Nortel Networks Corporation

Rockstar Consortium US LP

Mohamed Samy El-Hennawey

Rafik Alphonse Goubran

Ick Don Lee

Ayman M Radwan

Lijing Ding

John Cornelius Lynch

Marc Saunders

Masoud Ahmadi

Zhihong Qian

Joachim Fouret

Can Liao

Saeed Gazor

Provided is a method and apparatus for objectively and non-intrusively measuring voice quality on live calls without disrupting the call session or the network. A communication system includes plural communities each including a switch that controls access to a packet-based data network for call sessions. Each of the communities is coupled to the data network by respective packet-based trunks. Quality of service (QoS) monitoring devices are coupled to the respective packet-based trunks to monitor quality levels of routes between any two given communities. Each QoS monitoring device receives packets containing streaming data (which may be actual packets or test packets). From the received packets, the QoS monitoring device can derive QoS parameters, particularly for audio and speech signals on live calls without disrupting the call session.

Non-intrusive monitoring of quality levels for voice communications over a packet-based network

An apparatus (), method, and computer program to assess VoIP speech quality () using access to degraded signals is provided. Different types of impairment () have different effect, on speech quality. Preferred embodiments address up to four different types of impairment that affect VoIP signal quality: packet loss (), speech clipping in time (), noise () and echo. An overall assessment algorithm factors in degradation due to various impairment factors to generate an overall speech quality assessment score or value.

Method and apparatus for non-intrusive single-ended voice quality assessment in VoIP

An apparatus (), method, and computer program to assess VoIP speech quality () using access to degraded signals. Different types of impairment () have different effect on speech quality. Preferred embodiments address two distinct types of impairment that are unique to VoIP: packet loss () and speech clipping in time ().

Transcoders and mixers having reduced algorithmic delay and processing complexity. An improved mixer for signals having encoded speech parameters wherein the parameters obtained through decoding are used by a parameter estimator to improve the encoding by providing a parameter estimate for the mixed signal. In the case of pitch parameters, the mixer uses the principle of strong-pitch-domination. The mixing of wideband signals is simplified by performing mixing of individual lower and upper sub-bands. A transcoder and a mixer that converts a wideband signal into a narrowband signal relies upon high frequency suppression. A transcoder and a mixer that converts a narrowband signal into a wideband signal relies upon filter combination.

Transcoders and mixers for voice-over-IP conferencing

A fast DTMF detector for filtering a packetized linear voice signal to identify whether the voice signal contains a DTMF tone. The detector includes a plurality of parallel notch filters for knocking down DTMF tones. The energies of the filtered signals output by the notch filters are calculated and criteria based upon the calculated energies are applied to determine if the voice signal contains a DTMF tone. The criteria include an energy differential test, a twist test, a low frequency tolerance test, a high frequency tolerance test, a packetized linear voice signal energy level test, a dial tone filtered signal energy level test, and a 1004 Hz filtered signal energy level test.

Method and apparatus for fast DTMF detection

A method for discriminating noise from signal in a noise-contaminated signal involves decomposing a frame of samples of the signal into decorrelated components, and using a difference between probability distributions of the noise contributions and the signal contributions to identify signal and noise. A Gaussian distribution is used to determine whether the components are only noise whereas a Laplacian distribution is used to determine whether the components contain the signal. Such discrimination may be used in speech enhancement or voice activity detection apparatus.

Method and system for speech processing for enhancement and detection

A given DSP, when processing signals of a channel, stores channel instance data in a multiplexed memory in association with itself. If the channel processing task changes to one for which the given DSP is not optimised, the given DSP moves the channel to an appropriate new DSP. Additionally, the given DSP transfers the association of the channel instance data to an association with the new DSP. In this way, the new DSP is provided with the channel instance data without the requirement for physically transferring such data.

Growing community of inventors

Belleville, Canada

Mohamed Samy El-Hennawey