Art der Publikation: Beitrag in Zeitschrift

Mechanism Design for Machine Scheduling Problems: Classification and Literature Overview

Autor(en):

Kress, D.; Meiswinkel, S; Pesch, E.

Titel der Zeitschrift:

OR Spectrum

Jahrgang (Veröffentlichung):

40 (2018)

Heftnummer:

3

Seiten:

583-611

Sprache:

Englisch

Schlagworte:

Algorithmic mechanism design, Machine scheduling, Game theory, Incentive compatibility

Digital Object Identifier (DOI):

doi:10.1007/s00291-018-0512-8

Volltext:

Mechanism Design for Machine Scheduling Problems: Classification and Literature Overview (287 KB)

Zitation:

Download BibTeX

Kurzfassung

This paper provides a literature overview on (direct revelation) algorithmic mechanism design in the context of machine scheduling problems. Here, one takes a game-theoretic perspective and assumes that part of the relevant data of the machine scheduling problem is private information of selfish players (usually machines or jobs) who may try to influence the solution determined by the scheduling algorithm by submitting false data. A central planner is in charge of controlling and designing the algorithm and a rewarding scheme that defines payments among planner and players based on the submitted data. The planner may, for example, want to design algorithm and payments such that reporting the true data always maximizes the utility functions of rationally acting players, because this enables the planner to generate fair solutions with respect to some social criterion that considers the interests of all players. We review the categories and characterizing problem features of machine scheduling settings in the algorithmic mechanism design literature and extend the widely accepted classification scheme of Graham et al. (Ann Discrete Math 5:287–326, 1979) for scheduling problems to include aspects relating to mechanism design. Based on this hierarchical scheme, we give a systematic overview of recent contributions in this field of research.