Multistage online maxmin allocation of indivisible entities

Abstract

We consider an online allocation problem that involves a set P of n players and a set E of m indivisible entities over discrete time steps 1,2,…,τ. At each time step t∈[1,τ], for every entity e∈E, there is a restriction list Lt(e) that prescribes the subset of players to whom e can be assigned and a non-negative value vt(e,p) of e to every player p∈P. The sets P and E are fixed beforehand. The sets Lt(⋅) and values vt(⋅,⋅) are given in an online fashion. An allocation is a distribution of E among P, and we are interested in the minimum total value of the entities received by a player according to the allocation. In the static case, it is NP-hard to find an optimal allocation the maximizes this minimum value. On the other hand, ρ-approximation algorithms have been developed for certain values of ρ∈(0,1]. We propose a w-lookahead algorithm for the multistage online maxmin allocation problem for any fixed w⩾1 in which the restriction lists and values of entities may change between time steps, and there is a fixed stability reward for an entity to be assigned to the same player from one time step to the next. The objective is to maximize the sum of the minimum values and stability rewards over the time steps 1,2,…,τ. Our algorithm achieves a competitive ratio of (1−c)ρ, where c is the positive root of the equation wc2=ρ(w+1)(1−c). When w=1, it is greater than ρ4ρ+2+ρ10, which improves upon the previous ratio of ρ4ρ+2−21−τ(2ρ+1) obtained for the case of 1-lookahead.