TCP Congestion Control Simulation

This project simulates TCP Congestion Control mechanisms, illustrating how the Congestion Window (cwnd) changes over time in response to network events such as timeouts, fast retransmissions, slow start, and congestion avoidance.

Quickstart Guide

Running the Simulation

1. Copy all the contents from this repository.
2. Open a terminal and navigate to the folder containing congestion.py.
3. Run the program using:

   python congestion.py

4. The simulation will:
   • Print log entries showing cwnd changes due to different congestion control mechanisms.
   • Plot a graph showing how cwnd evolves over Round Trip Times (RTTs).

Core Concepts

• Slow Start: Doubles cwnd until it reaches ssthresh.
• Congestion Avoidance: Increases cwnd linearly after ssthresh is reached.
• Timeouts: Drops cwnd to 1 and updates ssthresh.
• Fast Retransmit: Quickly reduces cwnd but avoids a full restart.

Preview of Simulation Output

Log Output Example

Slow start at RTT 0: cwnd=2, ssthresh=16
Slow start at RTT 1: cwnd=4, ssthresh=16
Congestion avoidance at RTT 5: cwnd=17, ssthresh=16
Timeout at RTT 10: cwnd=1, ssthresh=8
...

Graph Example

The simulation generates a graph plotting cwnd over RTTs, illustrating how congestion control operates dynamically.

Notes:

• The probability of timeouts and fast retransmits is configurable within the script.
• The initial ssthresh value can be modified to observe different behaviors.
• The model follows the fundamental principles of TCP Tahoe & Reno congestion control.

Future Enhancements

• Implement Selective Acknowledgments (SACK) for a more advanced simulation.
• Add real-world network trace integration.
• Extend support for TCP Cubic.