Presentation | 2024-02-28 Test Point Selection Method for Multi-Cycle BIST Using Deep Reinforcement Learning Kohei Shiotani, Tatsuya Nishikawa, Shaoqi Wei, Senling Wang, Hiroshi Kai, Yoshinobu Higami, Hiroshi Takahashi, |
---|---|
PDF Download Page | PDF download Page Link |
Abstract(in Japanese) | (See Japanese page) |
Abstract(in English) | Multi-cycle BIST is a test method that performs multiple captures for each scan pattern, proving effective in reducing test patterns in in-system testing. However, multi-cycle capture can lead to a decrease in the testability (controllability and observability) of the temporally unfolded logic circuit, potentially causing a reduction in fault detection capability and, consequently, hindering the reduction of test patterns. This study aims to improve the testability of multi-cycle BIST by proposing a control point selection method that combines spatio-temporal graph neural networks with deep reinforcement learning. The proposed method selects optimal control points by considering the testability and logical structure of the signal lines in the temporally unfolded logic circuit in terms of their spatio-temporal relationships. Specifically, it uses spatio-temporal graph neural networks to model the testability of signal lines from the structural and temporal characteristics of the logic circuit and employs deep reinforcement learning to efficiently search for control point positions that maximize the testability of the temporally unfolded circuit. Evaluation results on benchmark circuits have confirmed the effectiveness of the proposed method. |
Keyword(in Japanese) | (See Japanese page) |
Keyword(in English) | Multi-cycle Test / Time-Expansion Circuit / Time-Series Variables / Graph Convolutional Neural Networks / Test Point insertion |
Paper # | DC2023-98 |
Date of Issue | 2024-02-21 (DC) |
Conference Information | |
Committee | DC |
---|---|
Conference Date | 2024/2/28(1days) |
Place (in Japanese) | (See Japanese page) |
Place (in English) | Kikai-Shinko-Kaikan Bldg. |
Topics (in Japanese) | (See Japanese page) |
Topics (in English) | |
Chair | Tatsuhiro Tsuchiya(Osaka Univ.) |
Vice Chair | Toshinori Hosokawa(Nihon Univ.) |
Secretary | Toshinori Hosokawa(Nihon Univ.) |
Assistant |
Paper Information | |
Registration To | Technical Committee on Dependable Computing |
---|---|
Language | JPN |
Title (in Japanese) | (See Japanese page) |
Sub Title (in Japanese) | (See Japanese page) |
Title (in English) | Test Point Selection Method for Multi-Cycle BIST Using Deep Reinforcement Learning |
Sub Title (in English) | |
Keyword(1) | Multi-cycle Test |
Keyword(2) | Time-Expansion Circuit |
Keyword(3) | Time-Series Variables |
Keyword(4) | Graph Convolutional Neural Networks |
Keyword(5) | Test Point insertion |
1st Author's Name | Kohei Shiotani |
1st Author's Affiliation | Ehime University(Ehime Univ.) |
2nd Author's Name | Tatsuya Nishikawa |
2nd Author's Affiliation | Ehime University(Ehime Univ.) |
3rd Author's Name | Shaoqi Wei |
3rd Author's Affiliation | Ehime University(Ehime Univ.) |
4th Author's Name | Senling Wang |
4th Author's Affiliation | Ehime University(Ehime Univ.) |
5th Author's Name | Hiroshi Kai |
5th Author's Affiliation | Ehime University(Ehime Univ.) |
6th Author's Name | Yoshinobu Higami |
6th Author's Affiliation | Ehime University(Ehime Univ.) |
7th Author's Name | Hiroshi Takahashi |
7th Author's Affiliation | Ehime University(Ehime Univ.) |
Date | 2024-02-28 |
Paper # | DC2023-98 |
Volume (vol) | vol.123 |
Number (no) | DC-389 |
Page | pp.pp.23-28(DC), |
#Pages | 6 |
Date of Issue | 2024-02-21 (DC) |