Online Allocation with Unknown Shared Supply

A recent study presents the Online Shared Supply Allocation (OSSA) problem, which is a dynamic online framework for resource distribution. In this model, a central hub allocates a limited and unknown supply to various locations experiencing sequential demand. It incorporates fixed transportation costs and penalties for lost sales, thus eliminating the possibility of backlogging. The researchers suggest a deterministic threshold-proportional policy (GPA) that attains a 4/3-approximation of the offline optimal solution, with an additional term. Lower bounds validate that this 4/3 ratio is indeed tight. Potential applications of this model include humanitarian logistics and the distribution of vaccines.