Generative AI Meets 6G and Beyond: Diffusion Models for Semantic Communications

Semantic communications mark a paradigm shift from bit-accurate transmission toward meaning-centric communication, essential as wireless systems approach theoretical capacity limits. The emergence of generative AI has catalyzed generative semantic communications, where receivers reconstruct content from minimal semantic cues by leveraging learned priors. Among generative approaches, diffusion models stand out for their superior generation quality, stable training dynamics, and rigorous theoretical foundations. However, the field currently lacks systematic guidance connecting diffusion techniques to communication system design, forcing researchers to navigate disparate literatures. This article provides the first comprehensive tutorial on diffusion models for generative semantic communications. We present score-based diffusion foundations and systematically review three technical pillars: conditional diffusion for controllable generation, efficient diffusion for accelerated inference, and generalized diffusion for cross-domain adaptation. In addition, we introduce an inverse problem perspective that reformulates semantic decoding as posterior inference, bridging semantic communications with computational imaging. Through analysis of human-centric, machine-centric, and agent-centric scenarios, we illustrate how diffusion models enable extreme compression while maintaining semantic fidelity and robustness. By bridging generative AI innovations with communication system design, this article aims to establish diffusion models as foundational components of next-generation wireless networks and beyond.

• Publication year

  2025

• Venue

  arXiv.org

• Publication date

  2025-11-11

• Fields of study

  Mathematics, Computer Science, Engineering

• Identifiers
  DOI 10.48550/arXiv.2511.08416arXiv 2511.08416
• External record

  Open on Semantic Scholar

• Source metadata

  Semantic Scholar

• No claims are published for this paper.

• No concepts are published for this paper.

• LRGD: Low-Rank Guided Diffusion for Robust Image Transmission in Semantic Communication

  2026cited by this paper
• Intellicise Wireless Networks From Semantic Communications: A Survey, Research Issues, and Challenges

  2025cited by this paper
• Advanced Architectures Integrated with Agentic AI for Next-Generation Wireless Networks

  2025cited by this paper
• Reconstruction vs. Generation: Taming Optimization Dilemma in Latent Diffusion Models

  2025cited by this paper
• The Principles of Diffusion Models

  2025cited by this paper
• Compute Only 16 Tokens in One Timestep: Accelerating Diffusion Transformers with Cluster-Driven Feature Caching

  2025cited by this paper
• Generative Diffusion Models for Wireless Networks: Fundamental, Architecture, and State-of-the-Art

  2025cited by this paper
• Neural Coding Is Not Always Semantic: Toward the Standardized Coding Workflow in Semantic Communications

  2025influential reference
• Mean Flows for One-step Generative Modeling

  2025cited by this paper
• PICD: Versatile Perceptual Image Compression with Diffusion Rendering

  2025cited by this paper
• AdaDiff: Adaptive Step Selection for Fast Diffusion Models

  2025cited by this paper
• Unified Multimodal Discrete Diffusion

  2025cited by this paper
• Towards Better Alignment: Training Diffusion Models with Reinforcement Learning Against Sparse Rewards

  2025cited by this paper
• A Contemporary Survey on Semantic Communications: Theory of Mind, Generative AI, and Deep Joint Source-Channel Coding

  2025cited by this paper
• Direct Discriminative Optimization: Your Likelihood-Based Visual Generative Model is Secretly a GAN Discriminator

  2025cited by this paper
• Show-o: One Single Transformer to Unify Multimodal Understanding and Generation

  2024cited by this paper
• Transfusion: Predict the Next Token and Diffuse Images with One Multi-Modal Model

  2024cited by this paper
• DiffSG: A Generative Solver for Network Optimization with Diffusion Model

  2024cited by this paper
• Diffusion-Driven Semantic Communication for Generative Models With Bandwidth Constraints

  2024cited by this paper
• Agent-Driven Generative Semantic Communication With Cross-Modality and Prediction

  2024cited by this paper
• Improving Diffusion Inverse Problem Solving with Decoupled Noise Annealing

  2024cited by this paper
• Diffusion Forcing: Next-token Prediction Meets Full-Sequence Diffusion

