NS-2 Network Simulation Project

This project involves the simulation of various network scenarios using the NS-2 simulator. The main objective is to examine how changing network parameters (e.g., packet size, interval, queue type, bandwidth) affects key performance metrics such as delay, throughput, and packet loss.

The project was collaboratively developed by @berkinyl, @silanazaslan, and @erenyurtcu as part of the Simulation and Modelling course at Manisa Celal Bayar University.

---

🖥️ Environment Setup

⚠ NS-2 is a legacy simulator developed in the early 2000s, which makes installation and compatibility a challenge on modern systems.

In this project, the team utilized a Linux Mint environment.

---

📁 Project Structure

├── case1/       # Interval and Bandwidth analysis (TCP + UDP)
├── case2/       # Packet size effect (UDP)
├── case3/       # Queue type comparison (DropTail vs RED)
├── case4/       # Bottleneck link (output bandwidth) variation
├── analyze_trace.py   # General trace analyzer
└── venv/        # Python virtual environment

Each case directory contains:

• .ns → NS-2 scenario (topology and traffic)
• .trace → Simulation logs (packet-level events)
• .nam → GUI animation file
• caseX_analyzer.py → Custom Python analyzer script

---

🔬 Metrics Analyzed

Using Python scripts, the following metrics were extracted:

• 📦 Sent / Received / Dropped packets
• 📉 Drop Rate (%)
• 🕒 Average Waiting Time (overall & per-link)
• 📈 Throughput (Kbps)
• ⏱️ Simulation Duration

---

📊 Case Summaries

➤ Case 1: Interval & Bandwidth

• Lower interval → higher traffic → more congestion
• Higher bandwidth → better throughput & lower drops

➤ Case 2: UDP Packet Size

• Larger packets → fewer drops, higher throughput, but higher risk on loss
• Smaller packets → increased drop due to queuing

➤ Case 3: Queue Type (DropTail vs RED)

• DropTail: Drops when queue is full (simple, high burst loss)
• RED: Random early drops to trigger TCP congestion control

🔁 Trade-off: RED reduces delay but may decrease throughput.

Case 3 Analysis Output

➤ Case 4: Bottleneck Bandwidth

• Narrow link (64Kb) → high drop & delay
• Wider link (256Kb) → more stable, higher throughput

Case 4 Analysis Output

---

🧪 Validation & Verification

✔ Verification:
Trace outputs were matched with Python script outputs.

✔ Validation:
Observed behaviors aligned with expected network theory:

• Throughput rises with bandwidth
• Queue type affects delay vs throughput balance
• Packet size alters risk vs efficiency

---

🛠 How to Run

# Run a simulation
$ ns case1_1_tcp_udp_004_1mb.ns

# Analyze output
$ python case1_analyzer.py

Make sure NS-2 is installed and sourced, and Python virtualenv is activated.

---

Example Outputs

Trace Analyzer Output

Graphical Output (Interval vs Throughput)