Abstract:Generative artificial intelligence (GenAI) and communication networks are expected to have groundbreaking synergies in 6G. Connecting GenAI agents over a wireless network can potentially unleash the power of collective intelligence and pave the way for artificial general intelligence (AGI). However, current wireless networks are designed as a "data pipe" and are not suited to accommodate and leverage the power of GenAI. In this paper, we propose the GenAINet framework in which distributed GenAI agents communicate knowledge (high-level concepts or abstracts) to accomplish arbitrary tasks. We first provide a network architecture integrating GenAI capabilities to manage both network protocols and applications. Building on this, we investigate effective communication and reasoning problems by proposing a semantic-native GenAINet. Specifically, GenAI agents extract semantic concepts from multi-modal raw data, build a knowledgebase representing their semantic relations, which is retrieved by GenAI models for planning and reasoning. Under this paradigm, an agent can learn fast from other agents' experience for making better decisions with efficient communications. Furthermore, we conduct two case studies where in wireless device query, we show that extracting and transferring knowledge can improve query accuracy with reduced communication; and in wireless power control, we show that distributed agents can improve decisions via collaborative reasoning. Finally, we address that developing a hierarchical semantic level Telecom world model is a key path towards network of collective intelligence.

What problem does this paper attempt to address?

The paper aims to address the following issues: 1. **Incompatibility of current wireless networks with Generative Artificial Intelligence (GenAI)**: The current wireless network design is primarily for data transmission and does not consider supporting large-scale AI device deployments, especially large AI models, which are very costly in terms of communication and computation. 2. **How to efficiently embed Large Language Models (LLMs) into resource-constrained devices**: Although there are techniques like compression and accelerated inference, these methods may reduce the generalization ability of the models. 3. **Communication efficiency in multi-agent collaboration frameworks**: To achieve efficient collaboration among multiple agents, a new communication paradigm is needed to reduce communication, computation, and storage costs. The paper proposes a new framework called GenAINet, which enables distributed GenAI agents to communicate knowledge (such as concepts or abstract information) to perform effective reasoning and solve tasks in specific applications (such as vehicular networks, smart grids, and the Internet of Things). Specifically, the GenAINet framework leverages LLMs to extract semantic concepts from multimodal raw data and build a knowledge base for planning and reasoning. Additionally, agents can quickly learn from each other's experiences, making better decisions based on efficient communication. Through two case studies (wireless device querying and wireless power control), the effectiveness of knowledge transfer and collaborative reasoning is demonstrated, proving that this approach not only improves task completion accuracy but also reduces communication overhead. Finally, the paper discusses constructing a hierarchical semantic-level telecommunication world model as a key path to achieving collective intelligent networks.

GenAINet: Enabling Wireless Collective Intelligence via Knowledge Transfer and Reasoning