Please use this identifier to cite or link to this item:
http://bura.brunel.ac.uk/handle/2438/31694
Title: | Convolutional Versus Large Language Models for Software Log Classification in Edge-Deployable Cellular Network Testing |
Authors: | Ihalage, A Taheri, S Muhammad, F Al-Raweshidy, H |
Keywords: | cellular networks;LLM;log analysis;machine learning;NLP |
Issue Date: | 8-Jul-2025 |
Publisher: | Institute of Electrical and Electronics Engineers (IEEE) |
Citation: | Ihalage, A. et al. (2025) 'Convolutional Versus Large Language Models for Software Log Classification in Edge-Deployable Cellular Network Testing', IEEE Access, 13, pp. 134283 - 134296. doi: 10.1109/access.2025.3587029. |
Abstract: | Software logs generated by sophisticated network emulators in the telecommunications industry, such as VIAVI TM500, are extremely complex, often comprising tens of thousands of text lines with minimal resemblance to natural language. Only specialised expert engineers can decipher such logs and troubleshoot defects in test runs. While AI offers a promising solution for automating defect triage, potentially leading to massive revenue savings for companies, state-of-the-art large language models (LLMs) suffer from significant drawbacks in this specialised domain. These include a constrained context window, limited applicability to text beyond natural language, and high inference costs. To address these limitations, we propose a compact convolutional neural network (CNN) architecture that offers a context window spanning up to 200,000 characters and achieves over 96% accuracy (F1>0.9) in classifying multifaceted software logs into various layers in the telecommunications protocol stack. Specifically, the proposed model is capable of identifying defects in test runs and triaging them to the relevant department, formerly a manual engineering process that required expert knowledge. We evaluate several LLMs; LLaMA2-7B, Mixtral_8 × 7B, Flan-T5, BERT and BigBird, and experimentally demonstrate their shortcomings in our specialized application. Despite being lightweight, our CNN achieves strong performance compared to LLM-based approaches in telecommunications log classification while minimizing the cost of production. Our defect triaging AI model is deployable on edge devices without dedicated hardware and is applicable across software logs in various industries. |
URI: | https://bura.brunel.ac.uk/handle/2438/31694 |
DOI: | https://doi.org/10.1109/ACCESS.2025.3587029 |
Appears in Collections: | Dept of Electronic and Electrical Engineering Research Papers |
Files in This Item:
File | Description | Size | Format | |
---|---|---|---|---|
FullText.pdf | Copyright © 2025 The Authors. Published by Institute of Electrical and Electronics Engineers (IEEE). This work is licensed under a Creative Commons Attribution 4.0 License. For more information, see https://creativecommons.org/licenses/by/4.0/ | 1.45 MB | Adobe PDF | View/Open |
This item is licensed under a Creative Commons License