Securing the Road Ahead: Physical Layer Security for Connected Cars—Architectures, Protocols, and Future Directions
Abstract
The wireless vehicle-to-everything (V2X) communication is open, which leaves connected cars susceptible to eavesdropping, jamming, spoofing, and signal level manipulation. This paper introduces and develops a MIMO-based physical layer security architecture for connected vehicular networks (C-VN), which combines secure beamforming, artificial-noise injection, channel state aware power allocation, and secrecy rate oriented transmission design. In contrast to vehicular security systems based solely on cryptography, the proposed architecture exploits spatial diversity, channel randomness, antenna correlation, and rapidly varying vehicular channels to enhance physical layer confidentiality. The system model is used to derive the legitimate link SNR, eavesdropper SINR, achievable secrecy rate, and secrecy outage probability for dynamic V2X scenarios. A simulation methodology is developed to evaluate the proposed architecture under controlled channel realizations using appropriate baseline schemes, imperfect channel state information, and Monte Carlo trials. The simulation based validation examines average secrecy rate, secrecy outage probability, legitimate link BER, the effect of artificial noise power allocation, and computational complexity. Practical aspects such as CSI aging, Doppler shifts, synchronization mismatches, RF-chain constraints, antenna scaling, and vehicular hardware limitation are also considered. The results demonstrate the security performance trade-offs of the proposed MIMO-based PLS architecture and show that artificial noise assisted secure transmission can improve secrecy performance while maintaining acceptable legitimate link reliability. The empirical complexity analysis further indicates the computational scaling associated with beamformer and null space processing and supports the use of RSU/MEC-assisted computation for computationally constrained vehicular platforms. The study also provides a practical mapping of vehicular attack cases from real world scenarios to upper layer and PLS countermeasures and discusses the applicability and limitations of PLS in connected car environments. Broader empirical validation using standardized vehicular channel models and hardware platforms remains an important direction for future work