Drone Defense

Download Datasheet (PDF)

Key Features

• Modular passive Counter-UAS detection architecture combining wideband RF sensing, LTE/5G network monitoring and acoustic drone-signature analysis
• 360° RF situational awareness from 400 MHz to 6 GHz in the base configuration, with optional 20 MHz to 8 GHz extended receiver coverage
• LTE/5G network scanning from 350 MHz to 6 GHz with LTE FDD/TDD and 5G NR SA/NSA measurements, broadcast-system-information decoding and spectrum analytics
• 8- or 16-element low-noise MEMS microphone array with 48 kHz, 24-bit capture for rotor-harmonic detection and direction-of-arrival estimation
• Edge classification and central fusion using time alignment, confidence/evidence fusion, track association and Kalman/EKF-based tracking
• Redundant Ethernet/fiber and LTE/5G backhaul, GNSS/PTP timing, secure web GIS, alarm engine and open API interfaces
• Urban and rural deployment profiles for critical infrastructure, airports, borders, prisons, government facilities and temporary event protection
• Detection-only architecture. No RF jamming or other mitigation transmitter is included in this reference specification. Important: values identified as "design target" or "reference target" are engineering objectives derived from public industry benchmarks and literature. Final contractual performance must be confirmed by selected hardware, FAT/SAT procedures and site acceptance trials. complex network needs for Wireless Networks. System Overview & Sensor Node Architecture The MS-CUAS-360 is a distributed passive detection system intended to identify, classify and track small UAS activity without relying on a single sensing modality. RF sensing provides long-range early warning for radio- controlled drones; LTE/5G monitoring adds awareness of cellular spectrum and network activity; acoustic sensing provides a complementary path for RF-silent or pre-programmed aircraft. The fusion layer correlates these measurements into a common track picture. Functional Architecture Subsystem Reference Architecture Wideband antenna / DF array -> preselector and LNA -> multi-channel SDR/receiver -> RF Detection Node spectrum features / protocol classifier -> AoA or line-of-bearing -> secure IP event output. Wideband cellular antenna -> autonomous network scanner -> LTE/NR measurements -> LTE / 5G Monitoring Node MIB/SIB and layer-3 broadcast decoding -> cell/spectrum metrics -> optional authorized operator-side metadata interface. 8/16 matched MEMS microphones -> synchronized digital audio capture -> wind/noise Acoustic Sensor Node pre-processing -> STFT/Log-Mel features -> rotor-signature classifier -> GCC- PHAT/SRP-PHAT direction estimation. GNSS/PTP time alignment -> spatial/temporal association -> confidence fusion -> Fusion & C2 Kalman/EKF tracking -> classification / geolocation -> GIS display, alarms and API integration. Detection Principles
• RF modality is the primary early-warning sensor for common commercial and FPV-type radio links. It is passive and can provide direction finding when a calibrated antenna array is used
• LTE/5G monitoring is used to characterize the cellular RF environment and local network configuration. A passive scanner alone cannot reliably attribute a specific subscriber device to a drone
• Authorized operator-side integration may add network telemetry useful for detecting airborne-UE behavior patterns. The reference architecture explicitly excludes IMSI-catcher functions and subscriber- content interception
• Acoustic sensing detects the tonal and broadband energy generated by motors, propellers and aerodynamic turbulence. It is short-range relative to RF but remains useful when a drone emits little or no detectable control-link RF
