AI-Driven Security in Streaming Scan Networks (SSN) for Design-for-Test (DFT)

Streaming Scan Networks used in chip architectures for DFT have been significantly improved and made efficient. We can send all the test data in one packetized format. However, as scan chains become more readily accessible, they bring some very real security vulnerabilities. For example, unauthorized test access can quickly be taken off the chain, or scan-based side-channel attacks could bring about severe intellectual property leaks (IP). In this paper, we'll investigate AI-driven methods to bolster the Security of SSNs. These include detecting threats, monitoring with statistical analysis, encrypted scan strategies, and integrated hardware security. The scan architecture’s integrity and confidentiality can be protected by AI-based anomaly detection and cryptographic protections without endangering test efficiency.