Methods and apparatus for predictive capacity allocation

• US 20080273591A1
• Filed: 05/04/2007
• Published: 11/06/2008
• Est. Priority Date: 05/04/2007
• Status: Active Grant

• Current Assignee: Time Warner Cable Enterprises LLC (Charter Communications Incorporated)
• Original Assignee: Time Warner Cable Enterprises LLC (Charter Communications Incorporated)
• Inventors: Rieger, Remi, Brooks, Paul D.
• Granted Patent: US 9,398,346 B2
• Time in Patent Office: Days
• US Class Current: 375/240.01
• CPC Class Codes: H04L 41/0896 Bandwidth or capacity management, i.e. automatically increasing or decreasing capacities flow or congestion control using dynamic resource allocation, e.g. in-call renegotiation, H04L47/76
• CPC Class Codes: H04L 41/147 for predicting network behaviour
• CPC Class Codes: H04N 21/2543 Billing , e.g. for subscription services payment schemes, architectures or protocols G06Q20/00; e-commerce G06Q30/00; arrangements for billing for the use of broadcast information or broadcast-related information H04H60/21
• CPC Class Codes: H04N 21/4181 for conditional access
• CPC Class Codes: H04N 21/47202 for requesting content on demand, e.g. video on demand
• CPC Class Codes: H04N 21/6118 involving cable transmission, e.g. using a cable modem
• CPC Class Codes: H04N 21/6168 involving cable transmission, e.g. using a cable modem
• CPC Class Codes: H04N 21/6373 for rate control , e.g. request to the server to modify its transmission rate flow control in packet networks H04L47/10
• CPC Class Codes: H04N 7/1675 Providing digital key or authorisation information for generation or regeneration of the scrambling sequence pseudo-random number generators in general G06F7/58
• CPC Class Codes: H04N 7/17318 Direct or substantially direct transmission and handling of requests

Abstract

Methods and apparatus for dynamically adjusting capacity allocation; e.g., to a group of services in a multimedia distribution network. In one embodiment, bandwidth allocation is adjusted by predicting bandwidth utilization of the group of services based on historical bandwidth utilization information. Behavioral templates and statistical models may also be employed for the predictions. The invention provides more efficient use of available bandwidth compared to conventional bandwidth allocation schemes where the overall bandwidth for a group of services is set to a constant value that is not changed frequently. The present invention further provides a bandwidth allocation method in a switched distribution network. The bandwidth allocation method allocates bandwidth such that the bandwidth required by the number of services does not exceed the allocation, i.e., the total usage falls within the allocation.

Claims

1. A method of operating a cable television network, comprising:

obtaining first data relating to the operation of said network over a period of time;

identifying at least one correlation between said first data and a future context; and

projecting at least one parameter for said future context based at least in part on said first data.

2. The method of claim 1, wherein said network comprises a broadcast switched network, and said first data comprises data indicating actual bandwidth demand within at least a portion of said period.

3. The method of claim 2, wherein said future context comprises a future period of time.

4. The method of claim 3, wherein said at least one correlation comprises a correlation between said period of time of said first data and said future period of time.

5. The method of claim 4, wherein said correlation between said period of time of said first data and said future period of time comprises the period and the future period being the same day of the week.

6. The method of claim 4, wherein said correlation between said period of time of said first data and said future period of time comprises the period and the future period being the same portion of a 24-hour day.

7. The method of claim 4, wherein said correlation between said period of time of said first data and said future period of time comprises the period and the future period being the same holiday.

8. The method of claim 1, wherein:

said identifying at least one correlation between said first data and a future context comprises matching at least a first portion of a pattern reflected in said first data to a pattern observed in a present context; and

said projecting comprises projecting based on a second portion of said pattern.

9. Apparatus for allocating bandwidth in a content-based network having at least one server, the apparatus comprising:

a computerized device; and

at least one management process capable of being run on said computerized device, said process being operative to;

collect data relating to historical bandwidth demand;

project, based at least in part on said collected data, bandwidth demand at some period in the future; and

allocate bandwidth within said network based at least in part on said projection.

