Using Hashing to Solve the Dictionary Problem (In External Memory)

Abstract

We consider the dictionary problem in external memory and improve the update time of the well-known buffer tree by roughly a logarithmic factor. For any λ>= max {lg lg n, log_{M/B} (n/B)}, we can support updates in time O(λ/ B) and queries in sublogarithmic time, O(log_λn). We also present a lower bound in the cell-probe model showing that our data structure is optimal. In the RAM, hash tables have been used to solve the dictionary problem faster than binary search for more than half a century. By contrast, our data structure is the first to beat the comparison barrier in external memory. Ours is also the first data structure to depart convincingly from the indivisibility paradigm.