First I would plot the data without an axis (Axis -> False) and without frame (Frame -> False).
I would turn the plot into an image (Rasterize). Then use (2D)autocorrelation (ListCorrelate and ImageData) and check for 'spikes'. This gives you the dimensions of the pattern...
What did you try yourself?
