CACHE MANAGER FOR SEGMENTED MULTIMEDIA AND CORRESPONDING METHOD FOR CACHE MANAGEMENT

• US 20140032849A1
• Filed: 02/02/2012
• Published: 01/30/2014
• Est. Priority Date: 02/07/2011
• Status: Active Grant

• Current Assignee: RPX Corporation
• Original Assignee: Alcatel-Lucent SA (Nokia Corporation)
• Inventors: Huysegems, Rafal, De Vleeschauwer, Bart, Wu, Tingyao
• Granted Patent: US 9,418,012 B2
• Time in Patent Office: Days
• US Class Current: 711/133
• CPC Class Codes: G06F 12/0862 with prefetch
• CPC Class Codes: G06F 12/0891 using clearing, invalidating or resetting means
• CPC Class Codes: G06F 16/9574 of access to content, e.g. by caching

Abstract

The invention concerns a cache manager (100) for managing the intermediate caching of segmented multimedia items. Multiple versions are available of each multimedia item, each version representing the multimedia item with a different quality. The cache manager (100) comprises control means (101, 102) to control the pre-fetching and eviction of segments. The control means (101, 102) are at least responsive to temporal and quality related inter-segment relationships (131).

Claims

1. A cache manager for managing the intermediate caching of segments of multimedia items of which multiple versions are available, each version representing a multimedia item with a different quality,

wherein said cache manager comprises control means to control pre-fetching of segments of said multimedia items into a cache memory and eviction of segments of said multimedia items from said cache memory, said control means being responsive to information indicative for temporal and quality related inter-segment relationships between segments of said multiple versions of said multimedia items.

2. A cache manager according to claim 1,

comprising means to extract said information from one or more multimedia presentation description file related to said multimedia items.

3. A cache manager according to claim 1,

wherein said control means are further adapted to be responsive at start-up to popularity information of said multimedia items.

4. A cache manager according to claim 1,

wherein said control means are further adapted to be responsive to one or more of the following:

user profile information;

segment popularity information;

user history information;

session history information;

content history information;

content information supplied by a content owner/producer;

network load between a multimedia server and said cache memory;

network load between said cache memory and end user equipment;

client side algorithm information.

5. A cache manager according to claim 1, further comprising means for machine learning said temporal and quality related inter-segment relationships.

6. A cache manager according to claim 1,

wherein said control means are further adapted to maintain and/or assign values to segments stored in said cache memory, each value representing a likelihood that a corresponding segment of a version of a multimedia item will be requested, considering said temporal and quality related inter-segment relationships.

7. A cache manager according to claim 1,

further comprising re-direction means adapted to instruct re-direction of a multimedia session from a first cache node to a second cache node, said re-direction means being responsive at least to network topology information mapping clients to one or more caching nodes.

8. A method for managing the intermediate caching of segments of multimedia items of which multiple versions are available, each version representing a multimedia item with a different quality,

wherein said method comprises controlling pre-fetching of segments of said multimedia items into a cache memory and eviction of segments of said multimedia items from said cache memory in response to information indicative for temporal and quality related inter-segment relationships between segments of said multiple versions of said multimedia items.

9. A method according to claim 8,

further comprising re-directing a multimedia session from a first cache node to a second cache node thereby using at least network topology information mapping clients to one or more caching nodes.