An optimal lower bound for the Gap-Hamming-Distance problem

Amit Chakrabarti, Dartmouth College, New Hampshire, USA.

In the Gap-Hamming-Distance problem (GHD), Alice and Bob receive n-bit strings and must decide whether they are "close" (Hamming distance <= n/2 - sqrt(n)) or "far" (distance >= n/2 + sqrt(n)). How many bits must they communicate to solve this problem, with error at most 1/3? Since being proposed in the early 2000s by Indyk and Woodruff, for its applications to space lower bounds for data stream algorithms, the problem has been extensively studied, with several interesting lower bounds having been shown. But these results were special-case, the most general being that any k-round GHD protocol requires Omega~(n/k^2) communication. I shall describe a new result, showing that any GHD protocol requires Omega(n) communication, which completely settles the problem. The result uses a new technique for communication lower bounds, that has since (and independently of this work) been subsumed by the "smooth rectangle" method of Jain and Klauck. This is joint work with Oded Regev.