On Revenue Generation When Auctioning Network Resources

While efficiency of mechanisms for control of communication networks has been extensively investigated, little attention has been paid to the critical metric of revenue generation. In this paper, we pursue such an investigation within a class of allocation schemes with the minimal signaling and computation costs necessary in communication network domains. We show that, within this space, linear cost rules for proportional allocation mechanisms are optimal for symmetric agent populations and reserving a portion of the resource can increase revenue even though less of the resource is being sold. While nonlinear cost functions can be better for asymmetric populations, intelligent agents can undermine this via signal splitting. We show how a resource can counter this phenomenon by declaring a linear cost.