• Multi-sensor confirmation is used to improve confidence, suppress nuisance alarms and maintain tracks when one modality is degraded by clutter, multipath, terrain or ambient noise. complex network needs for Wireless Networks. Wideband RF Detection Sensor Node The RF node is designed for continuous passive monitoring of radio-control, telemetry and video signals associated with commercial, industrial and custom UAS. Public C-UAS products demonstrate 400 MHz to 6 GHz class coverage, optional extension toward 8 GHz and direction-finding accuracy in the low-single-digit degree range. [R1][R2][R3] Parameter Specification Frequency range - base 400 MHz to 6 GHz continuous monitoring target Frequency range - option 20 MHz to 8 GHz extended wideband receiver target Priority UAS bands 400/433 MHz, 868/915 MHz, 1.2-1.4 GHz, 2.4 GHz, 5.2/5.8 GHz; configurable custom bands Receiver architecture Multi-channel SDR / monitoring receiver with RF preselection, attenuation and LNA Instantaneous bandwidth 40 to 100 MHz class, receiver dependent Antenna / coverage 360° omnidirectional detection; optional 4/8/16-sector or interferometric DF array Direction finding AoA / line-of-bearing, typical calibrated target <=3° in favorable LoS geometry Geolocation Multi-node AoA triangulation; optional TDOA with synchronized receivers Signal processing Energy/occupancy analysis, protocol features, known-signature matching and unknown-emitter anomaly detection Typical RF range target Urban 1-3 km; rural/open LoS 2-5 km for common commercial UAS with active control/video links Environmental / interfaces IP65 class; -20°C to +55°C target; GNSS, PTP/NTP, 1GbE, optional PoE++ and cellular backup RF range depends on transmitter power, protocol, site elevation, spectrum occupancy and line-of-sight. The 2-5 km class reflects public RF-C-UAS benchmarks and is not a guaranteed envelope. [R1][R2] complex network needs for Wireless Networks. LTE / 5G Network Monitoring Sensor The LTE/5G monitoring node complements generic RF detection by continuously measuring the cellular radio environment. Current autonomous network scanners support simultaneous LTE/5G NR measurements across 350 MHz to 6 GHz, including SIB/layer-3 decoding, while newer wideband scanners extend above 6 GHz for full FR1 coverage. [R4][R5] Supported Measurements
• LTE FDD/TDD, LTE-M, NB-IoT and 5G NR SA/NSA measurement support, depending on scanner software license and regional configuration
• RF metrics: RSSI, RSRP, RSRQ, SINR, channel power, spectrum occupancy and interference indicators
• Cell/network metadata: PLMN, band, EARFCN / NR-ARFCN, PCI, Cell ID / NCI where broadcast, TAC, SSB/beam information and decoded MIB/SIB system information
• Cell-change and spectrum-anomaly events can be fused with RF and acoustic detections to increase confidence and explain environmental changes
• Optional authorized mobile-operator interface may ingest network-side metadata or alarms for airborne- UE analytics. This is outside the capability of a stand-alone passive scanner and is subject to operator authorization, privacy law and local regulation. Reference Cellular Frequency Coverage Parameter Specification Scanner RF span - base 350 MHz to 6 GHz Optional wideband extension 10 MHz to 8.5 GHz class for complete 5G NR FR1 coverage through 7.125 Low bands GHz /[R5] 600 700 / 800 / 900 MHz families - e.g., LTE B12/B13/B20/B28 and NR n5/n8/n20/n28 where regionally deployed Mid bands 1.7/1.8 GHz, 1.9/2.1 GHz, 2.3 GHz, 2.5/2.6 GHz - LTE B1/B3/B7/B38/B40/B41 and related NR bands 5G FR1 C-band n77 / n78 approximately 3.3-4.2 GHz; n79 approximately 4.4-5.0 GHz Upper FR1 / shared spectrum n46 / 5 GHz family where applicable; optional receiver extension for 6- 7.125 GHz allocations FR2 option External downconverter/scanner module, typically 24-53 GHz class, only when mmWave monitoring is required Network decode LTE and 5G NR broadcast system information to Release 17 class, scanner dependent [R4] Typical update rate 1-20 Hz aggregate engineering target depending on number of configured carriers and technologies Timing / interfaces GNSS, PTP/NTP, 1GbE, API output; optional 5G/LTE backhaul modem Cellular-monitoring limitation: passive scanning provides RF and broadcast network awareness. It does not by itself prove that a particular LTE/5G device is a drone. Per-device attribution requires authorized operator-side telemetry or another independent sensor correlation. complex network needs for Wireless Networks. Acoustic Sensor Array & Rotor Signature Processing The acoustic node is a passive close-to-medium-range sensor that exploits the characteristic tonal structure of multirotor