Design and Implementation of a More Efficient Hashing Algorithm
Sam Barani, William Ou, Evan Katz
Jasper High School, Plano, Texas, United States

The purpose of this project was to develop and implement a cryptographic hashing algorithm that is more secure against “brute force” attacks but not at the cost of computation speed. Cryptographic hashing convert sets of data into fixed-size strings of bits. Functions like MD5 are commonly used when storing data such as passwords. The problem with MD5 is that the input used to generate the resulting hash can be easily retrieved using a “brute force” attack. A brute force program inputs every possible value until the output hash matches the one stored in a database. Upon researching how the MD5 function works, it was found that it consists of several ‘rounds’ of functions where four constants are modified and added together to create a single string of data based on the input. For this project, algorithm was created and implemented based on MD5. This new algorithm uses 8 constants, meaning the hash is twice as long (256 bits). This would result in a larger number of possible combinations (2^256 instead of 2^128) and by default computation times would be longer, making brute forcing exponentially harder, as millions of hashes have to be calculated. Computation times and brute force times were tested for both MD5 and the new algorithm using several “good” and “bad” passwords, and the ratios between increases in brute force times and increases in computation times were found. These ratios tended to be very high, indicating that the purpose of the project was fulfilled.