Since the advent of data communication over networks, it has become imperative to ensure security of information. Cryptography is a technique that is being employed. This paper takes a look at an important aspect of the public key encryption scheme, the modular exponentiation technique, with the view of optimizing it. Taking a look at some public key encryption schemes, it would be observed that the modular exponentiation process is primal to achieving high speed algorithms in data encryption. With special emphasis on the Montgomery exponentiation algorithm, a blend of this algorithm with the sliding window method of exponentiation is proposed. A detailed complexity analysis of the proposed and selected algorithms was carried out. Both algorithms were implemented and simulated using MATLAB 6.5. While the proposed algorithm did not prove to be faster than the classical Montgomery exponentiation algorithm, it was rather observed that it makes lesser number of calls to the Montgomery reduction sub-function. This means 10% lesser number of loops during execution and thus better optimized for lower memory applications.