Optimizing Explainable AI for 5G-Enabled Autonomous Systems: Enhancing Performance and Transparency

The growing use of the 5G technology in safety-critical autonomous systems requires the existence of high-performing and transparent artificial intelligence models. Despite high predictive accuracy with deep learning methods, they are not suitable due to low levels of trust, accountability, and real-time application in autonomous vehicles and healthcare diagnostics because they are not interpretable. This study will deal with this challenge by proposing optimized Explainable Artificial Intelligence (XAI) framework to enable autonomous systems that use 5G. The model is based on the combination of intrinsically interpretable neural architectures with model-agnostic explanation methods and optimization approaches to achieve a trade-off between the fidelity of explanations and computational efficiency. The performance is evaluated by the use of publicly accessible autonomous driving and healthcare data and it is measured with the accuracy, latency, clarity of explanation, and user-trust metrics. The statistical analysis shows that the framework proposed enhances the ease of explanation by about 30 percent and also halves decision time at 5G conditions and does not cause a drop in predictive accuracy. These findings underscore the potential to realize a combination of transparency and performance in real-time and mission critical autonomous systems in next generation wireless settings, and the proposed XAI framework is applicable in such environments.