Error detection/correction and fault detection/recovery – Pulse or data error handling – Digital data error correction
Reexamination Certificate
1999-08-27
2001-05-15
De Cady, Albert (Department: 2784)
Error detection/correction and fault detection/recovery
Pulse or data error handling
Digital data error correction
Reexamination Certificate
active
06233708
ABSTRACT:
BACKGROUND OF THE INVENTION
Field of the Invention
The invention relates to a method for frame error detection in a data transmission, and to a device for performing the method.
In order to make maximum use of the limited capacity of an available transmission channel, before data transmission in the transmitting direction, especially in video, audio and speech transmission, a source coding is done in order to eliminate redundance from the data and thus compress the data to be transmitted. Often the data to be transmitted are first divided up into so-called frames. These frames can be time slots of a certain duration or can also be spectral ranges or image excerpts or segments. Then in the source coding, parameters are ascertained that describe or determine the data to be transmitted as accurately as possible. This is known as parametric source coding. One example of it in the case of speech coding is GSM full-rate coding. In speech source coding (speech coding), the ascertained parameters are also called speech coefficients.
In the transmission of these compressed data, if the channel conditions are poor, errors can occur despite protective provisions such as channel coding; then the received speech, for instance, is greatly interfered with and may even becomes unintelligible. In the GSM standard, all the received speech frames are therefore first classified as good or bad with the aid of a frame error detection criterion. The bad frames are then replaced with a frame, which typically results from extrapolating one or more previously received “good” frames. The interference in the speech can thus be reduced, and the speech intelligibility is preserved.
In GSM full-rate transmission, for instance, the speech is subdivided into frames of 20 ms each, and each frame is compressed into 260 bits by the full-rate or enhanced full-rate coder. These bits are ranked by their importance as class 1a, 1b, and class 2. The 50 bits in class 1a are secured, among other things by way of a parity check, with 3-bits. If this CRC (cyclic redundancy check) shows an error, then the frame is declared bad. The 3-bit parity check in the full-rate speech coder has a certainty of approximately 7/8 for error detection; that is, good frames are found to be bad and vice versa. Bad frames that have been detected as good are often heard as a pronounced “crackling” in the speaker of the mobile phone.
SUMMARY OF THE INVENTION
It is accordingly an object of the invention to provide a method for frame error detection and a device for performing this method which overcomes the above-mentioned disadvantages of the heretofore-known methods and devices of this general type and which can detect the frame errors more reliably. In particular it is an object of the invention to reliably detect those frames that have a good quality as good frames and to detect the frames that have a bad quality as bad frames.
With the foregoing and other objects in view there is provided, in accordance with the invention, a method for frame error detection, which comprises:
determining, in a receiver, a plurality of comparison values including a given comparison value depending on at least one of a frame energy of a given speech frame and a frame energy change between the given speech frame and a preceeding speech frame; and
identifying the given speech frame as a bad speech frame if a logical combination of a plurality of criteria is met, one of the criteria being based on a comparison of the given comparison value with a threshold value.
In accordance with another mode of the invention, at least one of the comparison values is determined from a linear combination of at least one value selected from the group consisting of values based on a cyclic redundancy check, values depending on the frame energy of the given speech frame, values depending on the frame energy change between the given speech frame and the preceeding speech frame, values depending on parameters of the given speech frame ascertained from a parametric source coding process, and values depending on changes of the parameters of the given speech frame with respect to parameters of the preceding speech frame.
In accordance with yet another mode of the invention, a source coding is performed within a data transmission, the source coding including a linear predictive coding, and at least one of the comparison values is determined on the basis of a change in at least one linear predictive coding filter coefficient of the given speech frame compared with a corresponding linear predictive coding filter coefficient of the preceding speech frame.
In accordance with a further mode of the invention, a source coding is performed within a data transmission, the source coding includes a long-term prediction, and at least one of the comparison values is determined on the basis of a value of a long-term prediction delay.
In accordance with yet a further mode of the invention, a source coding is performed within a data transmission, the source coding includes a regular pulse excitation analysis, and at least one of the comparison values is determined on the basis of at least one block amplitude or on the basis of a rest energy value of a subframe.
In accordance with a further mode of the invention, an error propagation is reduced by correcting values of parameters to be stored in a memory for subsequent calculations.
With the objects of the invention in view there is also provided, a device for frame error detection, which is preferably included in a receiver, the device comprising:
a determining unit for determining a plurality of comparison values including a given comparison value depending on at least one of a frame energy of a given speech frame and a frame energy change between the given speech frame and a preceding speech frame;
a comparison unit connected to the determining unit, the comparison unit performing comparison operations including a comparison of the given comparison value with a threshold value;
a logic unit connected to the a comparison unit, the logic unit performing a logical combination of the comparison operations;
an identifier unit connected to the logic unit, the identifier unit identifying the given speech frame as a bad speech frame based on the logical combination; and a calculation unit connected to the determining unit, the calculation unit calculating the frame energy of the given speech frame.
In accordance with another feature of the invention, the calculation unit estimates the frame energy of the given speech frame.
In accordance with yet another feature of the invention, an evaluation unit is connected to the determining unit, the evaluation unit evaluating a cyclic redundancy check.
In accordance with a further feature of the invention, an ascertaining unit is connected to the determining unit, the ascertaining unit ascertaining parameters specific for data transmitted within the given speech frame and the preceding speech frame.
In accordance with another feature of the invention, a correction unit is connected to the ascertaining unit, the correction unit correcting the parameters such that an error propagation is reduced, and a storage unit is connected to the ascertaining unit, the storage unit storing corrected parameters for subsequent calculations.
The object of the invention is achieved by declaring frames as wrong if a certain logical combination of a plurality of different criteria is met. At least one criterion is to compare a so-called comparison value, which is ascertained in the receiving device, with a predetermined threshold value. If the threshold value is exceeded or undershot, then the criterion is considered to be met.
In a preferred embodiment of the invention, this comparison value is calculated by a mathematical function, whose arguments may be measured values or values that are available within the digital processing.
Another advantageous embodiment of the invention provides that at least one of the comparison values to be ascertained depends, among others, on the frame energy or its change relative to the frame energy of a
Erben Christian
Hindelang Thomas
Xu Wen
Cady Albert De
Chase Shelly A
Greenberg Laurence A.
Lerner Herbert L.
Siemens Aktiengesellschaft
LandOfFree
Method and device for frame error detection 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 frame error detection, we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and Method and device for frame error detection will most certainly appreciate the feedback.
Profile ID: LFUS-PAI-O-2549533