Distributed cache—adaptive multicast architecture for bandwidth reduction

• US 9,130,889 B2
• Filed: 06/21/2013
• Issued: 09/08/2015
• Est. Priority Date: 09/04/2009
• Status: Active Grant

• Current Assignee: ViaSat Incorporated
• Original Assignee: ViaSat Incorporated
• Inventors: Mills, Raymond L., Hudson, Erwin C.
• Primary Examiner(s): Duong, Frank
• Application Number:
  US13/924,220
• Publication Number:
  US 20140025783A1
• Time in Patent Office: 809 Days
• Field of Search: 370/252, 3703955-39552, 370400-401, 370464-465, 370/468, 370/477
• US Class Current: 1/1
• CPC Class Codes: H04B 7/18523 Satellite systems for providing broadcast service to terrestrial stations, i.e. broadcast satellite service arrangements specially adapted for satellite broadcast receiving H04H40/90; picture transmission via satellite H04N1/00103; television transmission via satellite H04N7/20
• CPC Class Codes: H04L 12/1886 with traffic restrictions for efficiency improvement, e.g. involving subnets or subdomains
• CPC Class Codes: H04L 43/0894 Packet rate
• CPC Class Codes: H04L 47/806 Broadcast or multicast traffic
• CPC Class Codes: H04L 67/1097 for distributed storage of data in networks, e.g. transport arrangements for network file system [NFS], storage area networks [SAN] or network attached storage [NAS]

Abstract

Disclosed is a method and system for maximizing the use of available bandwidth on an ISP communication system between an Internet Service Provider (ISP) and remote locations where at least one of the remote locations has a remote cache. An embodiment may create a pool of the cacheable objects being sent to the remote locations from the downstream traffic. An embodiment may determine bandwidth savings for each object in the pool of cacheable objects that would be achieved by remotely caching each object and prioritize the pool of cacheable objects based on the determined bandwidth savings for each object. An embodiment may create a queue of objects to multicast to the remote caches based on the pool of cacheable objects and the remaining multicast bandwidth and then multicast the queue to the remote caches. The remote caches may intercept and reply to requests for objects held in the remote cache without accessing the ISP communication system, thus, saving bandwidth on the ISP communication system.

Claims

1. A method for communication by an Internet Service Provide (ISP) in a communication system, said ISP connected to at least one remote location having a remote cache via an ISP communication system, said method comprising:

identifying, at a computerized harvester of said ISP, a pool of cacheable objects based on downstream unicast communication traffic, said pool of cacheable objects being requested objects contained in downstream unicast replies of said downstream unicast communication traffic of said ISP communication system;

prioritizing, at a computerized prioritizer of said ISP, said pool of cacheable objects based at least in part on a bandwidth savings from remote caching of each requested object in said pool of cacheable objects;

determining a sub-group of said requested objects in said pool of cacheable objects to place in a queue of multicast cacheable objects to multicast to said remote cache at said at least one remote location based on said prioritized pool of cacheable objects and an available downstream multicast bandwidth of said ISP communication system; and

delivering objects in said multicast queue to said remote cache at said at least one remote location using downstream multicast transmissions via said ISP communication system based on said available downstream multicast bandwidth of said ISP communication system, said delivered objects for use by said remote cache in intercepting upstream requests from said at least one remote location for objects and responding to said intercepted upstream requests with replies containing corresponding delivered objects contained in said remote cache.

2. The method of claim 1, further comprising:

determining, at said computerized harvester, said bandwidth savings from remote caching of each requested object in said pool of cacheable objects based at least in part on a delivery cost/size of each requested object, a Time To Live (TTL)/expiry time of each requested object, or a frequency of request for each requested object, or a combination thereof.

3. The method of claim 2, wherein said determining of said bandwidth savings from remote caching for each requested object in said pool of cacheable objects is further a function of a number of remote caches connected to said ISP communication system, said number of remote caches indicating a number of remote locations with said remote cache connected to said ISP communication system.

4. The method of claim 2, wherein said determining of said bandwidth savings from remote caching for each requested object in said pool of cacheable objects is further a function of a maximum total downstream bandwidth currently available to said ISP communication system.

5. The method of claim 2, wherein said determining of said bandwidth savings from remote caching of each requested object further comprises:

applying a weighting factor to one or more of said delivery cost/size of each requested object, said Time To Live (TTL)/expiry time of each requested object, or said frequency of request for each requested object.

6. The method of claim 1, wherein said pool of cacheable objects excludes objects designated as not being cacheable, objects with a Time To Live (TTL)/expiry time determined to be too short to deliver to said remote cache at said at least one remote location, objects originating from a domain forbidden to be cached, or objects designated as undesirable by a content control filter actively monitoring said requested objects monitored in said downstream unicast communication traffic, or a combination thereof.

7. The method of claim 1 further comprising combining a predefined list of objects with said requested objects monitored in said downstream unicast communication traffic in said pool of cacheable objects.

8. The method of claim 1 further comprising:

examining said Time To Live (TTL)/expiry time of each requested object in said queue of multicast cacheable objects to determine if said Time To Live (TTL)/expiry time of each requested object in said queue of multicast cacheable objects is indeterminate; and

overwriting/creating said Time To Live (TTL)/expiry time in a requested object contained in said queue of multicast cacheable objects with a pre-defined replacement expiry time if said Time To Live (TTL)/expiry time of said requested object contained in said queue of multicast cacheable objects is found to have no expiration, negative expiration, or some otherwise indeterminable expiration.

