LLM-Based Network Automation for Soho Networks

Noor, Muhammad Rafi Muhtaddin (2026) LLM-Based Network Automation for Soho Networks. Undergraduate thesis, UPN Veteran Jawa Timur.

	Text (Cover) 22081010201.-cover.pdf Download (1MB)
	Text (Bab 1) 22081010201.-bab1.pdf Download (138kB)
	Text (Bab 2) 22081010201.-bab2.pdf Restricted to Repository staff only until 22 July 2029. Download (357kB)
	Text (Bab 3) 22081010201.-bab3.pdf Restricted to Repository staff only until 22 July 2029. Download (1MB)
	Text (Bab 4) Bab 4.pdf Restricted to Repository staff only until 22 July 2029. Download (11MB)
	Text (Bab 5) 22081010201.-bab5.pdf Download (137kB)
	Text (Daftar Pustaka) 22081010201.-daftarpustaka.pdf Download (187kB)
	Text (Lampiran) 22081010201.-lampiran.pdf Restricted to Repository staff only Download (141kB)

Abstract

Managing network infrastructure in Small Office/Home Office (SOHO) environments faces challenges like manual configuration complexity and a lack of engineering expertise. This research proposes an Intent-Based Networking (IBN) automation system integrating the Qwen2.5-Coder-7B-Instruct Large Language Model (LLM) with n8n orchestration. The model, fine-tuned via QLoRA and deployed using vLLM, translates natural language instructions into JSON structures executed on MikroTik RouterOS via REST API. Testing across three SOHO clusters (Sidoarjo, Cilacap, and Surabaya) included seven test cases spanning Reliability, Latency, and Quality of Service (QoS). Results demonstrated a 100% execution success rate across 21 scenarios without runtime errors. The n8n backend execution ranged from 5.949 to 103.792 seconds. The system successfully validated operational protection mechanisms, including Conflict Detection Logic, an Enforce Fair Usage Guardrail capped at 90%, and place-before hierarchy injection, proving to be a highly reliable and secure network orchestrator for multi-site environments

Item Type:

Thesis (Undergraduate)

Contributors:

Contribution	Contributors	NIDN/NIDK	Email
Thesis advisor	Junaidi, Achmad	NIDN0710117803	achmadjunaidi.if@upnjatim.ac.id
Thesis advisor	Puspaningrum, Eva Yulia	NIDN0005078908	evapuspaningrum.if@upnjatim.ac.id

Subjects:

T Technology > TK Electrical engineering. Electronics Nuclear engineering > TK5105 Computer Network

Divisions:

Faculty of Computer Science > Departemen of Informatics

Depositing User:

Muhammad Rafi Muhtaddin Noor

Date Deposited:

22 Jul 2026 07:01

Last Modified:

22 Jul 2026 07:59

URI:

https://repository.upnjatim.ac.id/id/eprint/57675