Adaptive overlay network architecture
The adaptive overlay network architecture of the present invention improves the performance of applications distributing digital content among nodes of an underlying network such as the Internet by establishing and reconfiguring overlay network topologies over which associated content items are distributed. The present invention addresses not only frequently changing network congestion, but also interdependencies among nodes and links of prospective overlay network topologies. The present invention provides a prediction engine that monitors metrics and predicts the relay capacity of individual nodes and links (as well as demand of destination nodes) over time to reflect the extent to which the relaying of content among the nodes of an overlay network will be impacted by (current or future) underlying network congestion. The present invention further provides a topology selector that addresses node and link interdependencies while redistributing excess capacity to determine an overlay network topology that satisfies application-specific performance criteria.
1. A method of reconfiguring an existing overlay network topology for the distribution of one or more segments of a content item to a destination node according to an objective function, the method comprising the following steps:
receiving performance criteria relating to the distribution and consumption of the content item;
collecting node metrics associated with one or more parent nodes;
collecting link metrics associated with one or more links;
determining the node-relaying capacity of each of the one or more parent nodes based in part on the node metrics, wherein the node-relaying capacity reflects the ability of the parent node to relay content to one or more associated child nodes;
determining the link-relaying capacity of each of the one or more links based in part on the link metrics, wherein the link-relaying capacity reflects the ability to relay content to a child node connected by that link;
determining, based at least in part upon the determined node-relaying capacities and the determined link-relaying capacities, one or more potential overlay network topologies that satisfy the performance criteria, each of the one or more potential overlay network topologies comprising a portion of one or more parent nodes and a portion of the one or more links;
optimizing the objective function to identify a reconfigured overlay network topology from the one or more potential overlay network topologies; and
determining whether to replace the existing overlay network topology with the reconfigured overlay network topology based on an overhead associated with changing from the existing overlay network topology to the reconfigured overlay network topology.
2. The method of claim 1 wherein the objective function is maximizing the Quality of Experience.
3. The method of claim 1 wherein the objective function is minimizing the number of collectors in the overlay network topology.
4. The method of claim 1 wherein the objective function is minimizing excess relay capacity of parent nodes, and wherein the method further comprises calculating the excess capacity of the plurality of prospective parent nodes.
5. The method of claim 2 wherein determining one or more potential overlay network topologies that satisfy the performance criteria comprises:
sorting the one or more parent nodes according to their node-relaying capacity; and
assigning the one or more links to parent nodes based on the link-relaying capacities according to whether the link-relaying capacities satisfy the performance criteria.
6. The method of claim 1 wherein the prospective parent nodes comprise one or more content distribution networks (CDNs), and wherein determining the capacity of a plurality of prospective parent nodes comprises determining the capacity of one or more prospective CDNs.
7. The method of claim 6 wherein the reconfigured overlay network comprises one or more CDNs.
8. The method of claim 7 wherein the step of determining the node-relaying capacity of each of one or more parent nodes comprises predicting the capacity of each of the one or more parent nodes based on the node metrics; and wherein the step of determining the link-relaying capacity of each of one or more links comprises predicting the capacity of each of the one or more links based on the link metrics.
9. The method of claim 1 or 7 further comprising the steps of:
identifying node metrics and link metrics that are demand-limited, and
filtering to exclude demand-limited or upstream dependency-limited node metrics and link metrics prior to determining the node-relaying and link-related capacity.
10. The method of claim 1 or 7 further comprising the step of predicting the session duration of viewing nodes, and wherein optimizing the objective function to determine an overlay network topology comprises placing nodes with longer session duration at higher levels of the overlay network topology.
11. The method of claim 1 wherein the one or more parent nodes comprise a subset of nodes in the existing overlay network topology.
12. The method of claim 1 further comprising the step of performing periodic performance assessments of all or portions of the existing overlay network topology.
13. The method of claim 1 wherein the reconfigured overlay network topology is a peer-based overlay network topology.
14. The method of claim 1 wherein the steps of determining the node-relaying capacity and determining the link-relaying capacity comprise excluding capacity-limited training samples in the event that the apparent limited capacity of the sample is the result of an upstream dependency or a limitation imposed along the link itself.