Just dissolve them both in boiling water, mix, and jobs done... almost. The result will be a mixture of Sodium, Potassium, Chlorine and Nitrate ions. You now want the Potassium and Nitrate ions to fall out of solution in the form of Potassium Nitrate. Just allowing it to cool would result in Potassium Nitrate precipitating, as its solubility at 0*C is tiny compared to that at 100*C. However, you would have some Sodium Chloride in there too, which we don't want.
Another method could be fractional crystallisation. You will need a Potassium Nitrate crystal for this. You would get one by allowing the solution to cool, and picking out the largest Potassium Nitrate Crystal, then boiling the solution again to re dissolve the precipitate. I believe Potassium Nitrate will produce long needle crystals when precipitating out slowly, so look for the largest one of them. Then, suspend the crystal in the solution, and the Potassium Nitrate will form on the suspended crystal.
There is a bit more too it than that, but thats the general idea behind it anyway. The link below makes for an interesting read...
http://www.waynesthi...om/crystals.htm