Call-Based Audio Sockets for Voice Server Latency Reduction
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current telecommunications voice servers face processing delays due to dynamic audio socket allocation and frequent communication of socket information between the T&M subsystem and speech engines, especially under high loads, as they use turn-based speech engine allocation rather than call-based allocation, leading to bottlenecks in componentized architectures.
Innovation Solution
Establishing call-based audio sockets within a media converting component of the voice server, where audio sockets remain available for the duration of a call, allowing continuous communication between the media converter and speech engines without the need for frequent socket re-establishment, thereby reducing latency and bottlenecks.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If turn-based speech engine allocation is used, then speech engine resource utilization is maximized, but processing delays and bottlenecks increase due to frequent socket re-establishment and information conveyance
Solution Approach 1:
The patent applies preliminary action by pre-establishing audio sockets and conveying socket information to the T&M subsystem before speech processing begins. This allows the communication path to be ready in advance, eliminating the need for frequent socket re-establishment during turn-based processing, thereby reducing processing delays while maintaining high speech engine utilization
Solution Approach 2:
The patent introduces a socket information cache as an intermediary component that stores and manages audio socket information. This intermediary allows the T&M subsystem to quickly retrieve socket information without direct real-time communication with speech engines, reducing bottlenecks and processing delays in the componentized architecture
2Adaptability or versatility
If dynamic audio socket allocation is performed for each turn, then speech engine flexibility is maintained, but communication bottlenecks increase due to frequent socket information conveyance between T&M subsystem and speech engines
Solution Approach 1:
The patent merges the socket management function into the T&M subsystem by having it establish and cache socket information centrally. This consolidation eliminates the need for repeated socket establishment communications between the T&M subsystem and multiple speech engines, reducing communication overhead while maintaining the flexibility to allocate different speech engines to different turns
3Loss of time
If call-based speech engine allocation is used, then processing delays are reduced through stable communication paths, but speech engine resource utilization decreases due to 1-1 mapping
Solution Approach 1:
The patent segments the allocation strategy by applying call-based socket establishment for the communication path stability while maintaining turn-based speech engine allocation for resource optimization. The audio sockets are established at the call level to provide stable communication paths, while speech engines can still be dynamically allocated to different turns, achieving both reduced processing delays and high resource utilization
Data Source
AI summary
A method of interfacing a telephone application server and a speech engine can include the step of establishing one or more audio sockets in a media converting component of the telephone application server. The audio socket can remain available for approximately a duration of a call. A work unit that requires processing by a speech engine can be detected for the call. An identifier for the audio socket and a data for the work unit can be conveyed to a selected speech engine. Work unit results from the selected speech engine can be received by the media converting component via the previously established audio socket.