Multiplex communications – Diagnostic testing – Determination of communication parameters
Reexamination Certificate
1998-04-16
2002-04-23
Ngo, Ricky (Department: 2664)
Multiplex communications
Diagnostic testing
Determination of communication parameters
Reexamination Certificate
active
06377553
ABSTRACT:
TECHNICAL FIELD
The invention concerns a method for error masking and improvement of the signal quality in digital transmission systems, in which a distribution function for estimating transmitted parameters is determined at the receiving end.
BACKGROUND OF THE INVENTION
To enable masking errors on the receiving end of digital transmission systems for audio, speech or video signals, frame repetition methods are presently used which repeat the last correctly received bit frame (or parts thereof) (e.g. Recommendation GSM 06.11 “Substitution and Muting of Lost Frames for Full Rate Speech Traffic Channels”, ETSI/TC SMG, February 1992). This repetition of frames is initiated by binary frame reliability information, which can be obtained e.g. from the received field strength, from metric differences of a channel decoder, or also from the evaluation of an error detection method. Additional methods (e.g. T. Lagerqvist, T. B. Minde, P. Mustel and H. Nilsson “Soft Error Concealment in a TDMA Radio System”, U.S. Pat. No. #5,502,713, December 1993) are able to carry out a weighted combination of source-codec parameters of the current frame and of preceding ones, where the weighting reflects the error probabilities of the frame or the error probabilities of the parameter.
Disadvantages of the State of the Art I
Disadvantages of these methods are the relatively quick decline in the decoded audio-speech-video quality if the transmission channel becomes increasingly unreliable. This becomes noticeable as a function of the source coding method being used, e.g. by extremely disruptive click or modulation effects, which must often be suppressed by the additional use of muting switch mechanisms. In that case the quality of the reconstructed signals obviously declines as well.
Furthermore the weighted aggregation of current and preceding frames only models very vaguely the statistical behavior of the source-codec parameters, which leads to respectively inaccurate estimation results. In addition the use of an error probability alone for a received source-codec parameter value (or the bit combination representing it) is less than optimal as compared to the case where a respective probability is known at the receiving end for each possible transmitted parameter value.
State of the Art II
The error masking can be improved if, as is known from “Error Concealment by Softbit Speech Decoding”, ITG “Speech Communication” Conference Proceedings, Frankfurt am Main, September 1996, the quantized source-codec parameters are modeled as discrete value mark-off processes of the Nth order, and a probability distribution of all possible transmitted parameter values is known at every moment. This technique estimates every source-codec parameter by using individual parameter estimation methods.
Disadvantages of the State of the Art II
This method however has an exponentially increasing need for memory as a function of the number of bits M of the source-codec parameter to be estimated, in conjunction with the model order N, and an exponentially increasing numeric complexity. For that reason, source-codec parameters with a high number of bits M could only be estimated with low model orders N until now.
SUMMARY OF THE INVENTION
The object of the invention is to create a method and a device for masking errors in digital transmission systems, which achieves a far reaching improvement in the quality of the speech or audio or video signals, and only requires a small amount of memory and numeric complexity.
This object is achieved by a method and device for error masking and improvement of the signal quality in digital transmission systems, in which a distribution function for estimating transmitted parameters is determined at the receiving end, wherein a distribution function is adjusted around the output value of a predictor, and is integrated by sections into a new distribution function, and this new distribution function is multiplied by a distribution function which takes into account the reception quality and the result of an a posteriori distribution which can be used with conventional estimation methods for the final parameter estimation.
This object is also achieved by a device for error masking and improvement of the signal quality in digital transmission systems, in which a distribution function for estimating transmitted parameters is determined at the receiving end, by a distribution function estimating means, wherein the distribution function estimating means has means for causing the distribution function to be adjusted around the output value of a predictor, and to be integrated by sections into a new distribution function, and further having means for multiplying the new distribution function by a distribution function that takes into account the reception quality as well as the result of an a posteriori distribution which can be used with conventional estimation methods for determining the final parameter estimation.
Advantages of the Invention
The method of the invention is suitable e.g. for digital mobile radio receivers, digital cordless telephones, digital radio receivers, but also for the ATM transmission of speech and audio signals. It can also be applied to video picture transmission. In principle it can be used in all areas where reliability information is available for every received source-coded bit or bit groups as well. The predictor provided by the method enables the calculation of a probability distribution for every possible transmitted source-codec parameter, where a clearly reduced memory need and a clearly reduced numeric complexity can be achieved by comparison with the state of the art. The order of the predictor can be chosen with a marginal influence on the memory needed and the numeric complexity of the entire device, so that as much of the residual correlation of the parameter to be estimated as possible can be utilized for the error masking. This in turn promotes a clearly improved masking of the transmission error.
Beyond that, the method of the invention makes possible an efficient estimate of non-stationary source-codec parameters as well.
REFERENCES:
patent: 5212549 (1993-05-01), Ng et al.
patent: 5400076 (1995-03-01), Iwamura
patent: 5442400 (1995-08-01), Sun et al.
patent: 5502713 (1996-03-01), Lagerqvist et al.
patent: 5572622 (1996-11-01), Wigren et al.
patent: 5598506 (1997-01-01), Wingren et al.
patent: 5621467 (1997-04-01), Chien et al.
patent: 5901186 (1999-05-01), Jamal et al.
patent: 4121480 (1993-02-01), None
patent: 4137609 (1993-05-01), None
patent: 4305578 (1993-09-01), None
patent: 0574288 (1997-01-01), None
“Substitution and Muting of Lost Frames for Full Rate Speech Traffic Channels”, ETSI/TC SMG Feb. 1992.
“Error Concealment by Softbit Speech Decoding”, “Speech Communication” Conference Proceedings,Frankfurt am Main, Sep. 1996.
Fingscheidt Tim
Vary Peter
Alcatel
Ngo Ricky
Ware, Freesola, Van Der Sluys & Adolphson LLP
LandOfFree
Method and device for error masking in digital transmission... 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 device for error masking in digital transmission..., we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and Method and device for error masking in digital transmission... will most certainly appreciate the feedback.
Profile ID: LFUS-PAI-O-2885833