IoT and Embedded Systems for RF-Based Detection and Electromagnetic Sensing

Case Study: Airborne RF Monitoring for UAVs & Airspace Security

Šarūnas Paulikas, Raimondas Pomarnacki

Airspace Is No Longer Empty

• Low-altitude aerial objects are increasing
• Hobby drones
• Commercial UAVs
• Unknown objects
• Balloons
• Civil + hybrid security concerns
• Airports, borders, infrastructure affected

2

Monitoring Challenges in Rural RF Environments

• Technical Challenges:
• Limited cellular coverage
• High propagation variability (terrain, forests)
• Multipath reflections from tree canopy
• Embedded system implications:
• Spectrum sensing must adjust dynamically.
• Energy constraints (no grid power nearby).
• UAV-based sensing may provide LOS advantage.

3

Why RF-Based Detection Matters Today

• Increased tensions and hybrid activity along the Lithuania–Belarus border.
• Why balloons are problematic:
• Slow movement → filtered by radar clutter algorithms.
• No active transponder.
• No Remote ID.
• Hard to classify visually at high altitude.

4

RF‑Based Detection

• RF-based detection enables identification of both cooperative UAVs (with Remote ID) and non‑cooperative ones (no ID, spoofed ID, or intentionally silent).
• Ground sensors suffer from NLOS (non-line-of-sight) limitations in urban, forest environments where obstacles block radio propagation.
• Mobile airborne sensors mounted on UAVs can overcome these limitations by providing LOS access above rooftops and other dense structures.

5

5G for UAVs

• 3GPP Release 15: Initial UAV connectivity studies
• 3GPP recognized that UAVs are becoming large-scale network users.
• Unlike smartphones, UAVs operate in 3D space and require: Realible (BVLOS) connectivity; Low latency for control and telemetry, High uplink capacity for video and sensing data
• Release 17–18 (5G-Advanced): Enhanced UAV support
• Height-aware mobility management; Interference mitigation; Better uplink reliability
• However, networks are still optimized primarily for ground users
• Chalage - Interference management at altitude

6

What Changes When UAVs Use 5G?

• Base station antennas are down-tilted: it connects through antenna side lobes
• UAV “sees” multiple cells simultaneously
• Increased inter-cell interference
• Frequent handovers
• SINR instability at altitude: Signal strength may be high (RSRP), but quality (SINR) may degrade due to interference.

7

5G for RF-Based UAV Detection

• Real-time data uplink: The UAV is not only detecting signals — it must transmit results instantly
• Mission support
• Network slicing for priority traffic
• Secure licensed spectrum
• Integration with U-Space systems:
• without 5G, the UAV becomes an isolated sensor;
• with 5G, it becomes part of a connected airspace management system.

​

8

IoT Principles in RF Monitoring

RF sensing systems increasingly follow IoT architecture principles:

• Distributed sensor nodes equipped with SDRs, antennas, and embedded controllers.
• Edge computing at each sensor for local filtering and signal conditioning.
• Cloud‑based aggregation for data fusion, triangulation, and visualization. Such system could demonstrate this hybrid approach: edge DSP in base stations + cloud‑level triangulation + integration with national U‑Space systems.

9

RF Sensing Techniques

RF detection consists of several parallel methods:

• Passive radar: does not emit energy, listens to existing RF signals for reflections or emissions.
• Signal intelligence: identifies telemetry, video, control links typical of drone operation.
• Direct Remote ID (DRI) reception: Wi‑Fi/BLE beacons following ASTM F3411.
• Broadband spectrum scanning: using SDRs to detect unknown emitters, interference, or EMI patterns.
• Our approach integrates all three simultaneously in its hybrid detection nodes.

10

Embedded Hardware for RF Sensing

Modern RF sensing platforms rely on powerful yet lightweight embedded systems:

• SDR modules (BladeRF 2.0, USRP B210, KrakenSDR) enabling wideband and multi-channel reception.
• Embedded SBCs such as Raspberry Pi 4 or Odroid N2 used for data acquisition, DRI decoding, and preprocessing.
• RF front‑ends with LNAs, band‑pass filters, and high‑dynamic‑range ADCs to capture weak UAV emissions.

These components are specifically identified in testing and UAV RF‑monitoring prototypes.

