First step would be to create a function that computes the distance of the LED point to a single LDR from the analog reading of the LDR. On your piece of paper I would lay down markings from the center of a single LDR at ~5mm intervals, point the laser at each point, and record the analog reading. I'd throw these values into a spreadsheet and then calculate the "line of best fit" (a feature in both Excel or OpenOffice) to generate the function.
Given the above you should now be able to calculate the distance of the laser point to each LDR you would use trilateration to calculate the location of the laser point. I think I'll leave that to you and Google to figure out 