03. Figures-of-merit, design and characterization: friends or enemies?


  • Instrumentation improved a lot over the last 30 years without challenging the fundamental law governing a measurement system: be the best at 1 measurement under 1 set of operation conditions. In the meantime, the systems-under-test have experienced a major restyling: systems-on-a-chip increased the integration level and the flexibility, at the cost of a very large space fo settings.
  • That is why we need instrument 2.0. This instrument measures all the figures-of-merit of the complete system in 1 single experiment, with a level of accuracy that is tailored to the application. 
  • This topic contributes to the instrument 2.0. It brings instrumentation closer to the flexible, fast and adaptive tool that will be needed to tackle the characterization challenges posed by new applications such as self-driving cars or post 5G telecommunication.

Simple figures-of-merit such as the error-vector-magnitude are popular to assess communication systems performance, despite the weak feedback they provide to designers and users about the actual performance and the strengths and weaknesses of the characterized systems. You will bridge the gap between these figures-of-merit and high-level design challenges with a measurement based big data approach. You will develop a surrogate model: this is measurement-based model that accurately predicts the figures-of-merit on the one hand, but also empowers system level. This model therefore needs to be scalable, and will be dependent on several operating conditions, such as impedances, power levels, or bias conditions. Are you in to reunite the design and the characterization worlds?  We are always eager to provide you with more detailed information (virtually or face-to-face) without any obligation.


Back to top