Toward Coalition Resource Discovery and Representation for SDC using Generic Mathematical Programming Constraints as Basic Abstraction

Abstract A fundamental challenge in software defined coalition (SDC) is to discover resources controlled by multiple coalition partner networks with autonomy and privacy policies. To address this challenge, this paper first introduces generic mathematical programming constraints as a compact, secure representation of multiple properties (e.g., bandwidth, delay and loss rate) of available resources in SDC. We develop an obfuscating protocol, to address the privacy concerns by ensuring that no partner can associate the mathematical programming constraints with the corresponding partner networks. We also introduce a super-set projection technique to increase system scalability. We deploy the design in a small SDC network, and evaluate its performance using real topologies and traces. We show that the design (1) efficiently discovers available networking resources in a coalition network on average four orders of magnitude faster, and allows fairer resource allocations; (2) preserves the partner networks’ privacy with little overhead; and (3) scales to a coalition of 200 partner networks. Preliminary results appeared in Supercomputing 2018, JSAC2019.
Authors
  • Qiao Xiang (Yale)
  • Franck Le (IBM US)
  • Yeon-sup Lim (IBM US)
  • Richard Yang (Yale)
Date Sep-2019
Venue Annual Fall Meeting of the DAIS ITA, 2019