12 July 2017 - 14h00
Feature selection and design for machine learning-based test of analog, mixed-signal and RF circuits
by Manuel Barragan from TIMA
Abstract: The test of analog, mixed-signal and RF (AMS-RF) blocks embedded in a complex system has become a challenging, costly and time consuming task that has been identified as one of the main bottlenecks in the production of current and future integrated systems. Machine learning-based test is a promising strategy for overcoming these issues. Indirect test reduces the complexity and cost of production tests by replacing conventional functional tests at the production line for a set of low-cost indirect observations, often called signatures. Test results are then inferred by post-processing these signatures by building non-linear multi-dimensional regression models. The underlying idea is that signatures are easier to measure than specifications and can be extracted using low-cost equipment, or even by simple on-chip built-in test instruments that can be integrated together with the Device Under Test. One key point that still remains as an open problem is the conception of adequate simple measurement candidates. This seminar presents our research towards efficient algorithms for selecting and designing information rich signatures.