LoRaWAN Scalability Boundary Validation Dataset — V5R Replacement Campaign

This dataset supports the revised manuscript “Scalability Boundaries in Dense EU868 LoRaWAN: A Calibrated Airtime–Recovery Model with ns-3 Validation” (Mansour Abdulaziz, 2026). It contains the frozen V5R replacement simulation campaign and the supporting provenance, analysis, and model-validation artifacts. The V5R campaign contains 2,020 independent ns-3.45 runs: 1,440 main runs: 72 scenario cells × 20 rngRun values 540 repetition-control runs: 27 cells × 20 rngRun values 40 confirmatory excess-coded RLNC runs: 2 cells × 20 rngRun values The main comparison evaluates Baseline, multi-channel reception (MC), single-channel generation-based RLNC (NC), and combined MC–RLNC. Repetition controls use R∈{2,3,4}, and the confirmatory RLNC sensitivity evaluates k=4 with n>k. Simulation environment Simulator: ns-3.45 with LoRaWAN and study-specific measurement extensions Region / MAC: EU868, LoRaWAN Class A Deployment: static 500 m × 500 m field End-device positions: independently and uniformly distributed Gateways: three fixed gateways Propagation: LogDistance model, exponent 3.76, L0=7.7 dB at 1 m Additional shadowing random variable: none Spreading factor: static link-budget assignment Runtime ADR: disabled Uplinks: unconfirmed Application payload: 50 bytes Traffic periods: 10 s, 5 s, and 2 s Device counts: 50, 100, and 200 Run duration: 1200 s Measurement warm-up: first 100 s excluded Experimental duty-cycle accounting: conservative shared 1% per-device uplink airtime budget across the evaluated channel set Operating profiles Baseline: single-channel uncoded uplink MC: uncoded multi-channel uplink using eight evaluated EU868 channels NC: generation-based GF(256) RLNC on one channel MC–RLNC: generation-based GF(256) RLNC with multi-channel reception Rep: repeated uncoded transmission control RLNC generation sizes are k∈{2,4,8}. The main RLNC campaign uses n=k. A separate confirmatory experiment evaluates excess coded transmissions n>k for k=4. Each coded frame carries a 2-byte generation identifier plus k GF(256) coefficient bytes, for a total coding metadata overhead of k+2 bytes. Measurement semantics Logical PDR is calculated from uniquely recovered original application packets. Multi-gateway reception is deduplicated using the actual uplink frame identity. Unique-PHY delivery/interference outcomes are reported separately from RLNC generation recovery. Queue-specific loss is not reported because the V5R traces do not support a validated queue-loss attribution. Duty-cycle limitation is characterized through admission / never-PHY behavior rather than the simulator DutyBlocked trace, which is zero throughout the frozen campaign. Latency is conditional on successful logical delivery. Simulator latency is stored in milliseconds and converted to seconds in manuscript-facing analysis. Cells with zero logical PDR have no defined conditional delivery latency. Energy results are TX-only energy metrics, derived from transmit airtime. They do not represent complete device, processing, receive, buffering, encoder/decoder, or battery energy. Repository contents The public V5R v3 record contains: CAMPAIGN_V5R_ALL_RUNS.csv — consolidated per-run results for all 2,020 frozen runs. CAMPAIGN_V5R_CELL_SUMMARY.csv — aggregate results for the evaluated scenario cells. PLANNED_CAMPAIGN_V5R.csv — frozen experimental plan. campaign_ledger.csv — campaign execution/provenance information, including per-run SHA-256 provenance fields. MODEL_V6_1_FIT.csv — MODEL V6.1 fitting results. MODEL_V6_1_VALIDATION.csv — MODEL V6.1 validation results. MODEL_V6_1_RHO_VALIDATION.csv — coded-airtime-ratio validation results. C_eff_sensitivity.csv — effective-parallelism sensitivity analysis. model_mae_by_profile.csv — model error summary by profile and generation size. paired_MC_minus_baseline_PDR.csv — paired MC-versus-Baseline PDR analysis. energy_per_packet_and_bit.csv — TX-energy-efficiency analysis. model_v6_1.py — calibrated airtime-recovery model implementation. analysis_v7.py — manuscript-facing analysis script. run_post_campaign_scientific_audit_v5r.py — post-campaign scientific-audit script. CAMPAIGN_V5R_ALL_RUNS.csv provides the consolidated per-run evidence for all 2,020 V5R runs. Individual raw per-process CSV files are not separately included in this Zenodo version. Per-run provenance and SHA-256 information are retained in the campaign data and campaign_ledger.csv. The repository preserves the evidence supporting the revised manuscript and distinguishes the V5R replacement campaign from the earlier Zenodo versions associated with the previous 5,237-run dataset. Important version note: this Zenodo version supersedes the earlier campaign as the evidentiary dataset for the revised IEEE Access manuscript. Earlier record versions are retained for provenance and should not be used to reproduce the numerical results reported in the revised manuscript.