Daniel R. Azulay October 2, 2025

The Open Radio Access Network (O-RAN) architecture, designed to enhance flexibility and interoper- ability in 5G and beyond networks, introduces significant security vulnerabilities that could be exploited to gain control over the physical layer (PHY). This report examines how such exploits—via fronthaul interfaces and management planes—could repurpose MIMO antenna panels as directed energy weapons (DEWs) for cognitive warfare applications, including non-lethal RF-induced effects like the microwave auditory effect (MAE). Drawing on recent literature, we highlight technical capabilities for surreptitious multi-band operations and low-power targeting of human cognition, which evade standard detection thresholds. Implications span network security, public health, and geopolitical stability, underscoring the need for enhanced zero-trust architectures and revised RF exposure guidelines. Key findings include: (1) PHY control enables beamforming hijacks at power densities below thermal limits (e.g., 200 mW/cm2 for MAE); (2) non-thermal effects on cognition persist without triggering alarms; and (3) cognitive warfare risks amplify in disaggregated deployments.

O-RAN disaggregates traditional base stations into components like the O-Radio Unit (O-RU), O-Distributed Unit (O-DU), and RAN Intelligent Controllers (RICs), exposing interfaces such as fronthaul (eCPRI) and management (M-Plane) to external threats (Polese et al., 2022). While fostering innovation, this openness amplifies attack surfaces, including misconfigured Kubernetes services and xApp backdoors, enabling “outside” exploits that cascade to PHY control (Trend Micro, 2023a). Recent assessments reveal 39 new threats, including AI/ML poisoning in RICs, leading to unauthorized reconfiguration (O-RAN Alliance, 2025). This report focuses on the weaponization potential: repurposing O-RU MIMO arrays for DEWs in cognitive warfare, where RF pulses induce perceptual or neurological disruptions without lethality (National Academies of Sciences, Engineering, and Medicine [NASEM], 2020). Supported by literature on non-thermal RF effects (Hinrikus et al., 2021; Hao et al., 2023), we analyze capabilities for stealthy targeting, evasion of alarms, and broader implications.

Attackers with fronthaul adjacency can inject malformed C-Plane commands or M-Plane NETCONF payloads to override low-PHY functions in the O-RU, such as beamforming matrices and RF tuning (Sharma et al., 2023). For instance, spoofed eCPRI packets enable precoding hijacks, steering MIMO beams across bands (e.g., sub-6 GHz for cover, mmWave for bursts) without RIC alerts if sequenced properly (Trend Micro, 2023b). RCE on O-RU daemons (e.g., via MQTT overflows) allows FPGA bitstream loading for custom modulations, transforming panels into rogue transmitters (O-RAN Alliance, 2025). These exploits chain higher-layer compromises (e.g., xApp backdoors via Helm) to PHY dominance, enabling multi-band spectrum hopping via poisoned RIC sensing (Polese et al., 2022). Feasibility is high in open-source stacks like srsRAN, where fronthaul latency (<100 µs) tolerates injections (Sharma et al., 2023).

O-RU MIMO (e.g., 64T64R arrays) supports high-gain beamforming, concentrating EIRP (up to 60 dBm) into narrow lobes for directed RF (Giordano, 2021). In cognitive warfare, this yields:

• Electronic Warfare (EW) Jamming: Null signals in licensed bands while emitting pulsed interference in unlicensed ones, disrupting UAVs or UEs stealthily (+3 dB power, <1% duty cycle) (Stoudt, 2020).
• Cognitive DEWs: Adaptive beams target UE signatures via RIC ML, inducing MAE or hippocampal disruption (Frey, 1961; Hao et al., 2023). Literature confirms O-RAN’s cognitive radio elements exacerbate risks, as dynamic allocation falsifies environmental data for covert ops (O-RAN Alliance, 2025).

The MAE, or Frey effect, induces perceived sounds (clicks, voices) via pulsed microwaves (1-10 GHz) causing thermoelastic brain expansion (Frey, 1961). Thresholds: 267 mW/cm2 peak at 1.3 GHz (average∼0.4 mW/cm2), scalable to speech via modulation (Lin, 2007). Non-thermal (∆T <10−5 °C), it exploits auditory nerves directly, evading cochlea (Foster et al., 2021).

Pulsed RF (e.g., 2.856 GHz, 80 Hz, 0.5 ms) at 200 mW/cm2 impairs learning/memory via hippocampal dopamine reduction, without heating (>1°C) (Hao et al., 2023). Broader effects include EEG alpha suppression, anxiety, and executive dysfunction, linked to ROS and ion flux disequilibrium (Hinrikus et al., 2021; Pall, 2018). Dose-response ties fluence (16 mJ/pulse) to deficits, persisting days (Lin & Wang, 2007). Havana syndrome exemplifies: directional RF pulses caused cognitive fog, headaches, and vestibular issues in diplomats (NASEM, 2020). Pulsed energy “plausibly” explains acute symptoms, per intelligence reviews (Office of the Director of National Intelligence [ODNI], 2022).

