Load balancing based upon speech processing specific factors

Inactive Publication Date: 2007-06-21
IBM CORP
View PDF17 Cites 26 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Benefits of technology

[0009] The present invention can be implemented in accordance with numerous aspects consistent with material presented herein. For example, one aspect of the present invention can include a load balancer that uses one or more plug-in receptors linked to one or more speech utilization algorithms. The speech utilization algorithms can calculate a voice server load based upon a speech processing factors. The load balancer can use these calculations to route speech processing requests to voice servers having available capacity for handling the requests.
[0010] The plug-in receptor can comply with a known industry standard and voice server administrators can use standard interfaces to configure speech utilization algorithms. For example, the industry standard can be a JAVA 2 ENTERPRISE EDITION (J2EE) based standard. The plug-in receptor can conform to the JAVA Connector Architecture (JCA). The load balancer can be a WEBSPHERE Application Server (WAS) Edge Server Load Balancer. One or more voice servers that are managed by the load balancer can be WEBSPHERE Voice Servers. The industry standard and compliant components are provided for illustrative purposes only and the invention is not to be construed as limited in this regard.
[0011] Another aspect of the present invention is a computer based method for automatically selecting voice servers to handle speech processing requests based upon a speech utilization load. A computer queries one or more voice servers for speech processing load data. A speech utilization score can be calculated for each queried voice server so that each speech processing request is routed to the least loaded server as determined by the results of the query.
[0012] Yet another aspect of the present invention can include a system for handling speech processing request. The system can include two or more voice servers and at least one load balancer. The voice servers can handle speech processing requests, such as ASR requests and TTS requests. Each of the voice servers can include a speech load balancer servlet and a J2EE compliant resource adaptor. The resource adaptor can include a load balancing algorithm that computes a load for an associated server based upon at least one speech processing specific factor. The speech load balancer servlet can finction as an interface between the load balancer and the resource adaptor.

Problems solved by technology

Most load balancers in use today fail to take into consideration server-specific factors that differentiate one voice server from another when assigning requests to request handling servers.
Additionally, no known voice server uses an industry standard load balancer in any fashion.

Method used

the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
View more

Image

Smart Image Click on the blue labels to locate them in the text.
Viewing Examples
Smart Image
  • Load balancing based upon speech processing specific factors
  • Load balancing based upon speech processing specific factors
  • Load balancing based upon speech processing specific factors

Examples

Experimental program
Comparison scheme
Effect test

Example

EXAMPLE 1

Dynamic Factor Computation

[0045] Algorithm 300 computes a speech utilization load based upon a number of sessions, an arrival rate, a duty cycle, and a prediction. A duty cycle percentage equals a number of seconds in a session (ASR or TTS) divided by the number of seconds for a total session. The number of seconds per second and the total seconds per session are statistics that are often maintained by a voice server.

[0046] An ASR utilization equals a number of sessions per second times ASR duty cycle times prediction period, plus the number of ASR engines in use divided by the number of configured ASR engines. The TTS utilization equals a number of sessions per second times TTS duty cycle times prediction period, plus the number of ASR engines in use divided by the number of configured TTS engines.

[0047] An example of an ASR utilization based upon algorithm 300 is presented below: 1)⁢ ⁢Seconds⁢ ⁢ASR⁢ ⁢in⁢ ⁢Session=12⁢ ⁢seconds⁢ ⁢(determined⁢ ⁢by⁢ ⁢system⁢ ⁢stats)⁢⁢2)⁢ ⁢S

Example

EXAMPLE 2

ASR Utilization Computation

[0048]FIG. 4 is a flow chart of a method 400 for implementing speech utilization algorithms in accordance with an embodiment of the inventive arrangements disclosed herein. The method 400 can be performed in the context of a load balancer that selects one of multiple voice servers to handle incoming client speech processing requests. In one embodiment, the method 400 can be performed in the context of system 100. The speech utilization algorithm of method 400 can be implemented as algorithm 200, as algorithm 300, or as any algorithm that determines a speech utilization score or speech utilization level using one or more speech processing specific factors.

[0049] Method 400 can begin in step 405, where one or more speech utilization algorithms can be installed and configured for one or more voice servers In one embodiment, step 405 can be performed using an administration console that deploys a resource adaptor containing the speech utilization algo

the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
Login to view more

PUM

No PUM Login to view more

Abstract

A machine readable storage can include a set of instructions for load balancing. The storage can include a plug-in receptor of a load balancer. The plug-in receptor can be compliant with a known industry standard and can be is associated with a two or more load balancing algorithms. The load balancer can utilize selected ones of the load balancing algorithms to determine which of two or more voice servers are to handle incoming speech processing requests. Selected ones of the load balancing algorithms can include a speech utilization algorithm. The speech utilization algorithm can calculate a speech utilization score for at least one of the voice servers based upon speech processing specific factors.

Description

the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
Login to view more

Claims

the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
Login to view more

Application Information

Patent Timeline
no application Login to view more
Owner IBM CORP
Who we serve
  • R&D Engineer
  • R&D Manager
  • IP Professional
Why Eureka
  • Industry Leading Data Capabilities
  • Powerful AI technology
  • Patent DNA Extraction
Social media
Try Eureka
PatSnap group products