Next-gen telecom AI: mastering prompt engineering for innovation
Вантажиться...
Файли
Дата
2025
Автори
Науковий керівник
Назва журналу
Номер ISSN
Назва тому
Видавець
National Technical University of Ukraine "Igor Sikorsky Kyiv Polytechnic Institute"
Анотація
Background. Since 2021, prompt engineering has emerged as a cornerstone of artificial intelligence (AI), revolutionising telecommunications by 2023 through optimised large language models (LLMs). Objective. This review synthesises existing research to evaluate prompt engineering’s transformative role in telecommunications, emphasising practical applications, technical challenges, and future directions. Methods. This analysis draws on 2021–2025 literature from 31 sources, including IEEE journals, ACM Transactions on Information Systems, NeurIPS proceedings, and arXiv preprints, examining prompt engineering techniques like few-shot learning, chain-of-thought prompting, multi-step prompting, Named Entity Recognition (NER), Retrieval-Augmented Generation (RAG) and more, with a telecom focus (6G and hypothesised 5G applications) contextualized within Natural Language Processing (NLP) advancements. Results. Although research on prompt engineering specifically for 5G telecommunications is currently limited, it presents substantial opportunities for optimising network performance, diagnostics, documentation handling, enhancing customer support, and driving innovation across both 5G and future 6G networks. Conclusions. Prompt engineering bridges AI capabilities with telecommunications needs, with techniques like NER and RAG contributing to the enhancement of mobile communications. The dearth of 5G-specific research highlights the urgent need for specialised LLMs in telecommunications and automated prompting to advance solutions for 5G and 6G.
Опис
Ключові слова
prompt engineering, large language models, telecommunications, network management, 5G, 6G, few-shot learning, chain-of-thought prompting, multi-step prompting, NER, RAG, інженерія запитів, великі мовні моделі, телекомунікації, управління мережами, навчання за кількома прикладами, ланцюжкові запити, багатокрокові запити, розпізнавання іменованих сутностей (NER), генерація з доповненням через пошук (RAG)
Бібліографічний опис
Ilchenko, M. V. Next-gen telecom AI: mastering prompt engineering for innovation / Mykhailo V. Ilchenko // Information and telecommunication sciences : international research journal. – 2025. – Vol. 16. – N. 1. – Pp. 22-29. – Bibliogr.: 31 ref.