T2P Delay Buffer for DTMF Signal Packetising
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
In-band DTMF signalling leaks in packet-switched networks can cause double digit detection issues when voice paths traverse these networks, leading to inaccurate DTMF interpretation and voice compression limitations, as existing methods fail to reliably handle DTMF signals amidst packet jitter, delay, and loss.
Innovation Solution
Implementing a T2P delay buffer that initializes with a configurable silence pattern during call setup, allowing DTMF detection and injecting RFC4733 RTP packets when a digit is detected, thereby managing in-band DTMF leaks and reducing latency and jitter.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If DTMF signalling is carried in-band through a packet-switched network, then the DTMF can be transmitted through the network, but packet jitter, delay, and loss cause breaks in DTMF signals that lead to double digit detection or undetected digits
Solution Approach 1:
The patent segments the DTMF transmission path by extracting DTMF tones from the voice stream at the ingress gateway, transporting them separately as out-of-band signalling, and reconstructing them at the egress gateway. This separation isolates the DTMF signalling from packet jitter and delay affecting the voice path, preventing double digit detection while maintaining reliable digit transmission.
2Measurement precision
If DTMF detectors are configured to ignore short interruptions to prevent false detection, then false signal detection is reduced, but legitimate DTMF signals interrupted by packet loss are also ignored
Solution Approach 1:
The patent introduces an intermediary out-of-band DTMF signalling path that mediates between the voice path and the DTMF detector. The extracted DTMF tones are transported separately and injected after the voice path, allowing the DTMF detector to receive complete, uninterrupted signals even when the voice path experiences packet loss or jitter, thus maintaining both detection accuracy and signal delivery reliability.
3Reliability
If out-of-band DTMF signalling packets are injected after DTMF detection, then the DTMF can be transmitted reliably, but in-band DTMF leaks continue during the detection processing time
Solution Approach 1:
The patent applies preliminary action by extracting DTMF tones from the voice stream at the ingress gateway before transmission through the packet network. This pre-extraction prevents in-band DTMF leakage during detection processing time, as the DTMF components are removed from the voice path and transported separately as out-of-band signalling, eliminating the source of the harmful leak while maintaining reliable DTMF transmission.
4Productivity
If most compression algorithms are used for voice compression in packet-switched networks, then voice bandwidth is optimized, but DTMF signals cannot be detected reliably after decoding
Solution Approach 1:
The patent extracts DTMF tones from the voice stream at the ingress gateway before compression and packetization. By removing the DTMF components separately, the voice stream can be compressed using standard algorithms without affecting DTMF integrity. The extracted DTMF tones are transported out-of-band and reconstructed at the egress gateway, ensuring reliable detection while maintaining voice compression efficiency.
Data Source
Figure 1
Figure 2
Figure 3
AI summary
A T2P (TDM to packet) delay buffer is provided. The delay buffer can prevent double digit detections caused by in-band DTMF leak when out-of-band DTMF is used. The T2P delay buffer is initialized with an audio pattern that represents silence in a configurable amount of delay. When a DTMF digit is detected, the system can stop taking the voice payload from the T2P delay buffer and start injecting RFC4733 RTP packets into the RTP stream at a pre-configured rate. The RFC4733 DTMF RTP packets continue to be injected into the RTP stream until the DTMF digit stops. Once the end of the DTMF digit is detected, the content of the T2P delay buffer can be discarded and the T2P delay buffer is reinitialized with an audio pattern that represents silence in a configurable amount of delay. After the T2P delay buffer is reinitialized, the voice packetisation can be continued.