Balanced allocation on hypergraphs
Catherine Greenhill,Bernard Mans,Ali Pourmiri
DOI: https://doi.org/10.1016/j.jcss.2023.05.004
IF: 1.043
2023-12-01
Journal of Computer and System Sciences
Abstract:We consider a variation of balls-into-bins which randomly allocates m balls into n bins. Following Godfrey's model (SODA, 2008), we assume that each ball t, 1 ⩽ t ⩽ m , comes with a hypergraph H ( t ) = { B 1 , B 2 , … , B s t } , and each edge B ∈ H ( t ) contains at least a logarithmic number of bins. Given d ⩾ 2 , our d-choice algorithm chooses an edge B ∈ H ( t ) , uniformly at random, and then chooses a set D of d random bins from the selected edge B. The ball is allocated to a least-loaded bin from D. We prove that if the hypergraphs H ( 1 ) , … , H ( m ) satisfy a balancedness condition and have low pair visibility, then after allocating m = Θ ( n ) balls, the maximum load of any bin is at most log d log n + O ( 1 ) , with high probability. Moreover, we establish a lower bound for the maximum load attained by the balanced allocation for a sequence of hypergraphs in terms of pair visibility.
computer science, theory & methods, hardware & architecture