Online Resource Allocation Using Distributed Bidding Approaches

Abstract One strategy for efficient resource distribution in edge-cloud computing is auctioning resources to the prospective clients. This work focuses on performing such auctions in a distributed, online environment. Such auctions realistically model how coalition partners might share cloud resources with one another, as opposed to the centralized, static case. In addition, some of the resources provided by the edge servers will be elastic; i.e., clients can receive more resources so that their jobs complete more quickly. Since there may be varying levels of trust in a coalition environment, we examined varying levels of job utility disclosure between clients and servers. We also implement a strategy to strongly discourage dishonesty when jobs disclose their utility to servers.
Authors
  • Caroline Rublein (PSU)
  • Tom La Porta (PSU)
  • Fidan Mehmeti (PSU)
Date Sep-2020
Venue 4th Annual Fall Meeting of the DAIS ITA, 2020