Perform a convolution of two densities and then optimize its parameters (in R)

Question

Given two densities f, g and a data set, I would like to perform a convolution c(t) = (f*g)(t) and then fit the result to the data. Unfortunately, it is not possible to deri