This video introduces the important signal processing application of
system identification; identifying the system impulse response or
transfer function through measurements. The video highlights the
advantages and disadvantages of the three key deterministic approaches,
using as the input an impulse, or step function, or harmonic input. A
fourth method which relies on a stochastic input is then presented,
namely driving a system with white-Gaussian noise (WGN). It is then
shown, using the theory presented earlier in the course, that the
cross-correlation between the input and output is the impulse response.
The sample cross-correlation is highlighted as a way of estimating the
cross-correlation from a single realisation of the random process, where
ergodicity of the output has been assumed. Finally, as simple exam is
implemented in MATLAB.
PGEE11164 Probability, Estimation Theory, and Random Signals Lectures -- School of Engineering, University of Edinburgh. Copyright James R. Hopgood and University of Edinburgh, Scotland, United Kingdom (UK). 2020.
Institute for Digital Communications, Alexander Graham Bell Building, The King's Buildings, Thomas Bayes Road, Edinburgh, EH9 3JL. UK.