Telecommunications – Transmitter and receiver at same station – Radiotelephone equipment detail
Reexamination Certificate
2000-12-07
2004-04-20
Trost, William (Department: 2683)
Telecommunications
Transmitter and receiver at same station
Radiotelephone equipment detail
C455S569100, C379S406010, C379S406100, C370S286000, C370S292000, C381S071800
Reexamination Certificate
active
06725066
ABSTRACT:
BACKGROUND OF THE INVENTION
The invention relates to a method of cancelling echoes in a telecommunications system wherein an exchange of information takes place between a local subscriber and a subscriber at the remote end of a transmission link and at least one subscriber is assigned an echo canceller, the parameters of which are set i.a. as a function of an echo delay time, and to an echo canceller for the execution of the method.
Disturbing echoes, so-called line echoes, occur due to undesired electric coupling between a receiving path and a transmitting path. In modern communications technology such couplings or echoes, caused by data reductions, junctions, frequent 2-wire/4-wire transitions etc., constitute an increasing problem. Therefore echo-cancelling measures, such as NLMS algorithms or compander algorithms, are used to improve the quality of speech- or data connections. In the case of these algorithms it is often useful to have knowledge of the delay time of the echo signals. To determine the echo delay time it must be ascertained when the transmitted signal arrives at the echo reception location. As telecommunications systems are generally operated bidirectionally, the echo signal can be superimposed with a signal fed-in at the remote end of the transmission link as useful signal. This occurs whenever speech takes place simultaneously at both ends of the transmission link. This is the so-called double talk situation. It is a general problem in all possible transmission situations to clearly identify an echo and distinguish it from useful signals in the double talk situation.
DE-A-42 29 0110 has disclosed a method of detecting an echo in which the transmitted signal is recorded and compared with the signal incoming at the reception location. For this purpose a correlation analysis is performed, in which the degree of correlation indicates the probability as to whether an echo or a mixed signal, produced by the double talk, is present. This decision is made by a threshold value. The greater the selected threshold value, the less often are existing echoes detected. False decisions therefore frequently occur. If on the other hand the threshold value is selected too low, the danger exists that useful signals will be filtered out in double talk. The computation outlay for the correlation analysis is very high while the echo detection reliability is unsatisfactory.
Additionally, DE-A-19 85 0272 has disclosed a method of detecting an echo wherein local extreme values are determined from the transmitted signal and the received signal. As the time difference between two local extreme values is dependent upon the curve formation of the transmitted signal, in the case of an echo the same time difference between consecutive extreme values occurs for the transmitted signal and the received signal. These extreme values are marked with a pulse, thereby giving rise to pulse trains which represent a characteristic of a speech sequence. A measure of the similarity is derived from the pulse train of the local extreme values of the transmitted signal and of the received signal. This method requires a high signal processing outlay, using similarity criteria to make a decision about the echo characteristic of a signal. The decision reliability is consequently relatively poor.
SUMMARY OF THE INVENTION
The object of the invention is to improve the reliability of the echo detection whilst endeavouring to achieve as low as possible a storage and computation outlay.
The signal received by the subscriber are determined at equidistant time intervals, so that sign sequences arise which are stored and compared with one another, and that in the event of a direct correspondence or a correspondence as a result of inversion of the sign sequences of the received signal and of the transmitted signal, the received signal is detected as an echo and the echo transmission time is calculated. In this way it is possible to make a very reliable decision as to whether an echo signal or a mixed signal is present without taking into account similarity criteria. The decision reliability is the better, the smaller the selected equidistant time intervals for the sign determination. For example a signal curve of the transmitted signal characterised by the following blocks can occur: 23 times positive sign, 40 times negative sign, 51 times positive sign, 24 times negative sign. Only when precisely this block classification is determined either exactly or in inverted form also for the received signal, is an echo signal present.
To determine the echo delay time from a determined correspondence of this type, it is provided that the time offset of the sign sequences and of the corresponding signal blocks is determined.
In accordance with an advantageous further development, the sign sequences of the transmitted signals are determined during a fixed time interval. This time interval is selected such that a sufficient number of signs for the unequivocal characterisation of the signal curve of the transmitted signal can occur within it. For example on average the time interval can comprise five sign blocks, the blocks being redetectably characterised by a sufficient number of sign determinations. In the event of a restriction to approximately five sign blocks, during this time interval at least one hundred sign determinations are to take place, i.e. the time interval is to amount to at least one hundred times the equidistant time interval between two sign determinations. The sign sequences of the received signals are determined during the maximum echo delay time. In this way, in spite of a reduced outlay, a reliable identification of echo signals occurs. A calculation of the echo delay time as a precondition of the echo cancellation naturally only takes place if the received signal exceeds a fixed threshold above the reception noise. After a delay time measurement, no further delay time measurement is to occur until the transmitted signal breaks off. The determination of the next sign sequence during the fixed time interval is not initiated until the commencement of the next speech-specific transmitted signal.
An echo canceller for the execution of the method. Means are provided for determining the sign sequences, as well as comparison means and a counter for determining the time offset between sign sequences which correspond directly or as a result of inversion, where the time offset forms a measure of the echo delay time.
REFERENCES:
patent: 4562312 (1985-12-01), Duttweiler
patent: 4577309 (1986-03-01), Barazeche et al.
patent: 4645884 (1987-02-01), Barazeche et al.
patent: 4965823 (1990-10-01), Nakagawa et al.
patent: 5737410 (1998-04-01), Vahatalo et al.
patent: 6044068 (2000-03-01), El Malki
patent: 6061444 (2000-05-01), Kawahara et al.
patent: 0 300 352 (1989-01-01), None
patent: 0 592 787 (1998-12-01), None
patent: 0 998 110 (2000-05-01), None
Alcatel
Le Danh C
Sughrue & Mion, PLLC
Trost William
LandOfFree
Method of cancelling echoes in a telecommunications system... 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 of cancelling echoes in a telecommunications system..., we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and Method of cancelling echoes in a telecommunications system... will most certainly appreciate the feedback.
Profile ID: LFUS-PAI-O-3206649