Classification of Hash Functions Based on Anti-Attack Ability

Hash functions are widely used in the network field to provide support for load balancing, identity authentication, etc. Since the security problems of traditional hash functions continue to be exposed, researchers have proposed many new hash functions, such as keyed hash functions and provably secure hash functions. However, due to the existing classification methods can't fit the real-life security requirements well, it is still difficult for designers to choose appropriate hash functions in practice. This paper proposes a new classification method of hash functions based on their anti-attack ability, and classifies 301 common hash functions. The method, which considers the performance of hash functions under different attack strategies and the characteristics of new hash functions, can provide a better classification for practical application scenarios. In addition, this paper uses classic supervised learning algorithms to study the classification results based on the performance testing indicators of hash functions. The experimental prediction accuracy of the unknown hash function can reach 88.52%, which verifies that the new classification method has good applicability.