Error detection/correction and fault detection/recovery – Pulse or data error handling – Digital data error correction
Reexamination Certificate
1998-02-06
2002-10-22
Ton, David (Department: 2133)
Error detection/correction and fault detection/recovery
Pulse or data error handling
Digital data error correction
C714S758000
Reexamination Certificate
active
06470470
ABSTRACT:
FIELD OF THE INVENTION
The present invention relates to an information coding method utilizing focused error correction and/or error detection, in which method the quality of the data transfer connection is used for selecting the coding mode for the data transfer connection. The invention also relates to a system and terminal devices applying the method. The invention is particularly suitable for use in connection with data transfer connections realized by radio.
BACKROUND OF THE INVENTION
While transferring information, such as speech or data, using transfer connections subject to transmission errors, the information to be transferred is in general protected using an error correction algorithm. Especially in digital connections an attempt is made to detect transmission errors, and to correct the erroneous information bits. How successfully this is done depends among other things on the number of transmission errors and on the error correction algorithm used. In speech coding systems prior known to a person skilled in the art, a major part of the bits comprising speech information are protected using an error correction code. This is the procedure e.g. in the so called full rate (FR, Full Rate) speech codec of the GSM system.
In the full-rate speech codec (which later is also called FR-speech codec) of the GSM system a RPE-LTP (Regular Pulse Excitation-Long term Prediction) based speech encoding system is used. It produces 260 speech parameter bits for each 20 ms speech frame. Out of these 260 bits, the 182 subjectively most important bits are protected using an error correction code. As the error correction code, ½-rate convolution encoding is used. The remaining 78 bits are transferred in the data transmission connection completely without error correction.
The number of transmission errors on a data transfer connection may temporarily exceed the error correction capacity of the ½-rate convolution coding used in the GSM-system. As a result, the important received speech parameter bits may contain transmission errors. It is important to detect that these transmission errors occurred, even if it were not possible to correct them. If the speech parameters which are the most important for speech quality contain transmission errors, they shall not be used for speech synthesizing in the receiver, but they must be rejected. In the full rate FR-speech codec of the GSM-system 3-bit CRC (Cyclic Redundancy Check)-error detection is used. CRC-error detection is focused on the 50 most important bits of speech coding. In a receiver the error detection code is used for verifying the correctness of the 50 most important bits of each 20 ms speech frame. If they contain errors, the frame is classified as bad and it is not used in speech synthesizing. Instead, an attempt is made to substitute the bad frame with an estimate, which is formed e.g. based upon chronologically preceding error-free frames.
The full-rate speech coding method of the GSM-system briefly presented above, operates reasonably well, provided that the relative share of transmission errors does not grow too high. Under these conditions the error correction algorithm is capable of correcting transmission errors sufficiently for obtaining a satisfactory transfer connection and through it a satisfactory speech quality. When the proportion of transmission errors grows to medium or high level, the error correcting capability of convolution coding having ½-rate coding ratio is exceeded. In this case a more efficient error correction algorithm would be needed, such as e.g. a convolution coding having ⅓-rate coding ratio. In this case, however, the total speech encoding efficiency will be reduced essentially, because more error correction information bits must be included in the data transfer connection. This naturally increases the data transfer rate required of the data transfer connection. Accordingly, this approach cannot be used for codecs with fixed line speed. Instead, the above presented method based upon making the error correction algorithm more efficient is suitable for systems with variable line speed.
For instance, the total bit rate of the data transfer system used for transferring speech can be kept constant, provided that at the same time when the number of bits used for the error correction of speech parameter bits is increased, the number of bits used for speech encoding itself is reduced. This in turn requires using several different speech codecs with different line speeds in both the transmitter and the receiver, which makes the structure of the system more complicated. Further, the lower the number of bits used for speech encoding, the more calculation capacity is normally required of the various components of the system. The above presented disadvantages increase the cost of the system. In addition to the above, the deterioration of speech quality cannot be avoided when more bits are used for error correction, because the fewer bits there are available for speech encoding, the more one has to compromise the voice quality. The voice quality deterioration due to the reduction of the number of bits used for speech encoding is particularly important in a case where there is background noise to speech, e.g. the noise from a car engine.
One problem occurring in speech coding methods according to prior art is the complete muting of the speech synthesizing in a receiver when data transfer connections containing a large number of transfer errors are used. This is due to the fact that when an error detection algorithm detects transfer errors in speech frames, it too easily mutes the speech synthesizer. This leads to the loss of speech information.
As is evident from the above description, there is a need to develop a better method of protecting information parameters on data transfer connections containing numerous transfer errors. In addition, there is a need to develop a system, the receiver of which better tolerates information parameter frames containing errors. In the following the information coding method according to the invention and the system utilizing it and the terminal devices are explained primarily using the speech coding in a mobile communication system as an example. Nothing however limits using the information coding system according to the invention for coding of data other than speech data. For the sake of clarity the invention is in the following also called a speech coding method, because it best describes one of the most important fields of application of the invention. It is possible to utilize the invention instead of a radio connection, also e.g. in connection with information transfer systems realized using wireline connections.
SUMMARY OF THE INVENTION
Now an information coding method utilizing focused error correction and error detection system has been invented, by use of which the above described problems can be reduced. One of the purposes of the present invention is to present a speech coding method which will be automatically adjusted as a function of the quality of a data transfer connection optimizing the speech quality on data transfer connections of any quality. The quality of the data transfer connection used is analyzed by measuring the parameters describing the quality of a data transfer connection, such as e.g. C/I (Carrier to Interference) ratio, S/N (Signal to Noise) ratio or bit error rate (Bit Error Rate, BER) as known to a person skilled in the art. In the information coding method according to the invention there is no need to reduce the number of bits used for speech coding in relation to the total bit rate used in the information transfer connection, in which case the voice quality of the speech preferably remains good. In the information coding method according to the invention error correction and/or error detection is focused on the bits most essential for voice quality as a function of the C/I-ratio or of some other parameter describing the quality of the data transfer connection. The muting of speech synthesizing occurring in prior art sy
Jarvinen Kari
Kajala Matti
Vainio Janne
Nokia Mobile Phones Limited
Ton David
LandOfFree
Information coding method and devices utilizing error... does not yet have a rating. At this time, there are no reviews or comments for this patent.
If you have personal experience with Information coding method and devices utilizing error..., we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and Information coding method and devices utilizing error... will most certainly appreciate the feedback.
Profile ID: LFUS-PAI-O-2993024