Abstract

When modelling high-speed electromagnetic interconnections, two aspects must be taken into account: the optimisation of the design parameters to obtain the desired performance, and the tolerance one should expect on the optimised performance due to the difference between the real and the computed design parameters. Two main categories of optimisation techniques exist: stochastic search techniques (e.g. genetic algorithms) and gradient techniques. Stochastic techniques are easier to implement but require a very large number of iterations. This is often unacceptable when using full-wave simulations, which require a lot of CPU-time. Gradient techniques require the computation of the derivatives of the performance criteria with respect to the design parameters, but this can be implemented efficiently using adjoint-style sensitivity techniques. A sensitivity-based statistical analysis is applied to the full-wave analysis of electric and dielectric waveguide structures.

Full Text
Paper version not known

Talk to us

Join us for a 30 min session where you can share your feedback and ask us any queries you have

Schedule a call