🌐 Congestion Detection System

---

📘 Overview

This project simulates a small congestion detection pipeline. It continuously checks the network path to a target machine and measures:

• ICMP RTT to estimate basic reachability and latency
• UDP RTT to check delay under UDP probing
• TCP RTT to check response delay over TCP
• Throughput using iperf3 to measure bandwidth

The idea is simple:

• If delay is low and throughput is healthy, the network is likely normal
• If delay increases and throughput drops, the network may be congested

This makes the project a good example of real-time network performance monitoring.

---

✨ Features

• ICMP latency measurement
• UDP latency measurement
• TCP latency measurement
• Throughput measurement with iperf3
• Congestion state detection
• Live console output
• Graph output for results
• Works in both:
  • single-laptop testing
  • two-laptop demonstration

---

📌 What this project does

The project repeatedly:

1. Pings a target laptop and collects RTT samples.
2. Tries UDP and TCP probe measurements as fallback.
3. Smooths RTT values.
4. Extracts features such as mean RTT, standard deviation, and elevation.
5. Predicts congestion using a trained SVM model.
6. Measures throughput using iPerf3.
7. Saves everything into network_log.csv.
8. Optionally plots the logs.

---

⚙️ Requirements

Install these before running the project:

• Python 3.10+ on Windows
• pip
• matplotlib
• numpy
• pandas
• scikit-learn
• iPerf3 (iperf3.exe)

Install the Python packages with:

pip install matplotlib numpy pandas scikit-learn

---

🚀 TWO MODES OF RUNNING THE PROJECT

iPerf3 setup

Download iPerf3 for Windows, extract it, and keep iperf3.exe somewhere easy to find.

You will need the full path to iperf3.exe in throughput.py.

Example:

r"C:\Users\YourName\Desktop\iperf-3.21-win64\iperf3.exe"

---

1. Configuration: single-laptop mode

Use this mode when the client and server are on the same laptop.

main.py

Set targets to your laptop’s Wi-Fi IP address:

targets = ["<YOUR_LAPTOP_IP>"]

Example:

targets = ["192.168.0.3"]

throughput.py

Set the iPerf target to localhost:

def measure_throughput(server="127.0.0.1"):

And inside the subprocess command, keep:

"-c", "127.0.0.1",

Why this works

• 192.168.0.3 (example) is your real network IP for ping/RTT checks.
• 127.0.0.1 is localhost, so iPerf client connects to iPerf server on the same laptop.

---

2. Configuration: two-laptop demo

Use this mode when you want to demonstrate the project across two laptops.

Laptop A

This is the monitoring laptop. It runs main.py.

Laptop B

This is the server laptop. It runs:

iperf3.exe -s

main.py on Laptop A

Set targets to Laptop B’s IP address:

targets = ["<LAPTOP_B_IP>"]

Example:

targets = ["192.168.0.10"]

throughput.py on Laptop A

Set the iPerf server value to Laptop B’s IP:

def measure_throughput(server="192.168.0.10"):

And inside the subprocess command, use:

"-c", "192.168.0.10",

Laptop B terminal command

Run the server on Laptop B:

iperf3.exe -s

If Windows Firewall asks for permission, allow access.

---

⌨️ How to run the project

1) Extract the ZIP file

Unzip the project folder.

2) Open the folder in VS Code

Open the extracted folder in Visual Studio Code.

3) Open a terminal

Use VS Code Terminal or CMD/PowerShell.

4) Install dependencies

Run:

pip install matplotlib numpy pandas scikit-learn

5) Edit the IP values

Update the following files:

• main.py → targets = [...]
• throughput.py → server="..."
• throughput.py → "-c", "..."

Choose either:

• single-laptop mode: use 127.0.0.1 in throughput.py
• two-laptop mode: use Laptop B’s real IP in both files

6) Start iPerf3 server

Single-laptop mode

Open a second terminal and run:

iperf3.exe -s

or, from the folder that contains iperf3.exe:

.\iperf3.exe -s

Two-laptop mode

Run the same command on Laptop B.

7) Run the main project

In the project terminal, run:

python main.py

8) Optional: plot the log

After the run creates network_log.csv, plot it with:

python plot_graph.py

---

Commands a user will type

Install packages

pip install matplotlib numpy pandas scikit-learn

Start iPerf server

.\iperf3.exe -s

Run project

python main.py

Plot results

python plot_graph.py

---

Expected output

During execution, you should see lines like:

• ICMP RTT: ...
• UDP RTT: ...
• TCP RTT: ...
• Throughput: ... Mbps
• State: NORMAL or State: CONGESTED
• 

---

Project Structure

project-folder/
│
├── main.py
├── throughput.py
├── congestion.py
├── utils.py
├── requirements.txt
├── 6th.png
├── image61.png
└── README.md