This repository contains the simulation framework for the master's thesis on adaptive and fault-tolerant leader election in resource-constrained IoT and robotic clusters. It implements and evaluates multiple leader election algorithms on top of Contiki-NG using the Cooja network simulator.

Full source code: https://github.com/Pushpit07/PraSLE-Simulation-in-Contiki-NG-Cooja

The framework implements and compares four leader election algorithms for resource-constrained IoT networks:

PraSLE and Adaptive-PraSLE additionally support multiple network topologies: clique (fully connected), ring, line, and mesh.

examples/leader-election/
  algorithms/         # Algorithm implementations (bully, ring, prasle, adaptive-prasle)
  common/             # Shared metrics collection code
  experiments/        # Experiment runners and CSC simulation templates
  scripts/            # Python automation and analysis tools
  runners/            # Shell wrapper scripts for each algorithm
  results/            # Output data and comparison charts
  build/              # Compiled binaries

Detailed documentation is available inside the framework:

• examples/leader-election/README.md -- full framework guide
• examples/leader-election/EXPERIMENTS.md -- detailed experiment guide
• examples/leader-election/COMPARISON.md -- algorithm comparison and parameter details

• Contiki-NG development environment (this repository)
• Cooja simulator (included with Contiki-NG)
• Java (required by Cooja)
• Python 3 with matplotlib and numpy (for analysis scripts)
• Bash shell

Refer to the Contiki-NG getting started guide for setting up the development environment and Cooja.

cd examples/leader-election

# Interactive mode -- presents a menu to select the algorithm
./build.sh

# Or specify the algorithm directly
./build.sh bully
./build.sh ring
./build.sh prasle
./build.sh adaptive-prasle
./build.sh all            # Build all algorithms

Build options (passed via environment or Makefile variables):

# Collect metrics for a single algorithm (60-second default)
./create-metrics.sh bully

# Specify duration and output name
./create-metrics.sh ring 120 my_test

The master script automates everything -- building, running experiments across all algorithms and node counts, and generating charts:

cd examples/leader-election

# Run the complete evaluation (all algorithms, all experiments, all node counts)
./experiments/run_complete_evaluation.sh

# Customize the run
./experiments/run_complete_evaluation.sh \
  --trials 10 \
  --algorithms "bully ring prasle adaptive-prasle" \
  --nodes "5 10 50 100" \
  --experiments "convergence fault_tolerance noise network_partition" \
  --parallel

# Convergence (cold-start election time)
./experiments/convergence/run_convergence_trials.sh

# Fault tolerance (leader crash and recovery)
./experiments/fault_tolerance/run_crash_trials.sh

# Network noise (packet loss resilience at 50%, 70%, 90% success rates)
./experiments/noise/run_noise_trials.sh

# Network partition (split-brain scenarios)
./experiments/network_partition/run_partition_trials.sh

# Quick analysis
./experiments/analyze.sh bully

# Generate comparison charts (Python)
python3 scripts/plot_node_comparison_charts.py
python3 scripts/plot_recovery_comparison_charts.py
python3 scripts/plot_noise_comparison_charts.py
python3 scripts/plot_partition_comparison_charts.py
python3 scripts/compare_algorithms.py

Results are organized under examples/leader-election/results/:

results/case1/                          # Parameter set 1 (original paper)
  bully/convergence/5nodes_<timestamp>/
    convergence_times.csv               # Summary metrics
    trial_1/
      cooja_output.log                  # Raw simulation log
      metrics.csv                       # Parsed metrics
    trial_2/
    ...

Experiments are run across node counts of 5, 10, 50, and 100, with multiple trials per configuration for statistical significance.

This project is built on Contiki-NG, an open-source OS for next-generation IoT devices, which provides IPv6/6LoWPAN, 6TiSCH, RPL, and CoAP networking stacks.

• Contiki-NG documentation: https://docs.contiki-ng.org/
• Contiki-NG releases: https://github.com/contiki-ng/contiki-ng/releases

Unless explicitly stated otherwise, Contiki-NG sources are distributed under the terms of the 3-clause BSD license.