International Conference on Communications, Signal Processing, and their Applications (ICCSPA)
Wasserstein Distributionally Robust Secrecy Rate Maximization for MISO Systems with Multiple Eavesdroppers
Hala Amer, Mahmoud H. Ismail, Mohamed S. Hassan
Abstract
This paper investigates the problem of secrecy rate maximization in a multiple-input single-output (MISO) system in the presence of multiple passive eavesdroppers with uncertain channel state information (CSI). Unlike conventional robust or stochastic approaches that rely on deterministic uncertainty sets or assumed parametric distributions, we adopt a Wasserstein distributionally robust optimization (DRO) framework to model the uncertainty in the eavesdropper channels. We formulate the secrecy rate maximization problem with artificial noise and derive a tractable semidefinite programming (SDP) reformulation. Simulation results demonstrate that the proposed method improves secrecy reliability compared to non-robust designs and norm-bounded worst-case approaches. In particular, the proposed framework yields superior average secrecy rates, lower outage probabilities, and improved lower-tail performance under eavesdropper CSI uncertainty.