Fine-Tuning the Red-Black Tree Delete Algorithm

Abstract

Abstract Red-Black trees have gained popularity as a data structure widely used for implementing dictionaries, associative arrays, and symbol tables in compilers like C++ and Java, as well as other systems. This paper introduces an enhanced delete algorithm for this type of binary search tree. The proposed algorithm shows promise by introducing a different coloring scheme to the tree, resulting in a significant reduction of approximately 29% and 11% in color changes and maintenance operations, respectively. The new coloring emphasizes the presence of red nodes, effectively accelerating the deletion operation. Moreover, it achieves a 4% improvement in running time when insert and delete operations are performed together, while preserving the search performance of the standard algorithm.