NetSketch: Automated network configuration from hand-drawn topologies

Lu, Yuantao

Permalink

https://hdl.handle.net/2142/122264 Copy

Description

Title

NetSketch: Automated network configuration from hand-drawn topologies

Author(s)

Lu, Yuantao

Issue Date

2023-12-05

Director of Research (if dissertation) or Advisor (if thesis)

Caesar, Matthew Chapman

Department of Study

Electrical & Computer Eng

Discipline

Electrical & Computer Engr

Degree Granting Institution

University of Illinois at Urbana-Champaign

Degree Name

M.S.

Degree Level

Thesis

Keyword(s)

Hand-Drawn Diagrams, Computer Vision, Network Synthesis

Abstract

Designing and configuring computer networks can be a complex and time-consuming process, often requiring expert knowledge and specialized tools. With the growing demand for more intuitive and efficient network design methodologies, this thesis explores the possibilities of computer vision technology as a solution. Specifically, this research seeks to address the following question: Can we simplify the network design process by utilizing computer vision technology to recognize hand-drawn computer networks and configure them automatically? We propose NetSketch, a novel system that integrates modern object detection techniques with classic image gradient-based computer vision algorithms. This system identifies network components from hand-drawn diagrams and employs network synthesis algorithms to automatically generate correct and resilient network models. Based on our evaluation, our network vision framework demonstrates over 99.5% accuracy in detecting network components and network links from hand-drawn diagrams. In addition, our vision synthesis component always generates configurations that meet network design requirements, leading to more efficient network planning and management. This work has the potential to improve network design processes, making them more intuitive and efficient, with applications in network management, education, and collaboration.

Graduation Semester

2023-12

Type of Resource

Thesis

Handle URL

https://hdl.handle.net/2142/122264

Copyright and License Information

Copyright 2023 Yuantao Lu

Owning Collections