Abstract: There is provided a method for use by an echo canceller to cancel a far echo at a variable delay and a near echo at a fixed delay. The method comprises constructing an echo signal model based on an incoming signal, determining a variable echo delay for a far echo caused by a far echo source, determining a fixed echo delay for a near echo caused by a near echo source, subtracting the echo signal model from an outgoing signal at a window placed around the variable echo delay to cancel far echo, e.g. when the echo canceller determines existence of the far echo, and subtracting the echo signal model from the outgoing signal at a window placed around the fixed echo delay to cancel near echo, e.g. regardless of existence of the near echo, wherein the fixed echo delay is smaller than the variable echo delay.

Abstract: There is provided system and method for receiving speech signals via a plurality of speech lines and communicating the speech signals over a phone line. For example, the communication system comprises a modem, a multiplexor in data communication with the modem, and a plurality of speech codecs in data communication with the multiplexor. According to this aspect, each of the plurality of speech codecs receives a portion of the speech signals via one of the plurality of speech lines and encodes that portion of the speech signals to generate encoded speech signals, and wherein the encoded speech signals from each of the plurality of speech codecs are multiplexed by the multiplexor to generate multiplexed encoded speech signals, and wherein the modem transmits the multiplexed encoded speech signals over the phone line.

Abstract: Various media stream processing systems and methods are provided. In one exemplary embodiment, a media stream processing method is provided that includes the brokering of a number of communications channels between a communications network and a packet switched network with a number of media stream processors. Each of the communications channels is specified by a profile having a number of channel parameters. The method also includes storing each of the profiles in a standby media stream processor when each of the media stream processors is operational. Finally, the method includes brokering respective ones of the communications channels originally brokered by one of the media stream processors with the standby media stream processor upon a failure of the one of the media stream processors.

Abstract: In exemplary embodiments, a dataport in an access server is suitably configured to emulate a modern connection for non-modem calls such as voice calls. Computer applications for administering, for example, voice-over-network services thus communicate with the port via standard modem calls. In accordance with one aspect of the invention, communications between the port hardware and the voice enabled application take place through an encapsulating protocol such as the Point-to-Point-Protocol (PPP)such that the port is addressable and able to receive controls or other instructions from a voice application residing on the access server or from a distributed location.