  2024cited by this paper
• Channel-Blind Joint Source–Channel Coding for Wireless Image Transmission

  2024cited by this paper
• DiffCom: Channel Received Signal Is a Natural Condition to Guide Diffusion Posterior Sampling

  2024influential reference
• Latent Diffusion Model-Enabled Low-Latency Semantic Communication in the Presence of Semantic Ambiguities and Wireless Channel Noises

  2024cited by this paper
• Deep Generative Modeling Reshapes Compression and Transmission: From Efficiency to Resiliency

  2024cited by this paper
• Learning-to-Cache: Accelerating Diffusion Transformer via Layer Caching

  2024cited by this paper
• Plug-and-Play Diffusion Distillation

  2024cited by this paper
• Improved Distribution Matching Distillation for Fast Image Synthesis

  2024cited by this paper
• PeRFlow: Piecewise Rectified Flow as Universal Plug-and-Play Accelerator

  2024cited by this paper
• Semantic regularization of electromagnetic inverse problems

  2024cited by this paper
• A Survey on Semantic Communication Networks: Architecture, Security, and Privacy

  2024cited by this paper
• Diffusion-Aided Joint Source Channel Coding for High Realism Wireless Image Transmission

  2024cited by this paper
• Align Your Steps: Optimizing Sampling Schedules in Diffusion Models

  2024cited by this paper
• LD-Pruner: Efficient Pruning of Latent Diffusion Models using Task-Agnostic Insights

  2024cited by this paper
• What makes an image realistic?

  2024cited by this paper
• PixArt-Σ: Weak-to-Strong Training of Diffusion Transformer for 4K Text-to-Image Generation

  2024cited by this paper
• Accelerating Diffusion Sampling with Optimized Time Steps

  2024cited by this paper
• Wavelet-Guided Acceleration of Text Inversion in Diffusion-Based Image Editing

  2024cited by this paper
• Idempotence and Perceptual Image Compression

  2024cited by this paper
• S2-DMs: Skip-Step Diffusion Models

  2024cited by this paper
• Generative AI Meets Semantic Communication: Evolution and Revolution of Communication Tasks

  2024cited by this paper
• Generative Semantic Communication: Architectures, Technologies, and Applications

  2024cited by this paper
• Autoregressive Models in Vision: A Survey

  2024cited by this paper
• DiP-GO: A Diffusion Pruner via Few-step Gradient Optimization

  2024cited by this paper
• T2V-Turbo-v2: Enhancing Video Generation Model Post-Training through Data, Reward, and Conditional Guidance Design

  2024cited by this paper
• SITCOM: Step-wise Triple-Consistent Diffusion Sampling for Inverse Problems

  2024cited by this paper
• Lightweight Diffusion Models for Resource-Constrained Semantic Communication

  2024cited by this paper
• ImageFolder: Autoregressive Image Generation with Folded Tokens

  2024cited by this paper
• A Survey on Diffusion Models for Inverse Problems

  2024cited by this paper
• Language-Oriented Semantic Communication for Image Transmission with Fine-Tuned Diffusion Model

  2024cited by this paper
• MonoFormer: One Transformer for Both Diffusion and Autoregression

  2024cited by this paper
• Diffusion-ES: Gradient-Free Planning with Diffusion for Autonomous and Instruction-Guided Driving

  2024cited by this paper
• Rate-Adaptive Generative Semantic Communication Using Conditional Diffusion Models

  2024cited by this paper
• P2P-Bridge: Diffusion Bridges for 3D Point Cloud Denoising

  2024cited by this paper
• DiffusionE: Reasoning on Knowledge Graphs via Diffusion-based Graph Neural Networks

  2024cited by this paper
• Consistency Trajectory Models: Learning Probability Flow ODE Trajectory of Diffusion

  2023cited by this paper
• GLIGEN: Open-Set Grounded Text-to-Image Generation

  2023cited by this paper
• BLIP-2: Bootstrapping Language-Image Pre-training with Frozen Image Encoders and Large Language Models

  2023cited by this paper
• Flow Matching on General Geometries

  2023cited by this paper
• Adding Conditional Control to Text-to-Image Diffusion Models