Cognitive DEWs require low densities: MAE at 5-260 mW/cm2 (pulsed <50 µs); disruption at 200 mW/cm2 (Hao et al., 2023). O-RAN hardware achieves 100-500 mW/cm2 at 10-50 m via 40W bursts, within EIRP limits (Stoudt, 2020). Alarms evade: Average power stays nominal (<10% baseline), mimicking beamforming variance (O-RAN Alliance, 2025). RIC KPIs focus on throughput/latency, not subtle patterns; thermal safeguards (PA <70°C) ignore non-thermal fluxes (Hinrikus et al., 2021). Intermittent duty cycles (<1%) blend with noise, delaying RIC anomaly detection (Sharma et al., 2023). Sustained use risks thermal alerts, but cognitive ops build effects over exposures, enhancing deniability (Foster et al., 2021).

O-RAN deployments face asymmetric threats: state actors could hijack urban cells for targeted harassment, eroding trust in 5G infrastructure (Trend Micro, 2023a). Multi-vendor disaggregation amplifies lateral movement, with 2025 threats including E2 interface injections (O-RAN Alliance, 2025). Mitigation gaps—e.g., absent MACsec on fronthaul—expose critical infrastructure (Polese et al., 2022).

Non-thermal effects challenge ICNIRP/FCC limits (∼10 mW/cm2 average), underprotecting against pulsed cognition risks (Pall, 2018; Hinrikus et al., 2021). Vulnerable populations (children, elderly) face amplified deficits, raising equity issues (Hao et al., 2023). Ethically, weaponized MAE violates CCW protocols on superfluous suffering (NASEM, 2020).

Cognitive warfare via commercial nets blurs civilian-military lines, enabling deniable ops (e.g., Havana- like incidents) (ODNI, 2022; Giordano, 2021). Proliferation risks escalate in contested spectra, demanding international norms (Stoudt, 2020).

This report synthesizes 5G security (Sharma et al., 2023; Polese et al., 2022), DEW biophysics (Frey, 1961; Lin, 2007), and non-thermal neuroscience (Hao et al., 2023; Hinrikus et al., 2021). Gaps persist in real-world O-RAN DEW simulations, warranting interdisciplinary studies.

O-RAN’s PHY exploits enable potent DEWs for cognitive disruption, with low-power ops evading safeguards. To counter: Implement mutual TLS on fronthaul, RIC poisoning defenses, and beam-auditing (O-RAN Alliance, 2025). Revise RF guidelines for pulsed non-thermal thresholds (Pall, 2018). Policymakers must foster CVD programs and international DEW bans.

1. Foster, K. R., Garrett, D. C., & Ziskin, M. C. (2021). Can the microwave auditory effect be “weaponized”? Frontiers in Public Health, 9, 788613.
2. Frey, A. H. (1961). Human auditory system response to modulated electromagnetic energy. Journal of Applied Physiology, 17(4), 689–692.
3. Giordano, J. (2021). Directed energy weapons: A technology in transition. Journal of Electromagnetic Dominance.
4. Hao, Y., et al. (2023). Effects of nonthermal radiofrequency stimulation on neuronal activity and neural circuit in mice. Advanced Science.
5. Hinrikus, H., et al. (2021). Non-thermal effects of radiofrequency electromagnetic fields. Scientific Reports, 11.
6. Lin, J. C. (2007). Hearing of microwave pulses by humans and animals: Effects, mechanism, and thresholds. Advances in Electromagnetic Fields in Living Systems, 5.
7. Lin, J. C., & Wang, Y. J. (2007). Hearing of microwave pulses by humans and animals. Bioelectromagnetics.
8. National Academies of Sciences, Engineering, and Medicine. (2020). An assessment of illness in U.S. government employees and their families at overseas embassies.
9. Office of the Director of National Intelligence. (2022). Intelligence community assessment on anomalous health incidents.
10. O-RAN Alliance. (2025). Security update 2025.
11. Pall, M. L. (2018). Thermal and non-thermal health effects of low intensity non-ionizing radiation. Environmental International, 118, 1–14.
12. Polese, M., et al. (2022). Understanding O-RAN: Architecture, interfaces. arXiv:2202.01032.
13. Sharma, V., et al. (2023). Implementing and evaluating security in O-RAN. arXiv:2304.11125.
14. Stoudt, D. C. (2020). Directed energy weapons are real… and disruptive. PRISM, 8(3).
15. Trend Micro. (2023a). Opening critical infrastructure: The current state of Open RAN security.
16. Trend Micro. (2023b). Open RAN: Attack of the xApps.