Abstract:The recent success of large language models (LLMs) has spurred their application in various fields. In particular, there have been efforts to integrate LLMs into various aspects of wireless communication systems. The use of LLMs in wireless communication systems has the potential to realize artificial general intelligence (AGI)-enabled wireless networks. In this paper, we investigate an LLM-based resource allocation scheme for wireless communication systems. Specifically, we formulate a simple resource allocation problem involving two transmit pairs and develop an LLM-based resource allocation approach that aims to maximize either energy efficiency or spectral efficiency. Additionally, we consider the joint use of low-complexity resource allocation techniques to compensate for the reliability shortcomings of the LLM-based scheme. After confirming the applicability and feasibility of LLM-based resource allocation, we address several key technical challenges that remain in applying LLMs in practice.

What problem does this paper attempt to address?

### Problems Addressed by the Paper This paper aims to explore how to utilize large language models (LLM) to achieve resource allocation in wireless communication systems. Specifically, the paper proposes an LLM-based resource allocation scheme to maximize energy efficiency or spectrum efficiency. Additionally, the paper considers combining low-complexity resource allocation techniques to compensate for the reliability shortcomings of the LLM scheme. ### Main Content 1. **Background Introduction**: - Resource allocation in wireless communication systems (such as transmission power, bandwidth, beamforming, etc.) is crucial because the openness of the wireless medium can lead to interference with neighboring nodes. - As the number of transmission nodes increases and communication goals diversify, resource allocation becomes more complex. - Traditional optimization methods (such as convex optimization) are effective but face challenges when dealing with large-scale nodes and dynamically changing wireless environments. - Deep learning (DL) methods perform well in resource allocation but require designing and training specific deep neural networks (DNN) for each task, limiting their applicability. 2. **Application of LLM in Resource Allocation**: - LLM has strong reasoning and understanding capabilities and can serve as a general optimizer to solve various resource allocation problems. - LLM can determine resource allocation strategies through few-shot learning without retraining for specific scenarios. - LLM can handle multimodal data, such as geographic information, enabling more efficient and proactive resource allocation. 3. **System Model and Resource Allocation Strategy**: - The paper considers a simplified resource allocation problem involving 2 transmission pairs. - By adjusting transmission power, the goal is to maximize total spectrum efficiency (SE) or energy efficiency (EE). - An LLM-based resource allocation scheme is proposed, and its feasibility is validated through simulations. 4. **Performance Evaluation**: - The performance of different LLM models in resource allocation is evaluated through simulation experiments. - Results show that as the number of training samples increases, the performance of the LLM method gradually approaches optimal performance. - Combining traditional binary transmission power control methods can further improve the reliability and performance of the LLM method. 5. **Research Challenges**: - **Latency and Computation Time**: The large number of parameters in LLM requires substantial computational resources during operation, potentially leading to long delays. Developing small LLM (sLLM) and edge AI technology is a future research direction. - **Optimizing LLM Architecture**: To improve performance, it is necessary to optimize the LLM architecture, including selecting appropriate pre-trained models and adjusting model parameters. ### Summary This paper addresses the resource allocation problem in wireless communication systems by introducing LLM, proposing an LLM-based resource allocation scheme, and validating its feasibility and effectiveness through simulation experiments. Although LLM performs well in resource allocation, it still faces challenges such as latency, computation time, and model architecture optimization.

LLM-Empowered Resource Allocation in Wireless Communications Systems

WirelessLLM: Empowering Large Language Models Towards Wireless Intelligence

Joint URLLC Traffic Scheduling and Resource Allocation for Semantic Communication Systems

Energy-Efficient Resource Block Allocation for Licensed-Assisted Access

Resource Management for Millimeter-Wave Ultra-Reliable and Low-Latency Communications

Accelerating Resource Allocation for D2D Communications Using Imitation Learning

Deep-Learning-Based Wireless Resource Allocation With Application to Vehicular Networks

Large Language Model Enabled Multi-Task Physical Layer Network

Tradeoff Between Co-Channel Interference and Collision Probability in LAA Systems

Large Multi-Modal Models (LMMs) as Universal Foundation Models for AI-Native Wireless Systems

The Power of Large Language Models for Wireless Communication System Development: A Case Study on FPGA Platforms

Learning to Branch: Accelerating Resource Allocation in Wireless Networks

Large Language Model (LLM) for Telecommunications: A Comprehensive Survey on Principles, Key Techniques, and Opportunities

Machine Intelligence at the Edge With Learning Centric Power Allocation

Large Language Models (LLMs) Assisted Wireless Network Deployment in Urban Settings

Large Language Model (LLM)-enabled In-context Learning for Wireless Network Optimization: A Case Study of Power Control

Large Language Model Enabled Semantic Communication Systems

LLM-Slice: Dedicated Wireless Network Slicing for Large Language Models

A Practical Resource Allocation Approach for Interference Management in LTE Uplink Transmission.

Large Language Models for Power Scheduling: A User-Centric Approach