motors/propellers and the broader aerodynamic noise spectrum. Research and deployed acoustic C-UAS systems show that useful drone energy extends from low hundreds of hertz through several kilohertz, with 48 kHz sampling commonly used for robust feature extraction. [R6][R7][R8] Parameter Specification Array configuration 8 microphones standard; 16 recommended; optional 32/64-element site array Microphone technology Matched low-noise MEMS, digital PDM or low-noise differential analog front end Microphone benchmark 69-70 dB(A) SNR class, >=124 dB SPL AOP; low-frequency cutoff <=30-75 Hz [R7] Capture bandwidth 20 Hz to 20 kHz nominal audio capture Primary analysis band 100 Hz to 8 kHz for rotor harmonic and blade-passing signatures; higher-frequency features retained when useful Sampling / resolution 48 kHz, 24-bit synchronized multichannel capture Array geometry Circular / hemispherical 250-500 mm aperture, phase calibrated, with outdoor wind protection Feature extraction Wind/high-pass filtering -> STFT -> Log-Mel spectrogram -> harmonic and spectral-envelope features Classification CNN / compact transformer: drone, non-drone, unknown; site- trained nuisance models Direction finding GCC-PHAT / SRP-PHAT / beamforming; reference DOA target <=5° at adequate SNR Reference range targets Small multirotor: urban 100-250 m, rural/open 200-500 m; louder/fixed-wing UAS: up to ~1 km favorable [R6] complex network needs for Wireless Networks. Data Fusion, Classification & Multi-Target Tracking The fusion engine converts heterogeneous sensor events into a single georeferenced track picture. The architecture uses metadata-level fusion for routine operation and stores raw RF/audio snapshots locally for forensic review or algorithm improvement. Kalman-family tracking and evidence-based fusion are well established approaches for asynchronous multi-sensor UAV tracking. [R9][R10] Fusion Pipeline Stage Function 1. Time alignment GNSS / PTP timestamps are normalized to a common time base; asynchronous measurements are interpolated within configurable time gates. 2. Event normalization Each sensor reports position/bearing, band/frequency, class probabilities, SNR/quality, timestamp and sensor-health indicators. 3. Association & gating Candidate measurements are associated by time, angle, estimated position, motion compatibility and sensor-specific quality. 4. Confidence fusion Weighted Bayesian scoring or Dempster-Shafer evidence combination merges partially independent class/confidence estimates while preserving uncertainty. 5. Tracking Kalman / Extended Kalman filtering is the default constant-velocity tracker; UKF or GM-PHD multi-target tracking is optional for dense or maneuvering scenarios. 6. Track management Tentative -> confirmed -> coast -> dropped track logic prevents single- sensor spikes from creating persistent operational alarms. 7. Threat / policy engine User-defined zones, time-of-day rules, whitelist/authorized UAS data and confidence thresholds generate alerts without automatic mitigation action. complex network needs for Wireless Networks. Network Backhaul, Edge Computing & Cybersecurity The reference system is designed as a distributed IP sensor network. Routine traffic consists primarily of metadata and compressed event features, while high-rate raw RF I/Q or multichannel audio is retained at the edge and transferred only on demand. Parameter Specification 1GbE copper or fiber Ethernet; PoE++ IEEE 802.3bt where Primary backhaul sensor power budget permits Integrated or external LTE/5G router; optional microwave / Redundant backhaul private LTE backhaul for remote sites 2.5/10GbE for sustained raw sensor transfer or multi-receiver High-rate option research configurations Typically <1 Mbps average per node; burst rate depends on Metadata bandwidth event frequency and logging policy Approx. 9.2 Mbps for 8 channels or 18.4 Mbps for 16 channels Raw acoustic data at 48 kHz / 24-bit before framing/compression Can exceed 1 Gbps for tens of MHz of complex bandwidth; Raw RF I/Q therefore normally captured locally and retrieved on demand GNSS disciplined clock plus IEEE 1588 PTP; NTP fallback for Time synchronization non-DF services REST/JSON, MQTT over TLS, WebSocket event feed, SNMPv3, Protocols syslog/SIEM; optional Cursor-on-Target gateway Encryption TLS 1.3 in transit; AES-256 class encryption at rest Mutual X.509 certificates, RBAC, MFA at C2, signed firmware Identity / access and secure-boot target Store-and-forward event queue, heartbeat / sensor health, Resilience watchdog restart, redundant C2 server option On-premises preferred for sensitive sites; private cloud or Deployment hybrid deployment optional complex network needs for Wireless Networks. Recommended Edge Compute Profile
