pptx
 
A THROUGHPUT ANALYSIS OF TCP IN ADHOC NETWORKS
A THROUGHPUT ANALYSIS OF TCP IN ADHOC NETWORKSA THROUGHPUT ANALYSIS OF TCP IN ADHOC NETWORKS
A THROUGHPUT ANALYSIS OF TCP IN ADHOC NETWORKS
 
A throughput analysis of tcp in adhoc networks
A throughput analysis of tcp in adhoc networksA throughput analysis of tcp in adhoc networks
A throughput analysis of tcp in adhoc networks
 
Efficient and Fair Bandwidth Allocation AQM Scheme for Wireless Networks
Efficient and Fair Bandwidth Allocation AQM Scheme for Wireless NetworksEfficient and Fair Bandwidth Allocation AQM Scheme for Wireless Networks
Efficient and Fair Bandwidth Allocation AQM Scheme for Wireless Networks
 
PERFORMANCE EVALUATION OF SELECTED E2E TCP CONGESTION CONTROL MECHANISM OVER ...
PERFORMANCE EVALUATION OF SELECTED E2E TCP CONGESTION CONTROL MECHANISM OVER ...PERFORMANCE EVALUATION OF SELECTED E2E TCP CONGESTION CONTROL MECHANISM OVER ...
PERFORMANCE EVALUATION OF SELECTED E2E TCP CONGESTION CONTROL MECHANISM OVER ...
 
Bg4101335337
Bg4101335337Bg4101335337
Bg4101335337
 
Predictable Packet Lossand Proportional Buffer Scaling Mechanism
Predictable Packet Lossand Proportional Buffer Scaling MechanismPredictable Packet Lossand Proportional Buffer Scaling Mechanism
Predictable Packet Lossand Proportional Buffer Scaling Mechanism
 
Icfcc conference real
Icfcc conference realIcfcc conference real
Icfcc conference real
 
TCP INCAST AVOIDANCE BASED ON CONNECTION SERIALIZATION IN DATA CENTER NETWORKS
TCP INCAST AVOIDANCE BASED ON CONNECTION SERIALIZATION IN DATA CENTER NETWORKSTCP INCAST AVOIDANCE BASED ON CONNECTION SERIALIZATION IN DATA CENTER NETWORKS
TCP INCAST AVOIDANCE BASED ON CONNECTION SERIALIZATION IN DATA CENTER NETWORKS
 
A Review on Congestion Control Approaches for Real-Time Streaming Application...
A Review on Congestion Control Approaches for Real-Time Streaming Application...A Review on Congestion Control Approaches for Real-Time Streaming Application...
A Review on Congestion Control Approaches for Real-Time Streaming Application...
 
Token Based Packet Loss Control Mechanism for Networks
Token Based Packet Loss Control Mechanism for NetworksToken Based Packet Loss Control Mechanism for Networks
Token Based Packet Loss Control Mechanism for Networks
 

Packet Loss Distributions of TCP using Web100

  • 1. 1. Introduction REU 2006-Packet Loss Distributions of TCP using Web100 Zoriel M. Salado, Mentors: Dr. Miguel A. Labrador and Cesar D. Guerrero 2. Methodology 3. Results 4. Conclusions 5. Acknowledgments Department of Computer Science & I want to thank to César D. Guerrero and Dr. Miguel A. Labrador for their orientations and the National Science Foundation for supporting this project. TCP congestion control mechanism regulates the rate at which packets are transmitted from sender to receiver. When a packet loss occurs, TCP reduces the transmission rate to avoid congestion. This reduction does not consider the available bandwidth in the connection, and so, it does not maximize bandwidth utilization. The motivation for this work is to provide a mechanism to evaluate how often a bandwidth estimation has to be performed for TCP to improve its congestion control mechanism. The objective of this research is to find the time distribution between packet loss events in TCP connections.  A TCP connection was establish and congested to obtain time intervals between congestion events. 2.1. Tools We set a Testbed composed with two end hosts communicated by a router machine based on Dummynet. Dummynet is a packet shaper that allows to set different bandwidths, delays, packet loss rates, and more. Congestion events were induced by variation in the packet loss rate and by generating cross traffic using MGEN. To read from the TCP connection the times when a congestion event occurs, we use Web100. It is a Linux kernel patch that allows to monitor TCP variables. Specifically, the readvar command gets the current value of a Web100/TCP variable from one connection using the connection id, obtained from a graphical user tool called Gutil shown on the right. Using a python script, we timestamp every congestion event read with readvar and used a Matlab program to plot the results. 2.2. Testing scenarios To induce congestion events into the network, we applied Poisson Cross Traffic to a 75% rate of the pipe’s total capacity. A FTP connection was established from server-client and a 3GB file transfer was sent. From that connection, time between congestion events was measured. Bandwidth Cross Traffic 512 Kbps (ADSL) 377 Kbps 1.5 Mbps (DS1/T1) 1.1 Mbps 50 Mbps (OC-1) 37.5 Mbps The Cumulative Distribution Functions (CDF) for the studied scenarios show that for less than two second congestion intervals: •For an ADSL connection, 83% of the congestion events occurs is less than 1s. •For a DS1/T1 connection, the same 83% of the congestion events occurs in less than 0.5 seconds. Bandwidth: 512 Kbps Bandwidth: 1.5 Mbps Bandwidth: 50 Mbps •For a OC-1 connection, we see that 80% and more of the congestion events occur in 0.25 seconds. In addition, in a lower rate of packet loss, congestion events are more often than in higher rates of packet loss. •The lower the bandwidth, the less number of congestion events and the longer the time between them. A low bandwidth increases the RTT (Round Trip Time) and reduces the frequency in which TCP reacts to congestion events. •In a high bandwidth link the number of congestion events decreases when the packet loss rate increases. In this case, the RTT is affected not by the transmission speed but by the packet loss rate. •For high loaded links (75% cross traffic and significant packet loss rates), a bandwidth estimation mechanism to provide channel information to TCP, has to be performed in less than 1 second. That is the time between the majority of the congestion events. PLR 0.01 0.05 0.10 Cong. Events 401 608 815 PLR 0.01 0.05 0.10 Cong. Events 1048 1596 1778 PLR 0.01 0.05 0.10 Cong. Events 7976 4216 1488