An IoT Honeypot Device for Malware Forensics Jingyu YANG@ Tencent Anti-Virus Lab Fan DANG@ Tsinghua University 2017.12.08
About Speakers Jingyu YANG Fan DANG Tencent Anti-Virus Lab HaboMalHunter Malware Analysis IoT Security Research Fan DANG Tsinghua University NFC & IoT Security Research 2017.12.08
Outline Introduction Architecture Implementation Case Study Conclusion 2017.12.08
Introduction High Interaction vs. Low Interaction Traditional Honeypots vs. IoT Honeypots Challenges in IoT environment 2017.12.08
Interaction Why interaction matters? Measures the capability of a honeypot More interaction, more knowledge More interaction, more risks 2017.12.08
Interaction Low Interaction Honeypots High Interaction Honeypots Limited interaction; normally work by emulating Limited to capturing mainly known attacks Easy to deploy High Interaction Honeypots Involve real operating systems Attackers are given everything Vulnerable 2017.12.08
Shell / Video Streaming Traditional vs IoT Traditional IoT Architecture x86 / x86-64 Heterogeneity Service Web / Shell Shell / Video Streaming Deployment Cloud Computing Physical (HIH) 2017.12.08
Challenges Heterogeneity architectures Deployment LIH: Lack of knowledge HIH: Hard to emulate Deployment Expensive Difficult 2017.12.08
Architecture Access Layer Storage Layer View Layer Backend Execution MongoDB API Provider Web Management System Evidence Requests Log Collector Load Balancer 2017.12.08
Load Balancer SSH Agent Port Proxy Telnet Agent Others More Information Accuracy Malicious Actions More Secure Threats Impact Port Proxy SSH Agent Telnet Agent Others 2017.12.08
Technical Challenge How to Forward IP of Real Attackers Hpfeeds Protocol Publish & Subscribe TCP Connection Pair How to Forward IP of Real Attackers 2017.12.08
Yet Another SSH Honeypot Design Principles Hardware Based Design Embedded Linux System Monitoring Malicious Actions Traditional Design (Cowrie) Virtual Environment Python Process Command Emulation 2017.12.08
Raspberry PI 3 2017.12.08
Technical Challenges Re-Initialization Response to Hardware Failure 2017.12.08
Relay 2017.12.08
Case Study Pivoting Attack root:admin File less Network ssh -D 18080 user@IoT ./DoSAttack --sock5 localhost:18080 Target Pivoting Attack File less Network Dynamic port forwarding 2017.12.08
Pivoting Attack 2017.12.08
Why Hardware Solution Accurate Information Immune for Anti-Honeypot CPUs Diversity 2017.12.08
Hardware Based Solution Conclusion IoT Environment HIH Hardware Based Solution SSH Honeypot 2017.12.08
Acknowledgements Authors Tencent Tsinghua University Jingyu YANG, Jie LI, Chen GENG, Zhao LIU, Guize LIU, Jinsong MA Tsinghua University Fan DANG, Yongfeng ZHANG, Prof. Zhenhua LI 2017.12.08
References https://github.com/tencent/habomalhunter https://github.com/threatstream/mhn https://github.com/threatstream/hpfeeds https://www.usenix.org/system/files/conference/woot15/woot15- paper-pa.pdf https://researchcenter.paloaltonetworks.com/2017/07/palo-alto- networks-showcase-iot-honeypot-research-black-hat-2017/ http://blog.malwaremustdie.org/2017/02/mmd-0062-2017-ssh- direct-tcp-forward-attack.html 2017.12.08
Thank you very much :) 2017.12.08