Source code for "Coherence-Key Deinterleaving: Exploiting the Physical Constraint of Coherent Integration to Separate Four-Parameter-Agile Radar Pulse Trains"

Complete mathematical model and source code for the manuscript "Coherence-Key Deinterleaving: Exploiting the Physical Constraint of Coherent Integration to Separate Four-Parameter-Agile Radar Pulse Trains", submitted to Radioengineering. Contents: - pdwgen.py PDW generator (scenario A and B, fixed seeds) - ckd.py coherence-key estimator and the generalised likelihood-ratio test - baselines.py implementations of CDIF, SDIF, PRI transform, k-means, DBSCAN, MLP - hwil.py semi-physical bench model (analytic, no measured data) - make_figs.py produces all five figures - run_baselines.py / final_eval.py / run_experiments.py experiment drivers - sync_results.py writes measured numbers back into the manuscript sources - matlab/ckd_demo.m MATLAB implementation of the core estimator Reproduce: cd src && python3 make_figs.py && python3 run_baselines.py && python3 final_eval.py && python3 hwil.py (about 15 min on one core) Environment: Python 3.10.12, NumPy 2.2.6, SciPy 1.15.3, Matplotlib 3.10.9, Ubuntu 22.04.5 LTS.