RTL Testability Analysis Based on Genetic Algorithm Implementation

RTL testability analysis, scan register

The paper deals with a new approach to the RTL testability analysis. The approach is based on genetic algorithm theory and applies registers selected into UUA (Unit Under Analysis) scan chain as a population in genetic algorithm process. Quality of selecting registers into UUA scan chain is evaluated by the fitness function which takes into account the number of registers selected into UUA scan chains, number of UUA testable nodes and cost measure needed to test all UUA nodes. Thus, the genetic algorithm is used to find the optimal solution for the the scan chain structure. On the testability analysis basis, UUA modificatins leading to UUA nodes testability improvement are suggested. The goal of this methodology is to maximally utilize original UUA structure for test purposes to reduce necessary UUA modifications.