Akademik Veri Yönetim Sistemi
IEEE Transactions on Consumer Electronics, 2026 (SCI-Expanded, Scopus)
This paper studies an intelligent reflecting surface (IRS)-assisted secure downlink vehicle-to-infrastructure (V2I) communication system, where a multi-antenna roadside unit (RSU) serves legitimate vehicles while multiple passive eavesdropping vehicles are present. By exploiting both the direct and IRS-reflected links, we formulate a secrecy sum-rate maximization problem through the joint design of the RSU beamforming vectors and the IRS phase shifts, subject to the RSU transmit power constraint and the unit-modulus IRS hardware limitations. To tackle the resulting non-convex problem, we develop an efficient alternating-optimization framework that iteratively updates the RSU beamforming and IRS configuration via tractable convex surrogates and unit-modulus projection. Numerical results verify the fast convergence of the proposed algorithm and demonstrate significant secrecy performance improvements over benchmark schemes, including random IRS and no-IRS transmission. The results further show that the secrecy gains increase with the number of IRS elements, while they decrease as the number of eavesdroppers grows due to the worst-case interception effect.