9. The method of claim 1, wherein said ISP communication system connects to at least one additional remote location that does not have a remote cache.

10. The method of claim 1, wherein said at least one remote location restricts storage of objects in said remote cache of said at least one remote location to objects requested by said at least one remote location and not other remote locations.

11. A distributed cache adaptive multicast system for use with an Internet service provider (ISP) communication system connecting at least one remote location to an ISP, said at least one remote location having a remote cache, said distributed cache adaptive multicast system comprising:

a harvester subsystem that identifies a pool of cacheable objects based on downstream unicast communication traffic, said pool of cacheable objects being requested objects contained in downstream unicast replies of said downstream unicast communication traffic of said ISP communication system;

a prioritizer subsystem that prioritizes said pool of cacheable objects based at least in part on a bandwidth savings from remote caching of each requested object in said pool of cacheable objects;

a multicast queue subsystem that determines a sub-group of said requested objects in said pool of cacheable objects to place in a queue of multicast cacheable objects to multicast to said remote cache at said at least one remote location based on said prioritized pool of cacheable objects and an available downstream multicast bandwidth of said ISP communication system; and

a multicast delivery subsystem that delivers objects in said multicast queue to said remote cache at said at least one remote location using downstream multicast transmissions via said ISP communication system based on said available downstream multicast bandwidth of said ISP communication system, said delivered objects for use by said remote cache in intercepting upstream requests from said at least one remote location for objects and responding to said intercepted upstream requests with replies containing corresponding delivered objects contained in said remote cache.

12. The distributed cache adaptive multicast system of claim 11, wherein said harvester subsystem determines said bandwidth savings from remote caching of each requested object in said pool of cacheable objects based at least in part on a delivery cost/size of each requested object, a Time To Live (TTL)/expiry time of each requested object, or a frequency of request for each requested object, or a combination thereof.

13. The distributed cache adaptive multicast system of claim 12, wherein said harvester subsystem determines said bandwidth savings from remote caching for each requested object in said pool of cacheable objects further as a function of a number of remote caches connected to said ISP communication system, said number of remote caches indicating a number of remote locations with said remote cache connected to said ISP communication system.

14. The distributed cache adaptive multicast system of claim 12, wherein said harvester subsystem determines said bandwidth savings from remote caching for each requested object in said pool of cacheable objects further as a function of a maximum total downstream bandwidth currently available to said ISP communication system.

15. The distributed cache adaptive multicast system of claim 12, wherein said harvester subsystem determines said bandwidth savings from remote caching of each requested object further based on a weighting factor applied to one or more of said delivery cost/size of each requested object, said TTL/expiry time of each requested object, or said frequency of request for each requested object.

16. The distributed cache adaptive multicast system of claim 11, wherein said pool of cacheable objects excludes objects designated as not being cacheable, objects with a Time To Live (TTL)/expiry time determined to be too short to deliver to said remote cache at said at least one remote location, objects originating from a domain forbidden to be cached, or objects designated as undesirable by a content control filter actively monitoring said requested objects monitored in said downstream unicast communication traffic, or a combination thereof.

17. The distributed cache adaptive multicast system of claim 11, further comprising a predefined list subsystem that combines a predefined list of objects with said requested objects monitored in said downstream unicast communication traffic in said pool of cacheable objects.

18. The distributed cache adaptive multicast system of claim 11, further comprising:

an expiry update subsystem that examines said Time To Live (TTL)/expiry time of each requested object in said queue of multicast cacheable objects to determine if said Time To Live (TTL)/expiry time of each requested object in said queue of multicast cacheable objects is indeterminate; and

overwrites/creates said Time To Live (TTL)/expiry time in a requested object contained in said queue of multicast cacheable objects with a pre-defined replacement expiry time if said Time To Live (TTL)/expiry time of said requested object contained in said queue of multicast cacheable objects is found to have no expiration, negative expiration, or some otherwise indeterminable expiration.

19. The distributed cache adaptive multicast system of claim 11, wherein said ISP communication system connects to at least one additional remote location that does not have a remote cache.

20. A computer program product for distributed cache adaptive multicasting at an Internet service provider (ISP), said ISP connected to at least one remote location having a remote cache via an ISP communication system, comprising:

a non-transitory computer-readable medium, comprising code for;

identifying, at a computerized harvester of said ISP, a pool of cacheable objects based on downstream unicast communication traffic, said pool of cacheable objects being requested objects contained in downstream unicast replies of said downstream unicast communication traffic of said ISP communication system;

prioritizing, at a computerized prioritizer of said ISP, said pool of cacheable objects based at least in part on a bandwidth savings from remote caching of each requested object in said pool of cacheable objects;

determining a sub-group of said requested objects in said pool of cacheable objects to place in a queue of multicast cacheable objects to multicast to said remote cache at said at least one remote location based on said prioritized pool of cacheable objects and an available downstream multicast bandwidth of said ISP communication system; and

delivering objects in said multicast queue to said remote cache at said at least one remote location using downstream multicast transmissions via said ISP communication system based on said available downstream multicast bandwidth of said ISP communication system, said delivered objects for use by said remote cache in intercepting upstream requests from said at least one remote location for objects and responding to said intercepted upstream requests with replies containing corresponding delivered objects contained in said remote cache.