Frequency estimation with small number of samples, Kalman Filter, periodogram

Hello everyone, I am trying to estimate the frequency of a set of data. Most methods require a long length of samples in order to compute accurate results. The length of my sample is only 1024. Does anyone know a good approach how to estimate the frequency accurate with only 1024 samples?

I already looked up for possible solutions. The matlab function periodogram seems to be a good approach. However, the results is not satisfying.

My code:

data; % set of data

t ; % time-vector

fs = 1000;

[pxx,f]= periodogram(data,hamming(length(data)),length(data),fs;

[Max_Values,frequency] = findpeaks(10*log10(pxx),f,'SortStr','descend');

real_frequency = frequency(1);

Does anyone know how to improve my code?

Does anyone has a code which I could use for implement the kalman filter?

Thank you in advance!