# US 12,000,891 B2 · Scan correlation-aware scan cluster reordering method and apparatus for low-power testing

> Agents: this data is available as JSON from the patentgrep API and MCP server — see https://patentgrep.com/llms.txt

- Number: US 12,000,891 B2
- Document ID: US12000891B2
- Status: Granted
- Kind: B2
- Application number: 17/931,044
- Filed: 2022-09-09
- Published: 2023-04-27
- Granted: 2024-06-04
- Earliest claimed: 2021-10-26
- Assignees: UIF (University Industry Foundation), Yonsei University
- Applicants: UIF (University Industry Foundation), Yonsei University
- Inventors: Sungho Kang; Sang Jun Lee
- Web page: https://patentgrep.com/patent/US12000891B2

## CPC

- G01R 31/318538: {Topological or mechanical aspects}
- G01R 31/318583: {Design for test}
- G01R 31/318555: {Control logic}
- G06F 30/333: Design for testability [DFT], e.g. scan chain or built-in self-test [BIST]
- G01R: MEASURING ELECTRIC VARIABLES; MEASURING MAGNETIC VARIABLES (indicating correct tuning of resonant circuits H03J3/12)
- G06F: ELECTRIC DIGITAL DATA PROCESSING (computer systems based on specific computational models G06N)

## Abstract

The exemplary embodiments of the present invention provides a scan cluster reordering method and apparatus which perform a scan correlation aware scan cluster reordering to reduce a power generated during the scan test and place scan cells having a high correlation to be adjacent to each other by analyzing the correlation between scan cells, and reduce a test power generated during the scan test.

## Claims

### Claim 1

A scan cluster reordering method, comprising:
constructing a distance matrix between scan cells using a test pattern;
performing scan correlation clustering using the constructed distance matrix and acquiring a hierarchical structure as a clustering result; and
performing scan chain reordering using the hierarchical structure,
wherein as the scan correlation clustering, a hierarchical agglomerative clustering (HAC) is performed based on the smallest distance in the distance matrix,
wherein in the acquiring a hierarchical structure, a dendrogram tree is output as a clustering result.

### Claim 2 (depends on 1)

The scan cluster reordering method according to claim 1, wherein an unassigned bit (X-bit) of the test pattern is filled with zero (0-filling).

### Claim 3 (depends on 1)

The scan cluster reordering method according to claim 1, wherein a distance matrix is generated for every scan chain and a normalized hamming distance is applied.

### Claim 4 (depends on 1)

The scan cluster reordering method according to claim 1, wherein distance matrices for an input vector and an output vector of the test pattern are generated, a first distance matrix for the input vector is output, and a second distance matrix for the output vector is output.

### Claim 5 (depends on 4)

The scan cluster reordering method according to claim 4, wherein a first weight for the input vector is calculated, a second weight for the output vector is calculated, and the first weight is applied to the first distance matrix and the second weight is applied to the second distance matrix and added to calculate a total distance matrix.

### Claim 6 (depends on 5)

The scan cluster reordering method according to claim 5, wherein a relative standard deviation obtained by dividing a standard deviation by a mean value is applied to the first weight and the second weight.

### Claim 7 (depends on 1)

The scan cluster reordering method according to claim 1, wherein a scan cell order is determined in accordance with the order of the cluster corresponding to the hierarchical structure.

### Claim 8

A scan cluster reordering apparatus including a processor,
wherein the processor constructs a distance matrix between scan cells using a test pattern, performs scan correlation clustering using the constructed distance matrix and acquires a hierarchical structure as a clustering result, and performs scan chain reordering using the hierarchical structure,
wherein as the scan correlation clustering, a hierarchical agglomerative clustering (HAC) is performed based on the smallest distance in the distance matrix,
wherein the processor, a dendrogram tree is output as a clustering result.