Evaluation of Collision Resolution Methods Using Asymptotic Analysis

Abstract

The basic idea of hashing is to use a hash function h(k) to reduce a giant universe to a reasonably small table such that U = {0, 1, .., |U | − 1} → T {0, 1, …, |T | − 1} with Prh{h(key) → T } = 1/|K| which is independent of h(y) for all x y ∈ U. Hashing has been an efficient implicit and explicit search technique for retrieving an element from a collection for many years. Collision is inevitable in a high load factor environment, thus several algorithms for resolving collision to find an alternative slot for a key x in a hash table have been presented. One of the important property of these algorithms is running time complexity. Here, we analysed the performance of these algorithms based on runtime analysis for retrieving and inserting a key using asymptotic analysis. We discovered that many of these algorithms have a non constant insertion and retrieval time. Furthermore, only one algorithm was discovered to have an exact constant time for both lookup and insertion of any arbitrary key into a hash table.