UAV Underwater Drones network traffic statistics

Dataset: UAV and Underwater Drone Network Anomaly Dataset (NTNU and SINTEF) Overview This dataset contains real time operational telemetry and network traces from the Trondheim Fjord region in Norway. Data originates from long running campaigns conducted by the Marine Cybernetics Laboratory at the Norwegian University of Science and Technology and the SINTEF Ocean Research Center. Platforms include autonomous underwater vehicles, remotely operated vehicles, unmanned aerial vehicle relays, coastal buoys, and a shoreside gateway. The collection spans February 2022 to March 2025 and reflects real world marine communication where acoustic, radio frequency, and satellite links operate together. The public release is anonymized and standardized for research in anomaly detection, graph and temporal modeling, and privacy aware federated learning. Sensitive identifiers were removed in line with Norwegian maritime governance and European Union GDPR requirements. What is included File format Comma separated values file with header One row per client per timestamp Time zone naive timestamps in ISO 8601 format Temporal coverage Start: 2022 02 01 End: 2025 03 31 Sampling interval: thirty minutes in this release Note: the file name may reference ten minute cadence from earlier internal builds. The current artifact uses thirty minute cadence. Participants and roles Ten nodes total across the fleet configuration: three autonomous underwater vehicles, two remotely operated vehicles, two unmanned aerial vehicles, two coastal buoys, one gateway. Column count Total columns: 113 General identifiers: 4 Feature columns: 97 Multi label targets: 12 Schema and feature groups Field names match column headers exactly. General timestamp, client_id, role, mission_id Network and transport pkt_tx, pkt_rx, pkt_drop, pkt_retx throughput_up_kbps, throughput_down_kbps latency_avg_ms, latency_p95_ms, latency_jitter_ms rtt_avg_ms, rto_events syn_cnt, fin_cnt, rst_cnt udp_flows, tcp_flows, flow_churn_rate payload_entropy, header_anomaly_score routing_updates, route_changes, ttl_avg queue_len_avg, queue_drop_rate Link layer and physical channel mac_tx_fail_rate, collision_rate snr_db, ber, bler tx_power_dbm, rx_power_dbm channel_utilization_percent, cs_busy_time_ms freq_shift_hz, multipath_delay_ms Underwater environment sound_speed_mps, absorption_coef_dbkm ambient_noise_db, shipping_noise_db, sea_state thermocline_depth_m, water_temp_c, salinity_psu, turbidity_ntu UAV relay and backhaul uav_link_rssi_dbm, uav_link_snr_db, uav_link_latency_ms handover_events, uav_backhaul_type uav_altitude_m, uav_speed_ms, uav_heading_deg uav_relay_duty_percent, uav_link_uptime_percent Kinematics and platform health depth_m, pitch_deg, roll_deg, yaw_deg speed_ms, accel_rms, turn_rate_dps prop_rpm, thruster_load_percent imu_health_flag, nav_mode Energy and thermal battery_soc_percent, battery_voltage_v, battery_temp_c, energy_draw_w cpu_util_percent, mem_util_percent, gpu_util_percent board_temp_c, thermal_throttle_events Security telemetry port_scan_hits, syn_flood_score, frag_anomaly_score dns_req_rate, dns_nxdom_rate arp_spoof_score, mac_spoof_score auth_failures, cert_errors, ids_alerts_count, ml_ids_score Topology and context neighbors_count, link_changes, gw_hops path_stability_score, link_distance_m, uav_distance_m, mission_phase Derived temporal indicators latency_trend, throughput_cv, pkt_drop_trend snr_min_last5, rtt_ema, anomaly_ewma Labels Multi label targets for twelve conditions. A value of one indicates presence at the timestamp, otherwise zero. Normal, Latency_Anomaly, PacketLoss_Anomaly, Throughput_Degradation, Jitter_Anomaly, Jamming_Attack, Spoofing_Attack, Replay_Attack, Routing_Blackhole, Channel_Distortion, Sensor_Fault, Energy_Fault