Predictive network system and method

• US 9,680,766 B2
• Filed: 09/28/2011
• Issued: 06/13/2017
• Est. Priority Date: 09/28/2010
• Status: Active Grant

• Current Assignee: Ohio State Innovation Foundation (Ohio State University), University of Southern California
• Original Assignee: Ohio State Innovation Foundation (Ohio State University), University of Southern California
• Inventors: Eryilmaz, Atilla, El Gamal, Hesham, Sha, Fei, McLaughlin, Margaret, Caire, Giuseppe
• Primary Examiner(s): MEHRA, INDER P
• Application Number:
  US13/876,781
• Publication Number:
  US 20140113600A1
• Time in Patent Office: 2,085 Days
• Field of Search: 4554121, 4554141, 4554143, 4554221, 4554321, 455440, 455441442, 455443, 4554561, 455517, 709201, 709202, 709203, 705 1458
• CPC Class Codes: H04L 47/823 Prediction of resource usage
• CPC Class Codes: H04L 47/83 based on usage prediction
• CPC Class Codes: H04W 28/02 Traffic management, e.g. flow control or congestion control
• CPC Class Codes: H04W 28/16 Central resource management; Negotiation of resources or communication parameters, e.g. negotiating bandwidth or QoS [Quality of Service]

Abstract

A proactive networking system and method is disclosed. The network anticipates the user demands in advance and utilizes this predictive ability to reduce the peak to average ratio of the wireless traffic and yield significant savings in the required resources to guarantee certain Quality of Service (QoS) metrics. The system and method focuses on the existing cellular architecture and involves the design and analysis of learning algorithms, predictive resource allocation strategies, and incentive techniques to maximize the efficiency of proactive cellular networks. The system and method further involve proactive peer-to-peer (P2P) overlaying, which leverages the spatial and social structure of the network. Machine learning techniques are applied to find the optimal tradeoff between predictions that result in content being retrieved that the user ultimately never requests, and requests that are not anticipated in a timely manner.

Claims

1. A computerized method for delivering data resources in a network comprising:

(a) collecting a plurality of prior requests from an individual user for dynamically delivered user demand content from said network, said requests initiated by an individual user mobile device;

(b) analyzing said plurality of prior requests for dynamically delivered user demand content against a statistical model to predict anticipated requests from said individual user mobile device, said statistical model utilizing dynamic data requests for user demand content from said network obtained from a plurality of mobile devices;

(c) generating a list of anticipated requests from said individual user mobile device for user demand content, said list comprising a deadline for each anticipated request;

(d) for a plurality of anticipated requests and deadlines in said list;

(1) downloading to a site accessible to said individual user mobile device in advance of said deadlines data resources responsive to said anticipated requests; and

(2) storing at said site said data resources responsive to said anticipated requests;

(e) notifying said individual user of the availability of at least one of said data resources responsive to said anticipated requests;

(f) providing said individual user with an incentive associated with requesting at least one of said data resources responsive to said anticipated requests; and

(g) in response to at least one actual request from said individual user for user demand content, transferring from said site to said individual user mobile device at least one of said data resources responsive to said actual request.

2. The computerized method of claim 1 wherein said deadline is an expiration time for downloading to said site a data resource responsive to said anticipated request.

3. The computerized method of claim 1 wherein analyzing said plurality of prior requests for dynamically delivered user demand content comprises clustering actual dynamic data requests according to user interests.

4. The computerized method of claim 1 wherein analyzing said plurality of prior requests for dynamically delivered user demand content comprises clustering user demand content from said plurality of mobile devices based on similarity in topics.

5. The computerized method of claim 1, wherein said incentive is a monetary incentive.

6. The computerized method of claim 1, wherein said incentive can only be obtained if said at least one of said data resources responsive to said anticipated requests is requested when said network is experiencing off-peak traffic demand.

7. A computerized method for delivering data resources in a network comprising:

(a) collecting a plurality of prior requests for data resources from said network from an individual user, said requests initiated by an individual user mobile device;

(b) generating a list of anticipated requests for data resources by analyzing said plurality of prior requests for data resources against a statistical model to determine anticipated requests from said individual user, said statistical model utilizing actual requests for data resources from said network obtained from a plurality of mobile user devices;

(c) generating a proactive schedule for downloading said data resources responsive to said anticipated requests to a site accessible to said individual user mobile device in advance of actual requests from said individual user for said data resources;

(d) downloading said data resources to said site in accordance with said proactive schedule;

(e) notifying said individual user of the availability of at least one of said data resources downloaded to said site; and

(f) in response to at least one actual request from said individual user for at least one data resource, transferring from said site to said individual user mobile device said data resource responsive to said actual user request.

8. The computerized method of claim 7 wherein said proactive schedule comprises expiration times for downloading to said site said data resources.

9. The computerized method of claim 7 wherein generating a list of anticipated requests for data resources by analyzing said plurality of prior requests for data resources comprises clustering actual requests according to user interests.

10. The computerized method of claim 7 wherein generating a list of anticipated requests for data resources by analyzing said plurality of prior requests for data comprises clustering data resources based on similarity in topics.

11. The computerized method of claim 7, wherein notifying said individual user of the availability of at least one of said data resources downloaded to said site comprises identifying the data resource in results for a search performed by said individual user on said individual user mobile device.

12. The computerized method of claim 11, further comprising the step of providing said individual user with a monetary incentive for selecting said data resource identified in the search results.

13. The computerized method of claim 7, further comprising the step of providing said individual user with an incentive associated with requesting at least one of said data resources downloaded to said site.