Telephonic communications – Echo cancellation or suppression – Disable or inhibit function
Reexamination Certificate
1999-08-31
2003-06-17
Tieu, Binh (Department: 2643)
Telephonic communications
Echo cancellation or suppression
Disable or inhibit function
C379S406010, C379S406080, C370S352000, C375S345000
Reexamination Certificate
active
06580793
ABSTRACT:
FIELD OF THE INVENTION
This invention relates to telecommunications systems, and more particularly to echo cancellation systems used in transmission, switching, and other components of telecommunications systems, including packet-based and Internet voice networks.
BACKGROUND OF THE INVENTION
Echo remains a significant problem in voice and certain other telecommunications systems that incorporate physically lengthy transmission paths or other sources of significant transmission or propagation delay. Echo is typically caused when a signal generated by a “talker” and transmitted from a first end of a communications link is partially regenerated at a second end and returned to the first end of the link. By convention in the telecommunications arts, the end of the link containing the echo source—that is, the second end—is considered the “near” end of the link, and the talker is located at the “far” end of the link. The regenerated, or “echo” signal is received by the talker at the far end of the link, and can degrade the talker's perception of naturally generated speech from the near end. Echo may occur even when the communications link is formed from two isolated unidirectional communications paths operating in opposite directions, because devices at the ends of the link (or elsewhere) may receive a signal on one path, and transmit a regenerated by-product on the other path.
Depending on the amplitude and delay of the echo signal and its similarity to the original signal, the echo signal may be more or less noticeable to a user. When the by-product or “echo” signal has significant amplitude, and is delayed by more than about 20-30 mS, the echo signal may be sufficiently distracting as to make conversation difficult.
Echo may be produced in a number of ways. In conventional telecommunications transmission plants, a typical producer of echo has been the “hybrids” used as converters between four-wire transmission facilities and two-wire loops. Despite excellence in hybrid design, some leakage nonetheless occurs from the inbound path to the outbound path. When leakage occurs at a point “close” to the talker, the echo signal typically arrives with so little delay that it is neither noticeable or disturbing. However, when leakage occurs at the near end, such as at the near-end hybrid, the arrival of the echo signal at the far end may be significantly delayed due to the physical length of the transmission path and certain other network components. In that case, the echo signal may be noticeable; in some cases, the echo signal is so disturbing that conversation is difficult.
Several other network devices may also cause delay, even where physical path lengths are relatively short. For example, in modern mobile or wireless telephone systems and Internet voice systems incorporate voice coding devices (sometimes known as “vocoders”), which may introduce a significant delay. Even as telecommunications providers convert the world's telecommunications networks from analog to digital technologies, and despite the continual improvement in the performance of network components, some existing sources of echo remain, and new ones are created.
A variety of systems have been developed to minimize the effect of echo on the quality of communications service provided. When long-distance telecommunications systems were dominated by analog transmission facilities, the characteristics of individual transmission paths were carefully engineered to insert a controlled amount of attenuation. The attenuation was intended to reduce the amplitude of the echo signal such that it was not noticeable to the user. Although this system worked relatively well, it was primarily applicable to analog transmission facilities, and required significant continuing maintenance efforts and expense to adjust the attenuation levels to their engineered values.
Other systems have been developed to eliminate the effect of the echo signals without requiring careful control of the attenuation of transmission facilities. These systems are particularly necessary for digital telecommunications transmission systems, in which it is not feasible or desirable to introduce attenuation in the message content being carried by such transmission systems, but they have also been applied to analog transmission systems. These echo-control systems have incorporated two different main technologies which have been broadly characterized as “echo suppressers” and “echo cancellers.”
Echo suppressers employ one or more speech detectors, and one or more switches in the audio paths of a telecommunications link. For example, in one known echo suppresser, a speech detector monitors the near-end receive path and responsively controls a switch that enabled the near-end transmit path. If speech is not detected (i.e., if the near end is not talking), the echo suppresser disables the transmit path, thus preventing the locally generated echo signal from being transmitted to the far end. Echo suppressers of this type work well provided only one party talks at a time, but work poorly when the parties interrupt one another or talk simultaneously, as is characteristic of normal conversation.
Improved echo suppressers have incorporated speech detectors on both transmit and receive paths, and enable the transmit audio path responsive to a comparison of speech levels on the respective paths. When both parties talk simultaneously, the suppresser may leave the transmit path enabled, resulting in no echo suppression during that period, or may attenuate the transmit path to an intermediate level. Echo suppressers have not provided entirely satisfactory results, in part because some echo remains detectable during periods of simultaneous speech, and because the frequent switching of audio paths results in numerous abrupt changes in speech amplitude which are noticeable to the users.
Echo cancellers construct a model of the round-trip signal path through the network (e.g., the path from the echo canceller to the leakage source at the near end, and back to the echo canceller) that results in the echo signal. Using the model, and based on the original signal transmitted from the far end to the near end, the echo canceller calculates an estimated echo signal which it expects to receive from the near end. The echo canceller then subtracts the estimated echo signal from the received near-end signal. If the model is good, the estimated echo signal closely approximates the actual echo component of the received signal, and the echo is effectively subtracted away or canceled. Thus, substantially only the signal originally transmitted by the near end remains.
Although effective echo cancellers are available, they are expensive. Historically, echo cancellers have been permanently installed to serve particular telecommunications facilities (e.g., trunks). However, echo control is not always required or desirable on a facility at all times or for all calls. For example, facilities may not be continuously in use. Also, some calls, such as those carrying certain types of data may be hindered by the action of echo cancellers. Some existing echo cancellers can detect that a served facility is carrying a data call of a type for which echo cancellation is not desired and may responsively disable cancellation.
Even for calls which do not carry data, conditions at the near end and/or over the communications path may be such that echo control is unnecessary. For example, the near-end leakage signal may be of small amplitude, attenuation along the communications path may be significant, the length of the path may be short, or another echo cancellation device may be present on the call. Any of these conditions could produce an echo signal which either is not noticeable to the user or does not disturb communication. Because a facility served by an echo canceller may be used in conjunction with various other facilities and intermediate and near-end equipment from call to call, echo control may be essential on some calls and superfluous on others.
However, it is believed that no known echo cancelle
Dunn James Patrick
Zhang Peng
Lucent Technologies - Inc.
Tieu Binh
LandOfFree
Method and apparatus for echo cancellation with... does not yet have a rating. At this time, there are no reviews or comments for this patent.
If you have personal experience with Method and apparatus for echo cancellation with..., we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and Method and apparatus for echo cancellation with... will most certainly appreciate the feedback.
Profile ID: LFUS-PAI-O-3155039