A Stock Options Metaphor for Content Delivery Networks

TL;DR

This paper introduces a novel resource management framework for Content Delivery Networks using stock options and predictive reservation schemes to enhance efficiency and robustness through secondary markets and automated trading.

Contribution

It proposes a new resource allocation model for CDNs employing stock options, predictive reservation, and secondary markets to optimize resource utilization and availability.

Findings

Improved resource utilization in CDNs.

Enhanced robustness through secondary market mechanisms.

Effective resource allocation using stock options.

Abstract

The concept of Stock Options is used to address the scarcity of resources, not adequately addressed by the previous tools of our Prediction Mechanism. Using a Predictive Reservation Scheme, network and disk resources are being monitored through well-established techniques (Kernel Regression Estimators) in a given time frame. Next, an Secondary Market mechanism significantly improves the efficiency and robustness of our Predictive Reservation Scheme by allowing the fast exchange of unused (remaining) resources between the Origin Servers (CDN Clients). This exchange can happen, either by implementing socially optimal practices or by allowing automatic electronic auctions at the end of the day or at shorter time intervals. Finally, we further enhance our Prediction Mechanism; Stock Options are obtained and exercised, depending on the lack of resources at the end of day. As a result, Origin Servers may acquire resources (if required) at a normal price. The effectiveness of our mechanism further improves.