One easy way to remove EOG from your EEG signal is to calculate the correlation for each pair of EOG and EEG electrodes and to subtract the respective percentage of EOG from your measured EEG signal. In an ideal situation your EOG signal would be free of any EEG component. If you want a simulated EEG with EOG contamination you can just calculate the other way around. That means, generate a pseudo EOG signal (simple low frequency wave) and calculate the sum of the weighted artifical EOG and an EEG channel - if you like you can choose the weighting factor with respect to the channel position (frontal/big vs. occipital/small).

There are free EEG data sets available for download. -> Have someone who knows EEG interpretation to mark the 'eye blink events' -> Run waveform analysis on the events marked on the real EEG data (MATLAB-EEGLAB or R EEGKit) -> add those waveform features (from real EEG data) at intervals of 4 - 10 seconds to your simulated EEG signal -> That should give you a signal that is close to natural eye blink artifact. . . . . . . Once you have this signal you can test your blink artifact rejection algorithms.

Hopefully, there is a strong reason why you choose to use a simulated signal and not original EEG data. It would be more useful if your "blink artifact rejection algorithm" were performing better on real data than on simulated data.

Ganne Chaitanya Frank Schmitz Thank you very much.

Ganne Chaitanya , Do you recommend any data set that contains both EEG contaminated with EOG artifact along with EOG recording?