Related Experiment Video
Updated: Jun 25, 2026
05:30
Large Scale Energy Efficient Sensor Network Routing Using a Quantum Processor Unit
Published on: September 8, 2023
A heterogeneous graph attention network based IPv6 target generation algorithm for nonseed prefixes
Yangxiang Zhou1,2, Liancheng Zhang3,4, Haojie Zhu1,2
1Information Engineering University, Zhengzhou, 450001, China.
Scientific Reports
|May 22, 2026
Summary
This study introduces 6HAN, a novel method for generating IPv6 target addresses, significantly improving hit rates and prefix coverage for nonseed prefixes. It overcomes limitations of previous algorithms by utilizing heterogeneous graph attention networks and Whois data.
Area of Science:
- Computer Science
- Network Security
- Data Mining
Background:
- The expansive and sparse nature of the IPv6 address space presents significant challenges for efficient network scanning.
- Existing IPv6 target generation methods often fail with nonseed prefixes due to reliance on seed address patterns and limited use of auxiliary data.
Purpose of the Study:
- To develop an advanced IPv6 target generation algorithm that addresses the limitations of current methods, particularly for nonseed prefixes.
- To enhance target address hit rates and prefix coverage in IPv6 scanning.
Main Methods:
- Proposed 6HAN, a heterogeneous graph attention network (HAN) approach to model multi-modal correlations between IPv6 prefixes and Whois metadata.
- Integrated Whois auxiliary attributes with prefix structural features for embedding generation and employed a dual-task joint self-supervised learning framework.
- Utilized graph neural networks for predicting missing Whois information and pattern migration from similar seed prefixes.
Main Results:
- 6HAN demonstrated substantial improvements in hit rates (13.81%-201.05%) and prefix coverage (17.41%-233.58%) compared to HMap6 and AddrMiner-N.
- The method effectively handles nonseed prefixes and missing Whois information, outperforming existing techniques.
Conclusions:
- 6HAN offers a superior approach to IPv6 target generation, particularly for nonseed prefixes, by leveraging heterogeneous graph attention networks and comprehensive data integration.
- The proposed method significantly enhances the efficiency and effectiveness of IPv6 network scanning.