Learning Network Sheaves for AI-native Semantic Communication

2025-12-02

AI Summary

To address severe semantic noise and weak task relevance when exchanging compressed latent representations among heterogeneous AI agents, this paper proposes a joint learning framework for AI-native semantic communication. Methodologically, it integrates semantic denoising with rate-distortion optimization, introducing an orthogonally constrained mapping layer to jointly learn communication topology and cross-agent alignment mappings. It further incorporates non-convex sparse dictionary learning with closed-form iterative updates to construct a shared global semantic space, simultaneously learning orthogonal alignment matrices and sparse latent representations. The key contribution is the first end-to-end co-optimization of semantic compression, denoising, and alignment—effectively mitigating semantic heterogeneity. Experiments on real-world image data demonstrate a favorable balance among multi-agent semantic clustering, enhanced interpretability, and high downstream task accuracy (exceeding 92% in classification).

Technology Category

• Search and Optimization: Learning to Search
• Multiagent Systems: Agent Communication
• Computer Vision: Representation Learning for Vision

Application Category

• Semantics and Knowledge: Methods to enhance, augment, integrate or synergize semantic models such as knowledge graphs and LLMs
• Search and Retrieval-Augmented AI: Agentic search
• Graph Algorithms and Modeling for the Web: Graph embeddings and representation learning for Web-related graphs

Abstract

Recent advances in AI call for a paradigm shift from bit-centric communication to goal- and semantics-oriented architectures, paving the way for AI-native 6G networks. In this context, we address a key open challenge: enabling heterogeneous AI agents to exchange compressed latent-space representations while mitigating semantic noise and preserving task-relevant meaning. We cast this challenge as learning both the communication topology and the alignment maps that govern information exchange among agents, yielding a learned network sheaf equipped with orthogonal maps. This learning process is further supported by a semantic denoising end compression module that constructs a shared global semantic space and derives sparse, structured representations of each agent's latent space. This corresponds to a nonconvex dictionary learning problem solved iteratively with closed-form updates. Experiments with mutiple AI agents pre-trained on real image data show that the semantic denoising and compression facilitates AI agents alignment and the extraction of semantic clusters, while preserving high accuracy in downstream task. The resulting communication network provides new insights about semantic heterogeneity across agents, highlighting the interpretability of our methodology.

Problem

• Enabling heterogeneous AI agents to exchange compressed latent-space representations
• Mitigating semantic noise and preserving task-relevant meaning in communication
• Learning communication topology and alignment maps for AI-native networks

Innovation

• Learning network sheaves with orthogonal maps
• Semantic denoising and compression module
• Nonconvex dictionary learning with closed-form updates