Error detection/correction and fault detection/recovery – Pulse or data error handling – Digital data error correction
Reexamination Certificate
1998-04-09
2001-01-23
De Cady, Albert (Department: 2784)
Error detection/correction and fault detection/recovery
Pulse or data error handling
Digital data error correction
Reexamination Certificate
active
06178535
ABSTRACT:
FIELD OF THE INVENTION
The present invention relates to a method according to the preamble of the appended claim
1
for decreasing the frame error rate in a data transmission system in which information is transmitted in the form of data frames. The invention is further related to a data transmission system according to the method and a mobile station according to the preamble of the appended claim
11
.
BACKGROUND OF THE INVENTION
In data transmission in the form of data frames, the information to be transmitted is divided into data frames of usually regular length. In addition to the primary information, the data frames can also contain title information and other information required in the transmission of data frames. The data frames are transmitted from the transmitter to the receiver via a transmission channel which can comprise e.g. the radio channel or another wireless transmission channel. The transmission channel is exposed to disturbances, such as ignition interference caused by electrical devices, and on the other hand to interference caused by other devices of similar type used in wireless data transmission, such as radio transmitters. Still another considerable cause of disturbance, particularly in movable transmitter/receiver devices, is the fact that the signal to be received can enter the receiver via several routes of different lengths, thus distorting the signal to be received. Therefore, for eliminating transmission errors, error correction data or at least error detection data is usually added to the data frames. One method for adding error correction data is the use of so-called convolutional codes, i.e. the information to be transmitted is coded by using a suitable convolutional code, and the convolutional-coded information is transmitted to the data transmission channel. At the receiving stage, the operation is reversed for separating the transmitted information from the received data transmission stream. The error detecting data used is most commonly parity checking data which is calculated from the information to be transmitted, or at least from part of it. One such known parity checking method is the so-called cyclic redundancy check (CRC). Thus, at the receiving end, the received information is subjected to the corresponding operation, and the parity checking data formed at the receiving end is compared with the received parity checking data. If the data are equivalent, the receiving device concludes that the data was received correctly. If the calculated and received parity data are not equivalent, the receiving device concludes that the received data frame was at least partly incorrect. After this it is possible to request re-transmission or to try to decode the incorrect frame e.g. by interpolation.
Using error correction methods, it is possible to correct at least some of the potential transmission errors, wherein re-transmission is not necessary in all error situations. However, when an error checking method is used, only correctness or incorrectness is detected, and re-transmission is requested in an error situation, thus delaying the data transmission. These codes can also be called an outer code and an inner code. Outer coding is conducted before inner coding. Thus, the coding data formed by outer coding is further subjected by inner coding, thus improving the reliability of the transmission. The outer code is usually an error-detecting code and the inner code an error correcting code, but these can also be reversed. The coding can also concatenate more than two codes.
In current digital mobile communication systems, also speech is transferred in the form of data frames. For example in the GSM mobile communication system (Global System for Mobile Communications), in the full rate speech communication channel a majority of the digital information formed of the audio signal is protected by error correction coding. In the speech coder, 260 speech parameter bits are formed for each speech sequence of 20 milliseconds. Of these 260 bits, 182 bits with subjectively greatest significance are protected with an error correction code. These 182 bits are subjected to convolutional coding at a coding rate of ½, i.e. two bits to be transmitted to the transmission channel are formed for each information bit. The remaining 78 bits are transmitted with no protection, i.e. possible errors in them are not detected at the receiving stage.
The bit (symbol) error ratio in the received data frames can at times exceed the error correction capacity of the error correction method used in the transmission of the data frames. As a result, all errors cannot be corrected, wherein the most common procedure is to request re-transmission of such data frames or, e.g. in speech coding, to try to form a data frame synthesised on the basis of previously received data frames. Synthesisation of data frames can be used to some extent in the transmission of audio and video signals, but for example in the transmission of data signals, it is not possible to use a synthesised data frame.
When the error correction capacity of the receiver is exceeded, it is important to detect those errors which are still left after error correction. Such incorrect information should not be used in the receiver when reconstructing the transmitted information. For example in the full rate traffic channel in the GSM system, detection of uncorrected errors is conducted by CRC coding by forming three parity checking bits. When forming these parity checking bits, 50 bits from each data frame are used which are the most significant for the information to be transmitted. Thus these 50 data frame bits are subjected to the corresponding operation in the receiver, and the parity checking bits are compared with the parity checking bits transmitted with the data frame, wherein possible changes indicate that there was an error in the data transmission.
In the GSM system, speech decoding will reject all data frames in which it was not possible to correct all errors. These data frames are replaced by a data frame formed on the basis of previously received acceptable data frames. If the number of incorrect data frames is relatively small, the replaced data frames do not impair significantly the quality of the decoded speech signal. However, if the number of incorrect data frames increases, its effect can be gradually clearly audible in the speech signal. This can even lead to the fact that the decoded speech signal is no longer intelligible.
FIG.
1
a
is a block diagram showing a speech coding system according to prior art. This is an example of the full rate speech coding system of the GSM system. FIG.
1
a
is a block diagram showing speech coding, addition of parity checking bits, and convolutional coding.
FIG. 4
is a flow chart showing this channel coding used in the transmission of a speech signal of the GSM mobile communication system according to prior art. In the following, the operation of channel coding is shown with reference to the apparatus of FIG.
1
a
and the flow chart of FIG.
4
.
The speech signal is divided into frames, or time intervals of certain length, which in this system is 20 milliseconds. Each frame is coded separately. Thus each speech signal frame of 20 milliseconds yields a group of speech parameters in digital form. The digital speech samples
100
formed of the speech signal are coded in the speech encoder
101
in order to form a speech parameter frame. The speech encoder compresses the speech into a bit stream of 13.0 kbit/s. Of each speech frame of 20 milliseconds, the encoder forms
260
speech parameter bits which make up the speech parameter frame
102
(stage
401
).
This speech parameter frame
102
is further transmitted to a channel encoder
104
for grouping the bits, for example into bits to be protected with error correction coding and into bits to be left unprotected. Further, the channel encoder is used for forming error detection information, wherein some of the speech parameters are used for its calculation. In the speech parameter frame
102
, the bits for
Kajala Matti
Vainio Janne
Cady Albert De
Chase Shelly A
Nokia Mobile Phones Limited
Perman & Green LLP
LandOfFree
Method for decreasing the frame error rate in data... 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 for decreasing the frame error rate in data..., we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and Method for decreasing the frame error rate in data... will most certainly appreciate the feedback.
Profile ID: LFUS-PAI-O-2507573