A compact data structure for representing a dynamic multiset

We develop a data structure for maintaining a dynamic multiset that uses O(nlglgn/lgn) bits and O(1)O(1) words, in addition to the space required by the n   elements stored, supports searches in O(lgn) worst-case time and updates in O(lgn) amortized time. Compared to earlier data structures, we improve the space requirements from O(n)O(n) bits to O(nlglgn/lgn) bits, but the running time of updates is amortized, not worst-case.

Research highlights
► Gives a space-efficient dynamic multiset representation.
► Improves the space redundancy of earlier structure from constant bits per element to o(1)o(1) bits per element.
► Gives several improved memory management techniques.