By using the Fourier Transform, show that the following equation $$\int_{-\infty}^{+\infty} K(x-y) g(y) dy = f(x), \qquad -\infty < x < \infty$$ is ill-posed.

For overcoming ill-posedness of above problem what we can do?


By the convolution theorem, the solution to the above equation is the inverse FT of

$$G(k) = \frac{F(k)}{\mathbf{K}(k)} $$

where $\mathbf{K}(k)$ is the FT of $K(x)$. This solution is ill-posed because $G(k)$ is infinite whenever $\mathbf{K}(k)$ is zero.

To get around this, you may use additional information in the problem of which the original equation is a model. For example, if $K$ is the impulse response of a linear system exhibiting some statistical noise, you may use an expectation-maximization technique to find the most likely value of $G(k)$. See this as a primer of deconvolution.


Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.