An evolutionary approach to the design of on-chip pseudorandom test pattern generators

Abstract

Summary form only given. Weighted pseudorandom test generation (WPRTG) uses test sequences characterized by non-uniform distributions of test vectors in order to increase the detection probability of random resistant faults. Such non-uniform distributions are characterized by the values of signal probability of the CUT inputs (weights). Since different faults may require different distributions, a (small) number of distributions is typically used. The weights of such distributions are identified by analyzing the CUT The corresponding pseudorandom sequences are typically obtained by inserting a combinational network between the TPG and the CUT. Differently from the genetic-based approaches, where only numerical coefficients are computed, we have used an evolutionary programming (EP) algorithm that directly evolves the WGU network. In fact, evolutionary approaches have been shown to be effective in the design of digital circuits. In particular, we evolve a population were each individual represents a possible WGU and the fitness function considers the fault coverage as a primary target and the test length and the cost of the WGU as secondary ones. The fault coverage is evaluated here by means of fault simulation.