Genetic algorithms for scan path design

In this paper, we consider genetic algorithms for two problems connected with scan path design for VLSI circuits. The first problem, called partial scan path selection, is to optimize the number of flip-flops to be included in the scan path, at the same time maximizing the fault coverage. The second problem, called multiple scan configuration, is to divide the scan path into a specified number of subchains in order to minimize the test application time. Both problems are combinatorial optimization problems and we show that they lend themselves effectively to a genetic formulation. Our algorithms have been implemented on a Sun/SPARC in the OASIS design automation environment. We have tested our algorithms on standard benchmarks and the results are encouraging. Compared to other existing methods for scan design, the genetic paradigm is attractive due to its simplicity, potential to generate a global optimal solution, and capability to model compound objective functions.

[1]  Janak H. Patel,et al.  A fault oriented partial scan design approach , 1991, 1991 IEEE International Conference on Computer-Aided Design Digest of Technical Papers.

[2]  C. P. Ravikumar,et al.  Simulated annealing for target-oriented partial scan , 1994, Proceedings of 7th International Conference on VLSI Design.

[3]  Melvin A. Breuer,et al.  Configuring multiple scan chains for minimum test time , 1992, ICCAD.

[4]  Melvin A. Breuer,et al.  BALLAST: a methodology for partial scan design , 1989, [1989] The Nineteenth International Symposium on Fault-Tolerant Computing. Digest of Papers.

[5]  K.-T. Cheng,et al.  A Partial Scan Method for Sequential Circuits with Feedback , 1990, IEEE Trans. Computers.

[6]  D. E. Goldberg,et al.  Genetic Algorithms in Search , 1989 .

[7]  David E. Goldberg,et al.  Genetic Algorithms in Search Optimization and Machine Learning , 1988 .

[8]  S.M. Reddy,et al.  On determining scan flip-flops in partial-scan designs , 1990, 1990 IEEE International Conference on Computer-Aided Design. Digest of Technical Papers.