GAN-Enhanced Deep Reinforcement Learning for Semantic-Aware Resource Allocation in 6G Network Slicing

Collection of recent arXiv papers (Apr 2026) covering wireless/mobile systems, edge/kernel ML, cloud virtual networking, LLM-enabled control, post-quantum key exchange, and quantum network control. Notable items with security relevance: (1) GAN-DDPG for semantic-aware 6G resource allocation — improves spectral/latency metrics but increases dependence on ML-generated traffic models; (2) eBandit — moves adaptive-bitrate selection and a bandit learner into the Linux kernel via eBPF/sockops, yielding a new privileged attack surface if exploited or misconfigured; (3) HAP (high-altitude platform) ML