11

Case Study 1: UAV RF Sensing Platform

The three primary subsystems developed:

• Integrated Remote ID scanner-modem — receives RID frames in the air and relays them to U‑Space via 5G.
• RF seeker — detects and geolocates UAVs without RID by analyzing emitted RF signals.
• Drone laboratory — a flying RF laboratory with a spectrum analyzer for mapping cellular network quality at altitude.
• Together they form a modular airborne RF sensing platform.

12

System Architecture of the UAV Platform

Core architecture includes:

• SDR module for RF signal acquisition (telemetry/video links).
• 5G modem (Quectel RM500Q or similar) for low‑latency uplink streaming.
• Embedded processor for on‑board filtering, classification, and RID decoding.
• Server‑side platform interfacing with U‑Space and ANSP systems for air‑traffic integration.
• The architecture achieves real‑time data transmission from a UAV to aviation systems.

​

​

13

Non‑Cooperative UAV Detection Pipeline

Detecting drones without Remote ID relies on multi-stage RF processing:

• RF energy detection to locate any active transmitter.
• Spectrogram generation for identifying modulated signals.
• Cyclostationary feature extraction for identifying structured signals vs noise.
• Deep learning classification to distinguish UAV patterns (telemetry protocols, FPV links, motor EMI).

14

Geolocation: AoA, TDoA, Fusion

RF-based localization uses:

Angle-of-Arrival (AoA): Using multi‑element antenna arrays and algorithms like MUSIC.

Time Difference of Arrival (TDoA): Using multiple synchronized stations with nanosecond‑precision GPSDO oscillators.

Joint Fusion: Central fusion system combines hyperbolic TDoA positions with AoA vectors to reduce geometric ambiguity.

15

Case Study 2: Passive Radar System

System provides a full-scale passive RF radar for UAV detection:

• Multi‑node distributed SDR network for wide-area monitoring.
• On‑site embedded AI for noise filtering and signal classification.
• Cloud‑based TDoA/AoA fusion to compute UAV coordinates. Seamless integration to “AB Oro Navigacija” UTM interface.

This system represents state‑of‑the‑art IoT-based RF sensing for national airspace.

16

Technical Architecture

Each node consists of:

• RF front-end with broadband antennas, LNAs, filters.
• SDR devices (BladeRF) for synchronous multi-channel reception.
• RPi SBC for DSP, including FFT, energy detection, RID decoding.
• Cloud connection for data fusion and operator GUI.
• This architecture supports both broadband scanning and protocol‑specific decoding.

17

Direct Remote ID (DRI) Fundamentals

ID signals include:

• UAV ID, operator ID
• GPS position, altitude
• Ground station coordinates
• Velocity, heading, system status

RID uses Wi‑Fi or Bluetooth advertising according to ASTM F3411.

System implements full decoding of DRI messages in both BLE and Wi‑Fi.

18

Embedded RID Receiver (Example)

Hybrid DRI receiver developed using:

• Odroid N2 as main processing platform.
• BLE dongles (nRF52840) for Bluetooth RID.
• Wi‑Fi dongles (Alfa AWUS036ACHM) for Wi‑Fi RID capture.
• Local processing pipeline: scanning → CRC check → RID validation → UTM API upload.

Compatible with OpenDroneID and commercial RID devices (Dronetag, DroneScout).

19

AI in RF Classification

AI enables classification of RF emissions even when signals are weak or noisy:

• ANNs analyze spectrograms to classify UAV telemetry.
• ML methods detect anomalies such as spoofed RID.
• Deep models integrate AoA/TDoA measurements for improved localization accuracy.

20

Use of 5G Networks in Airborne Sensing

UAV-mounted 5G modules provide:

• Low-latency uplink for streaming RF data and RID messages.
• Ability to test 5G coverage, SINR, RSRP, RSRQ at altitude.
• Support for 5G Advanced features outlined in 3GPP Release 18.

​

The UAV should measure 5G quality up to 500 m altitude.

21

Engineering Challenges

RF sensing platforms must handle:

• EMI/EMC issues between SDRs, antennas, and 5G modems.
• Payload and power constraints on UAVs.
• Synchronization drift in multi-node TDoA systems.
• Regulatory limitations on RF monitoring frequencies.

22

​

​

​

Thank You ☺

23