Wireless Multi-Agent Generative AI: From Connected Intelligence to Collective Intelligence

TL;DR

The paper argues for wireless networks to adopt on-device multi-agent generative AI to achieve collective intelligence at the network edge, reducing reliance on cloud LLMs. It surveys architecture, planning/reasoning, MARL, and semantic communication as enabling technologies, and presents a case study with $K=4$ users, channel gains $g=(1.21,2.01,0.58,0.13)$, initial powers $p=(2,4,5,6)$ W, energy reduction target $\Delta p=0.85$ W, and rate thresholds $r=(3.50,15.80,4.40,1.00)$ kbps to illustrate potential gains. The work discusses challenges (e.g., TelecomLLM domain knowledge, hallucination, self-replication, and resource management) and outlines a research roadmap toward practical wireless collective intelligence. It envisions practical impact for 6G networks and edge devices through coordinated, autonomous, multi-agent planning and execution using on-device LLMs.

Abstract

The convergence of generative large language models (LLMs), edge networks, and multi-agent systems represents a groundbreaking synergy that holds immense promise for future wireless generations, harnessing the power of collective intelligence and paving the way for self-governed networks where intelligent decision-making happens right at the edge. This article puts the stepping-stone for incorporating multi-agent generative artificial intelligence (AI) in wireless networks, and sets the scene for realizing on-device LLMs, where multi-agent LLMs are collaboratively planning and solving tasks to achieve a number of network goals. We further investigate the profound limitations of cloud-based LLMs, and explore multi-agent LLMs from a game theoretic perspective, where agents collaboratively solve tasks in competitive environments. Moreover, we establish the underpinnings for the architecture design of wireless multi-agent generative AI systems at the network level and the agent level, and we identify the wireless technologies that are envisioned to play a key role in enabling on-device LLM. To demonstrate the promising potentials of wireless multi-agent generative AI networks, we highlight the benefits that can be achieved when implementing wireless generative agents in intent-based networking, and we provide a case study to showcase how on-device LLMs can contribute to solving network intents in a collaborative fashion. We finally shed lights on potential challenges and sketch a research roadmap towards realizing the vision of wireless collective intelligence.

Figures (5)

• Figure 1: Close-loop task planning, execution, optimization in wireless generative agents.
• Figure 2: Wireless generative agent network and device architecture.
• Figure 3: Playing repeated games in an example of power allocation on mobile users to minimize network power consumption.
• Figure 4: Users' transmission power given by LLMs, and power to maintain a minimum transmission rate v.s. round for reducing total power by $5\%$ at least.
• Figure 5: The difference of transmission rate w.r.t. minimum rate v.s. round for total power reduction by at least $5\%$.