A Multi-polynomial LFSR Based BIST Pattern Generator for Pseudorandom Testing
Haoqi Ren, Zhenya Xiong
Abstract
Haoqi Ren, Zhenya Xiong
Abstract
In this paper, a new linear feedback shift register (LFSR) structure for scan-based built-in self-test (BIST), which has at least two characteristic polynomials, is proposed. Multiple polynomials are utilized to generate the pattern sequences for feeding the scan chain of the circuit under test in pseudorandom testing phase. Using the proposed LFSR, same or even better fault coverage can be achieved in pseudorandom testing phase with less hardware cost, hence can also reduce the test application time in deterministic testing phase. Experimental results demonstrate the advantages of the proposed LFSR over the conventional ones.
OpenAlex reports 2 citations for this work. Citation counts describe recorded attention and do not establish research quality.
A contribution statement is not available in the OpenAlex record.
Method details are not available in the OpenAlex metadata.
Findings are not separately available in the OpenAlex metadata.
Limitations are not available in the OpenAlex metadata.
Application details are not available in the OpenAlex metadata.
In this paper, a new linear feedback shift register (LFSR) structure for scan-based built-in self-test (BIST), which has at least two characteristic polynomials, is proposed. Multiple polynomials are utilized to generate the pattern sequences for feeding the scan chain of the circuit under test in pseudorandom testing phase. Using the proposed LFSR, same or even better fault coverage can be achieved in pseudorandom testing phase with less hardware cost, hence can also reduce the test application time in deterministic testing phase. Experimental results demonstrate the advantages of the proposed LFSR over the conventional ones.
Key concepts: Pseudorandom number generator, Linear feedback shift register, Built-in self-test, Shift register, Fault coverage, Computer science, Algorithm, Generator (circuit theory)