10. The apparatus of claim 9, further comprising a database in data communication with said computerized device, said database comprising a plurality of said data.

11. The apparatus of claim 10, further comprising a substantially computerized process adapted to update said database with recently acquired data at least periodically, said recently acquired data relating to actual bandwidth demand within said network.

12. The apparatus of claim 9, further comprising a substantially computerized process adapted to, during or after said future period has elapsed, compare said projection of bandwidth demand with actual demand relating to said future period, and determine differences there between.

13. The apparatus of claim 12, wherein said substantially computerized process is further adapted to cause adjustment to said management process based at least in part on said differences.

14. The apparatus of claim 13, wherein said adjustment to said management process based at least in part on said differences comprises adjusting the frequency or periodicity with which said projection is performed.

15. The apparatus of claim 13, wherein said adjustment to said management process based at least in part on said differences comprises changing one or more assumptions on which said projection is based.

16. The apparatus of claim 13, wherein said adjustment to said management process based at least in part on said differences comprises switching from a first behavioral template to a second behavioral template.

17. The apparatus of claim 9, wherein said network comprises a cable television network, and said computerized device comprises a server disposed at a headend of said network.

18. The apparatus of claim 9, wherein said network comprises a broadcast switched architecture (BSA) cable television network, and said computerized device comprises a device disposed at a switching hub of said network.

19. The apparatus of claim 9, wherein said projection based at least in part on said data comprises correlating at least one feature within said collected data to said period in the future.

20. The apparatus of claim 19, wherein said at least one feature within said collected data comprises a particular time period of a particular day in the week, and said period in the future comprises the same time period of the same day of the week.

21. The apparatus of claim 19, wherein said at least one feature within said collected data comprises a particular holiday, and said period in the future comprises the same holiday.

22. A method of operating a cable television network, comprising:

providing a template describing at least one aspect of the operation of said network over a period of time;

correlating said template and a future context; and

projecting at least one parameter for said future context based at least in part on said template.

23. The method of claim 22, wherein said cable television network comprises a broadcast switched architecture network, and said providing a template describing at least one aspect comprises providing data indicating actual bandwidth demand over at least a portion of a historical period.

24. The method of claim 23, wherein said future context comprises a future period of time.

25. The method of claim 24, wherein said correlating comprises a correlation between said historical period and said future period of time.

26. The method of claim 25, wherein said correlation between said historical period and said future period of time comprises the historical period and the future period comprising the same day of the week.

27. The method of claim 25, wherein said correlation between said historical period and said future period of time comprises the historical period and the future period comprising the same portion of a 24-hour day.

28. The method of claim 25, wherein said correlation between said historical period and said future period of time comprises the historical period and the future period comprising the same holiday.

29. The method of claim 25, wherein said template describing at least one aspect comprises data indicating actual bandwidth demand over at least a portion of a historical period, said data having a given pattern or shape, and said act of projecting comprises generating a projection of bandwidth demand that substantially conforms to said pattern or shape.

30. The method of claim 23, wherein said data indicating actual bandwidth demand over at least a portion of a historical period has a given pattern or shape, and said act of projecting comprises generating a projection of bandwidth demand that substantially conforms to said pattern or shape.

31. A method of operating a cable network, comprising:

generating a projection of the capacity requirements for a plurality of services requested for delivery over said network, said projection being based at least in part on a model;

allocating capacity within the network based at least in part on the projection;

evaluating said projection based at least in part on operational data; and

determining whether adjustment of said model is required.

32. The method of claim 31, wherein said cable network comprises a broadcast switched architecture, and said plurality of services are requested by a plurality of customer premises devices of said network.

33. The method of claim 32, wherein said services comprise streaming or download requests for content.

34. The method of claim 31, wherein said method further comprises adjusting said model based at least in part on said determining.

35. The method of claim 34, wherein said adjusting comprises adjusting in an incremental fashion utilizing a feedback mechanism so as to avoid substantial overshoot.

