Large-Model-Driven Intelligent IoT Decision Systems
DOI:
https://doi.org/10.54097/92e9by14Keywords:
Internet of Things (IoT); Large Language Models (LLMs); Multimodal Semantic Fusion; Reinforcement Learning; Federated Learning.Abstract
With the development of Internet of things (IoT) technology, the continuous increase of IoT devices and the increasing complexity of application scenarios, the traditional decision system relying on rules has been difficult to meet the intelligent requirements of high dynamic, multi-modal, and global collaboration. In recent years, Large Language Models (LLM) have gradually become an important technical support for building a new generation of IoT decision systems with excellent semantic understanding, cross-modal reasoning, and contextual learning capabilities. This paper reviews the method implementation of large models in IoT decision-making systems, with a focus on elaborating key technical paths such as multimodal input fusion, semantic reasoning and decision generation, reinforcement learning collaboration, edge-cloud collaboration and federated learning, and analyzes the current engineering challenges and research directions. This article emphasizes that the IoT decision making system driven by large models not only promotes the transformation of the traditional IoT from a "data-driven" to a "semantic-driven" approach, providing a feasible path for the practicality of large models in real IoT scenarios, but also offers a technical foundation and research direction for the future development of fields such as intelligent manufacturing and smart cities.
Downloads
References
[1] Touvron H, Lavril T, Izacard G, et al. Llama: Open and efficient foundation language models. 2023. https://arxiv.org/abs/2302.13971.
[2] Achiam J, Adler S, Agarwal S, et al. GPT-4 technical report. 2023. https://arxiv.org/abs/2303.08774.
[3] Singh R, Gill S S. Edge AI: a survey. Internet of Things and Cyber-Physical Systems, 2023, 3: 71-92.
[4] Vaswani A, Shazeer N, Parmar N, et al. Attention is all you need//Proceedings of the 31st International Conference on Neural Information Processing Systems. Red Hook, NY, USA: Curran Associates Inc., 2017: 6000-6010.
[5] Radford A, Kim J W, Hallacy C, et al. Learning transferable visual models from natural language supervision//Proceedings of the 38th International Conference on Machine Learning. PMLR, 2021, 139: 8748-8763.
[6] Fangfang. Research on power load forecasting based on Improved BP neural network. Harbin: Harbin Institute of Technology, 2011.
[7] DeRieux A C, Saad W, Zuo W, et al. A transformer framework for data fusion and multi-task learning in smart cities. 2022. https://arxiv.org/abs/2211.10506.
[8] Wei J, Wang X, Schuurmans D, et al. Chain-of-thought prompting elicits reasoning in large language models. Advances in Neural Information Processing Systems, 2022, 35: 24824-24837.
[9] Hu E J, Shen Y, Wallis P, et al. Lora: Low-rank adaptation of large language models[J]. ICLR, 2022.
[10] McMahan B, Moore E, Ramage D, et al. Communication-efficient learning of deep networks from decentralized data//Proceedings of the 20th International Conference on Artificial Intelligence and Statistics. Fort Lauderdale, FL, USA: PMLR, 2017: 1273-1282.
[11] Sanh V, Debut L, Chaumond J, et al. DistilBERT, a distilled version of BERT: smaller, faster, cheaper and lighter. 2019. https://arxiv.org/abs/1910.01108.
Downloads
Published
Issue
Section
License
Copyright (c) 2026 Frontiers in Computing and Intelligent Systems
This work is licensed under a Creative Commons Attribution-NonCommercial 4.0 International License.
How to Cite
