Current popular asymmetrical cryptosystems are based on hardness of number theoretic problems. In the future, these problems may become practically solvable with the improvements of processing power, the development of quantum computation and distributed computation. So the need for new cryptosystems which are not based on these problems has risen. Researches on hardness of lattice problems have brought a new candidate for asymmetrical cryptography: Lattice Based Cryptography. NTRU is one of these cryptosystems. For practical use, the actual performance results of NTRU with respect to current asymmetrical cryptosystems should be known. This thesis is developed based on this purpose.