Answers are grounded in this paper and cite their sources.
Hyperspectral image (HSI) transmission over wireless networks is constrained by the large data volume of spectral cubes, stringent delay requirements, and the risk of information leakage caused by passive eavesdropping. These challenges become more pronounced when the eavesdropper channel state information (CSI) is imperfect and communication resources must be dynamically allocated. This article presents a secure STAR-RIS-assisted HSI transmission framework that jointly considers adaptive HSI payload control, queue evolution, transmission latency, energy consumption, and secrecy outage under imperfect cascaded eavesdropper CSI. The resulting resource allocation problem is formulated as a long-term stochastic optimization problem involving active beamforming, artificial-noise transmission, STAR-RIS transmission/reflection coefficients, and adaptive HSI payload ratios. To efficiently handle the coupled optimization variables and time-varying network dynamics, a Lyapunov-guided twin delayed deep deterministic policy gradient (LG-TD3) algorithm is developed, where Lyapunov optimization converts the long-term objective into queue-aware per-slot optimization, while TD3 learns an effective continuous control policy. A sample-based secrecy outage evaluation mechanism is further incorporated to provide robust secure transmission without requiring perfect knowledge of the eavesdropper channels. Simulation results demonstrate that the proposed approach converges faster than representative benchmark schemes while achieving higher secure HSI throughput, lower secrecy outage probability, reduced delivery latency, improved energy efficiency, and more effective adaptive payload control. These results demonstrate that the proposed framework provides an efficient and robust solution for secure HSI transmission in dynamic STAR-RIS-assisted wireless environments.
No takes yet. Share an insight, caveat, or question.
Waqas et al. (2026) studied this question.
Synapse has enriched 10 related papers. Consider them for comparative context: