The exponential growth of global data traffic has revealed fundamental limitations in traditional radio frequency (RF) communication, including spectrum scarcity, electromagnetic interference, and vulnerability to information leakage, which impede its capacity to meet 6G networks' demanding requirements for high speed, low latency, and reliable connectivity. Optical wireless communication (OWC), which utilizes light in visible, infrared, and ultraviolet bands as the transmission medium, emerges as a revolutionary solution, providing abundant spectrum resources, high data rates, and immunity to electromagnetic interference, among other benefits. However, despite its technological advantages, OWC encounters substantial security challenges including eavesdropping, signal interception, and channel disturbances at the physical layer that compromise its deployment in critical sectors such as military communications, satellite networks, and intelligent infrastructure. Addressing these security vulnerabilities is essential to realizing OWC's full potential and securing its position as a foundational technology for future intelligent communication systems.

free space optical (FSO) communication, visible light communication (VLC), and underwater wireless optical (UOWC) communication (Figs. 1?3). FSO systems employ near-infrared lasers to establish high throughput point-to-point links for long distance transmission, facilitating secure backhaul connectivity in urban networks, inter-satellite communications, and supporting communication requirements in emergency response scenarios (Fig. 1). Utilizing light emitting diodes operating in the 400?800 THz spectral band, VLC provides dual functionality of illumination and data transmission, enabling innovations in smart home automation, industrial internet of things ecosystems, and secure indoor positioning systems (Fig. 2). UOWC employs the blue-green light window where water absorption is minimal to overcome the bandwidth limitations of underwater acoustic communication, supporting applications in marine sensor networks, offshore exploration, and naval tactical operations (Fig. 3).

spatial eavesdropping emerges from beam divergence in FSO and the broadcast nature of VLC, enabling passive interception through receiver deployment or beam splitting. Channel induced vulnerabilities, such as atmospheric turbulence in FSO and underwater multipath fading in UOWC, compromise signal integrity and expand attack surfaces. Protocol level risks, including RF side channel sniffing in VLC and beam splitting attacks in FSO, challenge traditional encryption methods.

To address these, physical layer security technologies have been developed: chaotic encryption employs nonlinear systems like Lorenz and R?ssler models to generate unpredictable signals (Fig. 4), resisting eavesdropping in VLC and FSO despite synchronization challenges from atmospheric turbulence. Beam control techniques, such as orbital angular momentum (OAM) multiplexing, encode data into orthogonal spiral phase modes to complicate interception via mode specific decoding. In addition, optical fractal coding technology draws on the principles of fractal geometry and utilizes the self-similarity and recursiveness of fractal structures to efficiently compress and encrypt optical signals (Fig. 5). Quantum key distribution (QKD), based on the BB84 protocol, offers theoretically unbreakable key exchange via quantum mechanics but is limited by short transmission ranges and polarization decoherence (Fig. 6). Spread spectrum and optical code division multiple access enhance security through wideband spreading or orthogonal coding, though requiring high speed hardware and facing dynamic channel noise, while intelligent reflecting surfaces and artificial noise injection optimize signal phase and power in hybrid RF-FSO systems to boost legitimate links and degrade eavesdropper channels (Fig. 7). Adaptive enhancements, including dynamic power allocation and deep reinforcement learning, enable real time optimization of beam pointing and modulation, adapting to turbulent or underwater channel fluctuations, with anti-turbulence coding ensuring reliable transmission under varying conditions.

Hybrid OWC architectures, such as RF-FSO links (Fig. 8) and satellite-air-ground integration, utilize complementary technologies to address single system limitations, enhancing reliability and security by dynamically switching between high speed FSO and robust RF links or forming resilient global networks (Fig. 9). Collaborative mechanisms, including multi-hop relay and hybrid non-orthogonal multiple access, optimize resource allocation and increase attack complexity through signal path diversification. In practical applications, unmanned aerial vehicle secure communication employs laser-VLC hybrid links with OAM multiplexing and quantum noise encryption to resist jamming during dynamic aerial ground data exchange (Fig. 10). Military tactical systems utilize narrow beam FSO with chaotic encryption to avoid electromagnetic interference, integrating high energy lasers for both communication and directed energy countermeasures (Fig. 11). Finally, civil applications include VLC for secure indoor networks in smart homes and healthcare, vehicle-to-everything networks based on VLC and hybrid links technologies, as well as UOWC for underwater sensor networks using blue-green light and adaptive optics to mitigate water scattering (Fig. 12).

OWC has progressed significantly through physical layer innovations, hybrid system designs, and application specific security solutions, establishing a foundation for secure, high reliability communication. However, key challenges persist: QKD faces obstacles in long distance transmission and cost-effective hardware, chaotic encryption requires robust control in turbulent channels, and hybrid networks demand improved cross layer security optimization, particularly in complex scenarios like air-water cross domain communication.

Future research should prioritize three key directions to advance OWC technology. First, the integration of quantum security technologies, including advanced QKD protocols and quantum -resistant encryption into OWC systems, is essential for achieving absolute confidentiality. This necessitates the development of low-power, high -efficiency QKD modules compatible with both terrestrial and space -based OWC architectures. Second, the implementation of artificial intelligence and machine learning frameworks, particularly deep reinforcement learning for adaptive beamforming and neural network-based channel prediction, offers significant potential to enhance OWC resilience against dynamic eavesdropping and jamming threats. This approach enables real -time optimization of security parameters in response to environmental variations. Third, the expansion of OWC applications into emerging 6G and post 6G scenarios, such as ultra -reliable low -latency communication for autonomous vehicles, intelligent transportation systems, and distributed edge computing networks, requires interdisciplinary advancement across photonics, cryptography, and network engineering. Addressing these aspects will establish OWC as a foundation for next -generation networks, ensuring secure, efficient, and ubiquitous connectivity across terrestrial, aerial, and underwater domains, thus meeting the requirements of future intelligent systems.