Content delivery

Patent metadata

Field	Value
Publication number	US-9787745-B1
Application number	US-201615258976-A
Country	US
Kind code	B1
Filing date	Sep 7, 2016
Priority date	Sep 30, 2014
Publication date	Oct 10, 2017
Grant date	Oct 10, 2017

Abstract

A technology for content delivery is provided. In one example, performance of a caching network, performance of a delivery network, and customer demand are modeled. Instructions are provided for a client device on how to request content based on the modeled performance of the caching network, the modeled performance of the delivery network or the modeled customer demand.

First claim

The invention claimed is:

1. A non-transitory computer-readable medium comprising computer-executable instructions which, when executed by a processor, implement a video delivery method, comprising:
generating a content delivery model of performance of video delivery to a client through a content delivery network;
generating an internet service provider model of performance of the video delivery to the client from an internet service provider in communication with the content delivery network;
generating a customer demand model of customer demand for the video delivery based on historical and current demand for the video delivery wherein the content delivery model of performance, the internet service provider model of performance, and the customer demand model of customer demand are generated in a service provider environment separate from the content delivery network and the internet service provider; and
providing instructions for a client device, from the service provider environment, on how to request content based in part on the content delivery model, the internet service provider model or the customer demand model, wherein the instructions include at least one instruction selected from: which content delivery network to use for the video delivery, which edge of the content delivery network to use for the video delivery, or at what bitrate to receive the video delivery.

2. The computer-readable medium of claim 1 , wherein output from each of the content delivery model, the internet service provider model and the customer demand model is used as an input for at least one other of the content delivery model, the internet service provider model and the customer demand model.

3. The computer-readable medium of claim 1 , wherein the method further comprises modeling customer satisfaction based on implicit or explicit feedback from a customer to use as an input for at least one of the content delivery model, the internet service provider model or the customer demand model.

4. The computer-readable medium of claim 1 , wherein at least one of: modeling the performance of the content delivery network, modeling the performance of the internet service provider or modeling the customer demand comprises modeling using a multilayer neural network or regression machine learning method.

5. A computer-implemented content delivery method, comprising:
generating a content delivery model modeling performance of a caching network using a processor;
generating a delivery network model modeling performance of a delivery network using the processor;
generating a customer demand model modeling customer demand using the processor wherein the content delivery model of performance, the internet service provider model of performance, and the customer demand model of customer demand are generated in a service provider environment separate from the caching network and the delivery network; and
providing instructions for a client device, from the service provider environment, to request content based in part on the modeled performance of the caching network, the modeled performance of the delivery network or the modeled customer demand.

6. The method of claim 5 , further comprising instructing the client device to select a different caching network or to a different edge of a same caching network, or instructing client to select a different content delivery bitrate when content delivery is expected to improve based on the modeling of at least one of the performance of the caching network, the performance of the delivery network or the customer demand.

7. The method of claim 5 , further comprising creating projections of future performance and generating the instructions based on improved customer experience for customers.

8. The method of claim 5 , wherein the content delivery model includes one or more delivery inputs selected from the group consisting of: stream count, delivered bitrate, bitrate distribution, caching network capacity and events.

9. The method of claim 5 , wherein the delivery network model includes one or more inputs selected from the group consisting of: the performance of the caching network, client latency, caching network latency, delivery network throughput, delivered bitrate and customer demand.

10. The method of claim 5 , wherein the customer demand model includes one or more inputs selected from the group consisting of: recent demand, comparative demand, past demand trends, the performance of the caching network, the performance of the delivery network and events.

11. The method of claim 5 , further comprising receiving feedback from the client device regarding at least one of client performance, delivery network performance or caching network performance to use as input for modeling the performance of the caching network, modeling the performance of the delivery network or modeling the customer demand.

12. The method of claim 5 , further comprising modeling customer satisfaction based on customer scoring of delivery of the content and using the modeled customer satisfaction as an input for modeling the performance of the caching network, modeling the performance of the delivery network or modeling the customer demand.

13. The method of claim 5 , further comprising modeling customer satisfaction based on customer scoring of the content delivered.

14. The method of claim 5 , further comprising performing the modeling of the performance of the caching network, the performance of the delivery network or the customer demand for at least one of: predetermined time intervals, at times determined based on historical trends, or in response to events.

15. A non-transitory computer-readable medium comprising computer-executable instructions which, when executed by a processor, implement a content delivery system, comprising:
a content module configured to deliver content in response to a request for the content;
a performance module configured to monitor performance of the delivery of the content;
a learning module configured to model a caching network, a delivery network, and customer demand wherein the learning module is hosted in a service provider environment separate from the caching network and the delivery network; and
a delivery module configured to make a determination of how to deliver the content to a client device using the models from the learning module.

16. The computer-readable medium of claim 15 , further configured to implement a system comprising a reports data store to store the performance of delivery of the content received from a client.

17. The computer-readable medium of claim 15 , further comprising a communication module configured to receive communications from a client regarding the performance of the delivery of the content.

18. The computer-readable medium of claim 15 , wherein the learning module is a machine learning module using a multilayer neural network or regression.

19. The computer-readable medium of claim 15 , wherein output from each model is used as an input for each other models.

20. The computer-readable medium of claim 15 , wherein the learning module is further configured to model customer satisfaction.

Assignees

• Amazon Tech Inc

Inventors

• Binns Justin Michael
• Waggoner Charles Benjamin Franklin
• Barnard Joshua B
• Bordenet Matthew James
• Ulness Lars Christian
• Joliveau Marc
• Knowler Jim
• Backman Terje Kristian
• Longo Chris

Classifications

• G06N3/0499

  Feedforward networks · CPC title

• G06N3/09

  Supervised learning · CPC title

• H04L65/4076

  Electricity · mapped topic

• H04L65/4084

  Electricity · mapped topic

• H04L65/602 Primary

  Electricity · mapped topic