  2023cited by this paper
• T2I-Adapter: Learning Adapters to Dig out More Controllable Ability for Text-to-Image Diffusion Models

  2023cited by this paper
• I2SB: Image-to-Image Schrödinger Bridge

  2023cited by this paper
• Reduce, Reuse, Recycle: Compositional Generation with Energy-Based Diffusion Models and MCMC

  2023cited by this paper
• Consistency Models

  2023cited by this paper
• Diffusion policy: Visuomotor policy learning via action diffusion

  2023cited by this paper
• FreeDoM: Training-Free Energy-Guided Conditional Diffusion Model

  2023cited by this paper
• Masked Diffusion Transformer is a Strong Image Synthesizer

  2023cited by this paper
• Continual Diffusion: Continual Customization of Text-to-Image Diffusion with C-LoRA

  2023cited by this paper
• Shap-E: Generating Conditional 3D Implicit Functions

  2023cited by this paper
• A Variational Perspective on Solving Inverse Problems with Diffusion Models

  2023cited by this paper
• Structural Pruning for Diffusion Models

  2023cited by this paper
• Training Diffusion Models with Reinforcement Learning

  2023cited by this paper
• SnapFusion: Text-to-Image Diffusion Model on Mobile Devices within Two Seconds

  2023cited by this paper
• MotionDiffuser: Controllable Multi-Agent Motion Prediction Using Diffusion

  2023cited by this paper
• Generative Semantic Communication: Diffusion Models Beyond Bit Recovery

  2023cited by this paper
• Pseudoinverse-Guided Diffusion Models for Inverse Problems

  2023cited by this paper
• Solving Linear Inverse Problems Provably via Posterior Sampling with Latent Diffusion Models

  2023cited by this paper
• AnimateDiff: Animate Your Personalized Text-to-Image Diffusion Models without Specific Tuning

  2023cited by this paper
• Improving and generalizing flow-based generative models with minibatch optimal transport

  2023cited by this paper
• Enhancing Deep Reinforcement Learning: A Tutorial on Generative Diffusion Models in Network Optimization

  2023cited by this paper
• SwinJSCC: Taming Swin Transformer for Deep Joint Source-Channel Coding

  2023cited by this paper
• InstaFlow: One Step is Enough for High-Quality Diffusion-Based Text-to-Image Generation

  2023cited by this paper
• CDDM: Channel Denoising Diffusion Models for Wireless Semantic Communications

  2023cited by this paper
• LLM-grounded Video Diffusion Models

  2023cited by this paper
• CommIN: Semantic Image Communications as an Inverse Problem with INN-Guided Diffusion Models

  2023cited by this paper
• Latent Consistency Models: Synthesizing High-Resolution Images with Few-Step Inference

  2023cited by this paper
• Towards image compression with perfect realism at ultra-low bitrates

  2023cited by this paper
• SEINE: Short-to-Long Video Diffusion Model for Generative Transition and Prediction

  2023cited by this paper
• Deep Equilibrium Diffusion Restoration with Parallel Sampling

  2023cited by this paper
• One-Step Diffusion with Distribution Matching Distillation

  2023cited by this paper
• DeepCache: Accelerating Diffusion Models for Free

  2023cited by this paper
• Cache Me if You Can: Accelerating Diffusion Models through Block Caching

  2023cited by this paper
• LMD: Faster Image Reconstruction with Latent Masking Diffusion

  2023cited by this paper
• Reasons for the Superiority of Stochastic Estimators over Deterministic Ones: Robustness, Consistency and Perceptual Quality

  2022cited by this paper
• Refining Generative Process with Discriminator Guidance in Score-based Diffusion Models

  2022cited by this paper
• Wavelet Diffusion Models are fast and scalable Image Generators

  2022cited by this paper
• Less Data, More Knowledge: Building Next-Generation Semantic Communication Networks

  2022cited by this paper
• Generative Joint Source-Channel Coding for Semantic Image Transmission

  2022cited by this paper
• Parallel Diffusion Models of Operator and Image for Blind Inverse Problems

  2022cited by this paper

• No citing papers are available for this paper.