• RF node: FPGA/DSP capable receiver plus x86 or ARM edge CPU; optional CUDA-class GPU/NPU for wideband AI classification
• Acoustic node: low-power ARM/x86 processor with DSP acceleration; local 30-120 minute rolling audio buffer configurable by policy
• Fusion server: 8-core or higher CPU, 32-64 GB RAM, NVMe event storage and optional GPU for model inference / retraining; redundant server pair for critical sites
• All software components should support remote signed updates, rollback and configuration audit history. Recommended data-minimization principle: retain only the sensor data required for UAS detection, engineering analysis and incident evidence. Cellular content interception and unrelated communications collection are outside the reference design. Urban & Rural Operational Performance The following values are reference engineering targets for system sizing. They are not guaranteed detection envelopes. Final performance depends on target type, RF emissions, acoustic signature, terrain, antenna/sensor height, weather, RF congestion and site- specific nuisance sources. Metric Urban Rural / Open Engineering Notes RF detection - small 1-3 km typical 2-5 km typical Active control/video link; 360° commercial UAS antenna. Stronger/louder or elevated emitters may be detected Acoustic - small 100-250 m typical 200-500 m typical Strongly affected by traffic, farther. multirotor machinery, wind and microphone elevation. Acoustic - larger fixed- 200-500 m class Up to ~1 km favorable Consistent with public acoustic- wing / loud UAS vector-sensor benchmarks for louder aircraft classes. [R6] complex network needs for Wireless Networks. LTE/5G environment Local monitored cell Local monitored cell Not a conventional range metric. monitoring footprint footprint Per-device attribution requires authorized network-side integration. RF direction finding 3-6° practical target in <=3° target in calibrated Geometry, array calibration and clutter LoS multipath dominate error. Acoustic DOA 5-10° practical target <=5° target at adequate Array aperture and wind/noise SNR control dominate error. Initial alert latency 1-3 s target 1-3 s target Depends on confirmation policy and classification window. Fusion track update 1-5 Hz typical 1-10 Hz typical Higher update rates available when sensor events are frequent. Qualified fusion >=95% with >=2 >=95% with >=2 Qualification target only; must be detection target independent modalities independent modalities defined against a controlled inside validated range inside validated range reference drone set and flight Nuisance alarm target <=1 alarm/hour after <=1 alarm/hour after Reference commissioning target, profile. site tuning site tuning not a universal guaranteed FAR. Consolidated Technical Specification - Reference Configuration System-Level Specifications Parameter Specification System type Passive multi-sensor Counter-UAS detection and situational-awareness platform Sensor modalities Wideband RF detection / DF, LTE/5G network monitoring, acoustic array Coverage 360° RF and acoustic coverage with site-dependent elevation coverage Processing Distributed edge analytics with centralized track fusion and web GIS Node scale 1 to 64 sensor nodes per fusion-server reference target; software-scalable Concurrent tracks >=100 track software target, configuration dependent Time base GNSS + PTP; NTP fallback Event interfaces REST/JSON, MQTT/TLS, WebSocket, SNMPv3, syslog; optional third-party C2 gateway Cybersecurity mTLS, RBAC, signed updates, encrypted storage, audit logs Deployment Fixed site, temporary tripod, rooftop, mast, vehicle-supported sensor kit