Scholarlink

<p>Static random access memory physically unclonable function (SRAM PUF) exploits process variations in identically designed transistors during production to generate a unique hardware fingerprint for each chip. The error correction construction cost of conventional SRAM PUF is high, and there is a problem of unstable response due to factors such as external ambient temperature or power supply voltage. Therefore, this paper uses the characteristics of spatial autocorrelation inside the SRAM circuit to propose a PUF preselection method based on neighboring spatial majority voting, and designs a generation scheme that can stably generate keys without improving the original circuit and increasing the test cost. Test data shows that in the case of 6.2% unstable SRAM cells, this method can obtain a 128-bit key with a failure rate of 2.1 × 10<sup>-6</sup> using 3200 SRAM cells. Under the change of power supply voltage (± 10%) and temperature (-20℃~70℃), the bit error rate is still less than 3 × 10<sup>-6</sup>. The results show that the method can meet the requirements of uniqueness and randomness of PUF key generation and has high stability.</p>

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

SRAM PUF Preselection Method Based on Neighboring Spatial Majority Voting

  • 期刊论文

  • 发表日期：

    2026年2月2日

查看全文
  • Tianyao Shi,
  • Sining Chen,
  • Haijin Chen

摘要

Static random access memory physically unclonable function (SRAM PUF) exploits process variations in identically designed transistors during production to generate a unique hardware fingerprint for each chip. The error correction construction cost of conventional SRAM PUF is high, and there is a problem of unstable response due to factors such as external ambient temperature or power supply voltage. Therefore, this paper uses the characteristics of spatial autocorrelation inside the SRAM circuit to propose a PUF preselection method based on neighboring spatial majority voting, and designs a generation scheme that can stably generate keys without improving the original circuit and increasing the test cost. Test data shows that in the case of 6.2% unstable SRAM cells, this method can obtain a 128-bit key with a failure rate of 2.1 × 10-6 using 3200 SRAM cells. Under the change of power supply voltage (± 10%) and temperature (-20℃~70℃), the bit error rate is still less than 3 × 10-6. The results show that the method can meet the requirements of uniqueness and randomness of PUF key generation and has high stability.