Answers are grounded in this paper and cite their sources.
A class of binary trees is described for maintaining ordered sets of data. Random insertions, deletions, and retrievals of keys can be done in time proportional to log N where N is the cardinality of the data-set. Binary B-trees are a modification of B-trees described previously by Bayer and McCreight. They avoid the storage overhead encountered with B-trees and are suitable for processing in a one-level store.
No takes yet. Share an insight, caveat, or question.
Rudolf Bayer (1971) studied this question.