Method and system to bundle message over a network

• US 7,299,273 B2
• Filed: 12/16/2002
• Issued: 11/20/2007
• Est. Priority Date: 12/16/2002
• Status: Expired due to Fees

• Current Assignee: Google LLC (Alphabet Inc.)
• Original Assignee: International Business Machines Corporation
• Inventors: Krishnapuram, Raghuram, Kummamuru, Krishna, Kumar, Manoj
• Primary Examiner(s): Flynn; Nathan J.
• Assistant Examiner(s): WASEL, MOHAMED A
• Application Number:
  US10/319,966
• Publication Number:
  US 20040117469A1
• Time in Patent Office: 1,800 Days
• Field of Search: 709217-220, 709201-203
• US Class Current: 709/220
• CPC Class Codes: G06F 2209/505 Clust
• CPC Class Codes: G06F 9/5061 Partitioning or combining of resources

Abstract

The invention describes a method and system to optimize network bandwidth and obtain greater efficiency in transmission of messages/data in, a client-server network. The invention proposes the use of clustering of client requests and the data items in such a manner so as to optimize the network transmission as well as reduce the cost of processing involved in sending and picking/pruning the data items at server and client end respectively.

Claims

1. A method for minimizing an overall cost of processing client requests for items in a client-server network, said method comprising:

defining a range of values for a number of clusters into which requesting clients and requested items in said client requests from said requesting clients are to be clustered;

simultaneously clustering said requested items and said requesting clients into a set number of clusters, using a clustering algorithm, wherein said set number of clusters is a value within said range;

calculating a cost of processing said set number of clusters of said requested items and said requesting clients, wherein said cost is based on a difference between items received and items requested by said requesting clients;

repeating said processes of simultaneously clustering and calculating such that said cost is calculated for each of said values within said range; and

determining a cluster combination that provides a minimum overall cost of processing said client requests

wherein said process of simultaneously clustering said requested items and said requesting clients further comprises;

randomly initializing fuzzy membership values of said requested items and said requesting clients;

iteratively calculating said fuzzy membership values until a sum of changes of membership values is less than a configurable threshold value; and

defuzzifying said fuzzy membership values to obtain actual clusters of said requested items and said requesting clients.

2. The method as claimed in claim 1, further comprising assigning clusters of said requested items to respective clusters of said requesting clients.

3. The method as claimed in claim 1, wherein said requested items comprise at least one of data, information and messages requested by said requesting clients.

4. A system for minimizing an overall cost of servicing client requests for items, said system operable in a client-server network and comprising:

a selector operable for defining a range of values for a number of clusters into which requesting clients and requested items in said client requests from said requesting clients are to be clustered;

an organizer operable for simultaneously clustering said requested items and said requesting clients into a set number of clusters, using a clustering algorithm, wherein said set number of clusters is a value within said range and wherein said clustering algorithm is adapted to;

randomly initialize fuzzy membership values of said requested items and said requesting clients;

iteratively calculate said fuzzy membership values until a sum of changes of membership values is less than a configurable threshold value; and

defuzzify said fuzzy membership values to obtain actual clusters of said requested items and said requesting clients,;

a calculator operable for establishing a total processing cost for said set number of clusters of said requested items and said requesting clients, wherein said cost is based on a difference between items received and items requested by said requesting clients and wherein said organizer and said calculator are further adapted to repeat said simultaneous clustering and said establishing of said total processing cost for each of said values within said range; and

a component operable for establishing a minimum overall cost of processing said client requests.

5. The system as claimed in claim 4 further comprising a device operable for assigning clusters of said requested items to respective clusters of said requesting clients.

6. The system as claimed in claim 4, wherein said requesting clients comprise request generating computing devices that are connected to at least one other computing device, wherein said at least one other computing device is adapted to function as a servers servicing requests made by said requesting clients.

7. The system as claimed in claim 4, wherein said requested items comprise at least one of information, messages, and data requested by said requesting clients.

8. The system as claimed in claim 5, wherein said system resides on at least one server device:

a system bus;

a communications unit connected to said system bus;

a memory unit including a set of instructions connected to said system bus; and

a control unit executing said instructions in said memory unit.

9. The system as claimed in claim 6, wherein said requesting clients are interconnected to said system through a computer network comprising at least one of an Ethernet, an Internet, a LAN, a WAN, and a MAN.

10. A computer program product comprising computer readable program code stored on a computer readable storage medium embodied therein for performing a method for minimizing an overall a total cost of processing client requests for items in a client-server network, said method comprising:

defining a range of values for a number of clusters into which requesting clients and requested items in said client requests from said requesting clients are to be clustered;

simultaneously clustering said requested items and said requesting clients into a set number of clusters, using a clustering algorithm, wherein said set number of clusters is a value within said range;

calculating a total cost of processing for said set number of clusters of said requested items and said requesting clients, wherein said cost is based on a difference between items received and items requested by said requesting clients;

repeating said processes of simultaneously clustering and calculating such that said cost is calculated for each of said values within said range; and

determining a minimum overall cost of processing said client requests,

wherein said process of simultaneously clustering said requested items and said requesting clients further comprises;

randomly initializing fuzzy membership values of said requested items and said requesting clients;

iteratively calculating said fuzzy membership values until a sum of changes of membership values is less than a configurable threshold value; and

defuzzifying said fuzzy membership values to obtain actual clusters of said requested items and said requesting clients.

11. The computer program product as claimed in claim 10, said method further comprising assigning clusters of said requested items to respective clusters of said requesting clients.

16. The computer program product as claimed in claim 10, wherein said requested items comprise at least one of data, information and messages requested by said requesting clients.

12. A method of bundling messages over a client-server network to minimize an overall cost of processing client request for items in said client-server network, said method comprising:

clustering said incoming client requests from requesting clients into clusters of requested items and clusters of said requesting clients, wherein said process of clustering incoming client requests into clusters comprises;

randomly initializing fuzzy membership values of said requested items and said requesting clients;

iteratively calculating said fuzzy membership values until a sum of changes of membership values is less than a configurable threshold value; and

defuzzifying said fuzzy membership values to obtain actual clusters of said requested items and said requesting clients;

formulating costs associated with serving said incoming client requests in said network, wherein said costs are based on a difference between items received and items requested by said requesting clients; and

determining a cluster combination which provides a minimum overall cost of processing said incoming client requests.

13. The method of claim 12, further comprising assigning clusters of said requested items to respective clusters of said requesting clients.

14. The method of claim 12, further comprising defining a range of values for a number of clusters into which said requesting clients and said requested items are to be clustered prior to said process of clustering.

15. The method of claim 12, wherein said requested items comprise at least one of data, information and messages requested by said requesting clients.