Mitigation Not included in base system; external authorized effectors may be integrated through a separate C2 interface complex network needs for Wireless Networks. Outdoor Sensor Node Environmental / Power Targets Parameter Specification Ingress protection IP65 minimum reference target; IP66 option where selected hardware permits Operating temperature -20°C to +55°C standard target; -30°C to +60°C extended option Storage temperature -40°C to +70°C target, selected hardware dependent Power input 48 VDC PoE++ or 24/48 VDC; 100-240 VAC via external PSU option Typical power - RF node 25-60 W depending receiver / antenna electronics Typical power - LTE node 20-60 W depending network scanner and compute Typical power - acoustic node 10-25 W depending microphone count and edge compute Mounting Pole / mast / rooftop / tripod with surge protection and grounding GNSS antenna External active GNSS antenna or integrated sky-view module Maintenance Remote health monitoring, field-replaceable PSU/network components, modular sensor electronics complex network needs for Wireless Networks. Typical Applications & Public Technology Benchmarks Typical Applications Critical Infrastructure Airports & Airfields Borders & Remote Sites Power plants, oil & gas, telecom Perimeter monitoring, approach- Distributed sensor grids with fiber, sites, data centers and industrial area awareness and temporary high- private LTE or cellular backhaul. campuses. risk events. Prisons / Government VIP / Public Events Defense / Public Safety Unauthorized drone delivery and Rapidly deployable sensor nodes for Passive local airspace awareness and surveillance detection around temporary protected zones. C2 integration, subject to customer restricted facilities. authorization and local regulation. Representative Public Market References - Engineering Benchmark Only Dedrone RF-360 R&S TSMA6B Microflown SKYSENTRY Passive networked RF sensor. Public data Autonomous mobile-network scanner Passive acoustic-vector-sensor C-UAS sheet states 2 km normal and up to 5 km covering 350 MHz to 6 GHz with LTE/5G concept. Public tests cite up to 250 m for ideal detection for specific drones; AoA NR and broadcast SIB/L3 decoding a small 2 kg quadcopter and up to 1 km localization; IP65; LTE/Ethernet through Release-17 class capabilities. for a small fixed-wing aircraft. [R6] connectivity. [R1] [R4] The reference products above are not represented as EMTS components or committed suppliers. They are shown only to document the public technical benchmark used while defining this reference architecture. complex network needs for Wireless Networks. Engineering References [R1] Dedrone RF-360 Data Sheet V2.0 and RF-360 product page - passive RF detection, AoA localization, 2 km normal / up to 5 km ideal range, IP65, LTE/Ethernet. [R2] Rohde & Schwarz ARDRONIS Locate Compact - 400 MHz to 6 GHz monitoring and >5 km class threat detection in public product description. [R3] Aaronia AARTOS DDS X9 public technical material - 400 MHz to 6 GHz base RF coverage, optional extension toward 8 GHz, 1° to 3° class tracking accuracy and IP65 class hardware. [R4] Rohde & Schwarz R&S TSMA6B product brochure - 350 MHz to 6 GHz autonomous mobile network scanning, LTE/5G NR and SIB/layer-3 decoding. [R5] ETSI / 3GPP TS 38.101-1 - 5G NR FR1 frequency range defined through 7.125 GHz in current releases. [R6] Microflown AVISA SKYSENTRY - passive acoustic C-UAS sensor post; public initial test ranges up to 250 m for small quadcopter and 1 km for small fixed-wing aircraft. [R7] Infineon IM69D130 and TDK/InvenSense high-SNR MEMS microphone data sheets - 69-70 dB(A) SNR class, 124-130 dB SPL AOP class and wide audio response suitable for microphone arrays. [R8] Sensors / acoustic-UAS literature - 48 kHz capture, STFT and Log-Mel features for rotor harmonic extraction; drone harmonic energy commonly extends through several kHz. [R9] Multi-sensor UAV tracking literature - Kalman filtering for asynchronous heterogeneous sensor measurements and improved track continuity. [R10] MUTES acoustic/optical fusion research - microphone-array cueing and Kalman-based multi-sensor trajectory fusion. complex network needs for Wireless Networks