One-way games

This paper is motivated by optimization applications involving multiple stakeholders. Examples of such applications are found in large-scale restoration of interdependent infrastructures after significant disruptions (e.g.,[1, 2]), humanitarian logistics over multiple states or regions (e.g., [3]), supply chain coordination (e.g., [4]) and integrated logistics involving, say, a port, inland terminals, and railway and truck operators. Consider, for instance, the restoration of the power system and the telecommunication network after a major disaster. As explained in [3], there are one-way dependencies between the power system and the telecommunication network. This means, for instance, that some power lines must be restored before some part of the telecommunication network can become available. It is possible to use centralized mechanisms for restoring the system as a whole. However, in practice, it is often the case that these restorations are performed by different agencies with independent objectives and selfish behavior may have a strong impact on the social welfare. It is thus important to study whether it is possible to find high-quality outcome to these problems in decentralized settings when the stakeholders proceed independently and do not share complete information about their costs. This paper aims at taking a first step in this direction. We propose a class of one-way games that abstracts some of the salient features of these applications. In this first approach, we restrict the attention to one-way dependencies.