SoftPUF: a Software-Based Blockchain Framework using PUF and Machine Learning

2025-08-04

AI Summary

Traditional Physical Unclonable Functions (PUFs) rely on dedicated hardware, hindering deployment across heterogeneous and legacy devices. To address this, we propose the first hardware-software co-designed blockchain-based identity authentication framework. Our approach leverages SoftPUF—a software-only PUF that employs machine learning to model intrinsic hardware-software characteristics of a device and deterministically generate cryptographically secure, device-unique keys without requiring specialized circuitry. We integrate SoftPUF with a lightweight blockchain consensus mechanism and a multi-layer security protocol to achieve decentralized, Sybil-resistant, and 51%-attack-resilient identity verification. Experimental evaluation demonstrates that our framework significantly reduces hardware dependency and cost, enables secure cloud integration and seamless authentication for diverse legacy devices, and maintains high robustness against adversarial attacks—including modeling, emulation, and physical tampering—while ensuring strong compatibility, scalability, and authentication efficiency.

Technology Category

• Machine Learning: Hardware-aware ML
• Data Mining & Knowledge Management: Representing, Reasoning, and Using Provenance, Trust
• Multiagent Systems: Mechanism Design

Application Category

• Security and Privacy: Blockchains and distributed ledgers
• Systems and Infrastructure for Web, Mobile and WoT: Energy management for devices in mobile Web and WoT environments
• User Modeling, Personalization and Recommendation: Attacks and countermeasures in recommendation systems

Abstract

Physically Unclonable Function (PUF) offers a secure and lightweight alternative to traditional cryptography for authentication due to their unique device fingerprint. However, their dependence on specialized hardware hinders their adoption in diverse applications. This paper proposes a novel blockchain framework that leverages SoftPUF, a software-based approach mimicking PUF. SoftPUF addresses the hardware limitations of traditional PUF, enabling secure and efficient authentication for a broader range of devices within a blockchain network. The framework utilizes a machine learning model trained on PUF data to generate unique, software-based keys for each device. These keys serve as secure identifiers for authentication on the blockchain, eliminating the need for dedicated hardware. This approach facilitates the integration of legacy devices from various domains, including cloud-based solutions, into the blockchain network. Additionally, the framework incorporates well-established defense mechanisms to ensure robust security against various attacks. This combined approach paves the way for secure and scalable authentication in diverse blockchain-based applications. Additionally, to ensure robust security, the system incorporates well-established defense mechanisms against various attacks, including 51%, phishing, routing, and Sybil attacks, into the blockchain network. This combined approach paves the way for secure and efficient authentication in a wider range of blockchain-based applications.

Problem

• Overcoming hardware dependency of PUF for broader authentication
• Enabling secure blockchain authentication without dedicated hardware
• Integrating legacy devices into blockchain with machine learning

Innovation

• Software-based PUF mimics hardware PUF for authentication
• Machine learning generates unique software keys for devices
• Incorporates defense mechanisms against multiple attack types