Replication package: loop-gain analysis of blind retransmission in IEEE 802.11 DCF and NR-V2X Mode 2 sidelink

Instrumentation, raw result files and analysis scripts supporting the article "When Does Reliability Destabilise a Shared Medium? Loop-Gain Analysis from IEEE 802.11 and NR-V2X toward 6G Sidelink". The study asks when a retransmission mechanism, introduced to raise reliability, closes a positive feedback loop on the medium that carries it, and measures the gain of that loop in NR-V2X Mode 2 sidelink. The data here are what those measurements were computed from. The package contains four parts. (i) The trace header inserted into the Mode 2 resource-selection routine, which records the candidate set where it is formed and the choice where it is committed, together with a measurement-only registry of held resources used to count invisible occupancy: resources that are occupied but not excluded because the collision destroyed the announcement protecting them. Nothing the simulated station decides reads that registry. (ii) The run harness, which records a failed run as failed rather than as a zero. (iii) Twelve result files covering the transmitting-population sweep, the reselection-counter sweep, the resource-keep sweep at seven settings with ten seeds each, the selection-window sweep, the subchannel sweep, and the seed-matched perturbation experiment in which a run with a synchronised reselection is differenced against an otherwise identical unperturbed run. (iv) The analysis and figure scripts, including the statistics behind the interior minimum in the resource-keep probability: per-setting confidence intervals, Welch tests between adjacent settings and against the pooled plateau, and a bootstrap over the seed ensemble giving the distribution of the location of the minimum. A README maps every result file to the table or figure it produces, documents the column conventions, explains how failed runs are marked, and states what the model does and does not claim. The model implements the Mode 2 sensing, exclusion and reselection procedures as specified in 3GPP TS 38.214; it has not been through conformance testing and no claim of compliance is made. The statistical results of the packet-level section can be reproduced from the result files alone, without re-running the simulator. Running the keep-sweep script on the supplied CSV reproduces the seven per-setting summaries, the significance tests and the figure.