article · IEEE Open Journal of the Communications Society
Deploying large language models on edge devices introduces decentralised intelligence directly into local environments. Realising this potential requires addressing significant constraints related to system architecture, resource limits, security, and responsible implementation. Current developments focus on tailored optimisation and autonomy techniques that enable models to operate effectively under resource-constrained edge settings. In addition, implementing these models at the edge introduces unique vulnerabilities regarding data confidentiality and system integrity, necessitating specialised defence mechanisms. A broad range of practical applications across diverse domains can benefit from decentralised language capabilities, provided systems follow structured design decisions. Establishing guidelines, best practices, and trustworthy development principles is essential for managing ethical challenges and deploying these architectures securely across practical edge environments.
Bringing large language models directly to edge devices reduces reliance on central servers and enables intelligent processing locally. However, edge hardware has limited computing capacity and faces distinct security risks. Understanding how to optimise these systems while safeguarding data privacy and integrity allows developers and organisations to build trustworthy, privacy-preserving decentralised services.
The work addresses decentralised intelligence applicable to sectors adopting edge computing and local language processing. Potential users include system architects, software developers, and technology providers deploying intelligence on resource-limited hardware. As a broad review of architectures, optimisation strategies, and security defences, this work sits at an early-stage review level rather than demonstrating a specific commercial product or near-market solution.
AI-generated from the published abstract. Always read the original work before citing.
The integration of Large Language Models (LLMs) and Edge Intelligence (EI) introduces a groundbreaking paradigm for intelligent edge devices. With their capacity for human-like language processing and generation, LLMs empower edge computing with a powerful set of tools, paving the way for a new era of decentralized intelligence. Yet, a notable research gap exists in obtaining a thorough comprehension of LLM-based EI architectures, which should incorporate crucial elements such as security, optimization, and responsible development. This survey aims to bridge this gap by providing a comprehensive resource for both researchers and practitioners. We explore LLM-based EI architectures in-depth, carefully analyzing state-of-the-art paradigms and design decisions. To facilitate efficient and scalable edge deployments, we perform a comparative analysis of recent optimization and autonomy techniques specifically designed for resource-constrained edge environments. Additionally, we shed light on the extensive potential of LLM-based EI by demonstrating its varied practical applications across a wide range of domains. Acknowledging the utmost importance of security, our survey thoroughly investigates potential vulnerabilities inherent in LLM-based EI deployments. We explore corresponding defense mechanisms to protect the integrity and confidentiality of data processed at the edge. In conclusion, highlighting the essential aspect of trustworthiness, we outline best practices and guiding principles for the responsible development and deployment of these systems. By conducting a comprehensive review of these key components, our survey aims to support the ethical development and strategic implementation of LLM-driven EI, paving the way for its transformative impact on diverse applications.
This page summarises published work. The authoritative version sits with the publisher.
DOI: 10.1109/ojcoms.2024.3456549
Is something wrong with this record? Report it or request removal.
Discussion
Have you built on this work, tried to replicate it, or seen it applied in practice? Share what you know. Verified researchers and MARATTO™ domain experts can open a discussion, and any member can reply. Contributions are reviewed before they appear.
No discussion yet. Open the first thread.
New to MARATTO™? Create a free account.
