rhymesg / mc_crb_nongaussian

Monte-Carlo Calculation of Cramer-Rao Bound for non-Gaussian Recursive Filtering

MC_CRB_nonGaussian

MATLAB implementation of Monte-Carlo calculation of Cramer-Rao bound for non-Gaussian recursive filtering, with an example of 1-dimensional terrain-reference navigation application.

Citation

Youngjoo Kim and Hyochoong Bang, "Monte-Carlo Calculation of Cramer-Rao Bound for non-Gaussian Recursive Filtering", 2017 Asia-Pacific International Symposium on Aerospace Technology.

Copyright

This work is based on the FIM calculation by Sonjoy Das. See estimateFIM.html.

mc_crb_nongaussian's Issues

Calculation of G

Hello,
The equation 15 of your paper does not fit with your realization in line 131/138 of main_TRN_1d.m:
The value for e.g. thetp should be

            thetpp = thet + (c_tlde*Delk_tlde) + (c*Delk);
            thetp = thet - (c_tlde*Delk_tlde) + (c*Delk);

but in your code it is

            thetpp = thet + (c_tlde*Delk_tlde) + (c*Delk);
            thetp = thet + (c*Delk);

So I would have expected there the negative term of the perturbation. When I change this to my suggested solution, I get of course different results. Or did I missunderstood something?

Thank you