36. The method of claim 34, wherein said model comprises at least one behavioral template, and said adjusting comprises utilizing a different template.

37. The method of claim 31, wherein evaluating said projection based at least in part on operational data comprises:

determining a difference between actual capacity required at a given time and the projected capacity for that given time; and

comparing said difference to at least on one criterion.

38. The method of claim 37, wherein said determining whether an adjustment is required comprises determining if said at least one criterion is met;

and

said method further comprises, if said criterion is met, adjusting said model.

39. The method of claim 31, wherein said generating a projection comprises generating a function describing projected bandwidth requirement as a function of future time.

40. The method of claim 31, wherein said generating a projection comprises determining clamped bitrates associated with at least some of said services.

41. A method of allocating bandwidth to a group of programs in a network, comprising:

gathering data from a program source, said data relating to bandwidth requirements over a time interval for at least one program in said group of programs ingested in said network by said source,

storing said data into a database;

subsequently analyzing at least said data to identify one or more features of significance therein;

using at least one of said features to determine a value of at least one operational parameter used in operating said network; and

basing an allocation of bandwidth to said group of programs in said network at least in part on said value.

42. The method of claim 41, wherein said network comprises a cable television network, and said source comprises a program source ingress (PSI) server disposed at a headend of said network.

43. The method of claim 41, wherein said act of analyzing to identify one or more features comprises identifying at least one correlation between a historical event and a recently observed event.

44. The method of claim 41, wherein said act of analyzing data from said database comprises statistically analyzing data corresponding to a plurality of said programs.

45. Apparatus for allocating bandwidth in a content distribution network having at least one server and a plurality of consumer premises equipment (CPE), comprising:

a first network side process running on said at least one server, said first network side process configured to monitor bandwidth required to distribute content to a subset of said CPE;

a second network side process running on said at least one server, said second network side process configured to predict bandwidth required to distribute said content to said subset of said CPE at a future time;

a third network side process running on said at least one server, said third network side process configured to allocate bandwidth to distribute content to said subset of said CPE at said future time;

wherein said second network side process comprises at least one computer program adapted to analyze past bandwidth utilization by said subset of said CPE, and base said prediction at least partly thereon.

46. The apparatus of claim 45, wherein said network comprises a cable television network, said CPE comprises set-top boxes, and said server comprises a video-on demand (VOD) server.

47. The apparatus of claim 45, wherein said network comprises a cable television network, and said server comprises a broadcast switched architecture (BSA) server.

48. The apparatus of claim 45, further comprising a historical database in data communication with said at least one server.

49. A method of operating a cable network, comprising:

receiving a plurality of requests for content for delivery, said requests comprising at least two different types of delivery;

determining the type of delivery requested for at least a portion of said requests; and

scheduling delivery of content for at least a portion of said requests of a first type in the future.

50. The method of claim 49, wherein said at least two different types of delivery comprise:

(i) real-time or streaming delivery; and

(ii) trickle or delayed download.

51. The method of claim 50, wherein said trickle or delayed download comprises said first type.

52. The method of claim 49, wherein said scheduling delivery comprises scheduling based at least in part on a prediction of future bandwidth demand.

53. The method of claim 49, wherein said scheduling delivery comprises scheduling based at least in part on a difference between current actual bandwidth demand and capacity.

54. The method of claim 49, wherein said scheduling delivery of content for at least a portion of said requests of a first type in the future is used as a basis for predicting future bandwidth demand.

55. A method of doing business within a content-based network, said network being adapted to selectively provide program streams, and having a server and a plurality of subscribers, said subscribers each having a client device coupled to said network, the method comprising offering at least a portion of said subscribers an incentive to utilize opportunistic capacity present in said network, said opportunistic capacity being related to a difference between a predicted demand for a resource and an actual demand for said resource at a given point or period of time.

56. The method of claim 55, wherein said resource comprises downstream bandwidth, and said actual demand is related to a measured instantaneous bandwidth, and said predicted demand is based at least in part on historical data.