Error detection/correction and fault detection/recovery – Pulse or data error handling – Digital data error correction
Reexamination Certificate
1999-12-13
2001-09-18
DeCady, Albert (Department: 2133)
Error detection/correction and fault detection/recovery
Pulse or data error handling
Digital data error correction
Reexamination Certificate
active
06292922
ABSTRACT:
BACKGROUND OF THE INVENTION
FIELD OF THE INVENTION
The present invention relates to a method and a decoder for a source-controlled channel-decoding of data in a frame format.
Source signals such as voice, sound, image and television almost always include statistical redundancy in the sense of a correlation between different bits. This redundancy can be greatly reduced through the use of source-encoding so that an efficient transmission or storage of the source signal is made possible. This reduction in redundancy eliminates redundant signal contents prior to a transmission. The redundant signal contents are based on the previous knowledge of, for example, statistical parameters of the signal profile. After the transmission, these parts or portions are added again to the signal so that it is objectively impossible to demonstrate any loss of quality. Due to the incomplete knowledge of the source signals or restrictions in the complexity of the encoding method, the source-encoding can usually only be implemented in a less than optimum way, i.e. there is still a certain degree of redundancy in the compressed data even after the encoding.
On the other hand, when transmitting the signals it is necessary to selectively add redundancy again through the use of a channel-encoding in order to largely eliminate the effects of a channel interference on the transmission. Additional redundant bits thus make it possible for the receiver to detect errors, and possibly also correct them, without involving the transmitter.
For a long time, one of the basic premises of the information theory was that source-encoding and channel-encoding could be carried out independently of one another in order to achieve an optimum result. According to this principle, the configuration of the source-decoder depends only on the source properties, whereas the channel-encoding scheme will depend only on the channel properties. This principle may be correct if the source-encoder supplies statistically independent results which are thus noncorrelated and equally probable and if the decoding delay can be of any desired value. In practical applications, these conditions are however usually not fulfilled. The output signal of the source-encoder often has a residual redundancy, and at the same time the permitted delay is restricted, in particular in the case of voice transmission.
This residual redundancy of the source-encoded data may be used in the so-called source-controlled channel-decoding, in order to correct further bit errors. In this case, the decoding operation of the channel-decoder is controlled on the one hand by the transmitted code bits and on the other hand by an item of a-priori/a-posteriori information relating to the most probable value of a number of important source bits. The source information therefore affects the way in which the channel-decoding proceeds. In the case of Viterbi algorithm decoding, this process is referred to as an a-priori Viterbi algorithm. When such a method is used, a modification is necessary only at the receiver end.
The German Patent DE 42 24 214 and the publication “Source-Controlled Channel Decoding”, J. Hagenauer, IEEE Trans. Commun., vol. 43, pages 2449-2457, September 1995 disclose the use of the inter-frame correlation, i.e. the statistical dependence between chronologically and/or spatially adjacent signal samples in source-controlled channel-decoding.
The above-mentioned method will now be explained in more detail with reference to FIG.
4
. As is clear in
FIG. 4
, according to this method, a signal is firstly source-encoded
10
, then channel-encoded, transmitted over a transmission channel
12
, channel-decoded
13
and finally source-decoded
14
. Furthermore, the channel-decoding takes place under source control using an item of a-priori information and an item of a-posteriori information relating to the sources.
Here, a reliability, which determines in advance (a-priori) a probability for a decision in a subsequent step, is calculated subsequently (a-posteriori) from the statistic of the source and the information which has already been decided. Here, the a-posteriori information can be acquired directly downstream of the channel-decoder, but may also be acquired downstream of the source-decoder.
Parametric encoding methods are treated below. Here, the bit stream which is generated by the source (e.g. voice) is divided into blocks (e.g. chronological frames) and these are processed separately. The source-encoding supplies so-called parameters (e.g. voice coefficients) which reflect the properties of the source in the current block in a certain way (e.g. voice spectrum, filter parameters) and which are quantized with a certain number of bits.
Here, in particular a correlation of the source bits is taken into account. The fundamental idea of this method is that the most significant bits of a parameter do not change very often between two successive frames and there is thus redundancy in the transmission. This correlation between successive frames can be exploited at the receiver end by using an APRI-SOVA (a-priori soft decision Viterbi algorithm) decoder. In a soft-decision decoder, not only the actual decision but also the reliability of the decision is taken into account. In summary, according to this method, a source-controlled channel-decoding takes place using the inter-frame correlation.
However, it has become apparent that due to the unequal distribution of the parameter values, which is in turn due to the nonstationarity of the source signals, in particular when transmitting voice, there may be residual redundancy not only between bits in successive frames (inter-frame correlation) but also between the bits of a parameter within one frame. This redundancy between different bits of a frame is referred to as intra-frame redundancy. In other words an intra-frame correlation exists between bits belonging to the same frame.
SUMMARY OF THE INVENTION
It is accordingly an object of the invention to provide a method for a source-controlled channel-decoding of data in a frame format and a channel-decoder configuration for data in a frame format which overcome the above-mentioned disadvantages of the heretofore-known methods and devices of this general type and which perform an effective decoding with little expenditure.
With the foregoing and other objects in view there is provided, in accordance with the invention, a method for a source-controlled channel-decoding of data in a frame format, preferably in a GSM frame format, including the steps of channel-decoding a given frame of data, such as data of a voice channel according to the GSM standard; determining an a-posteriori information based on at least one of a reliability decision of the channel-decoding and a source-decoding subsequent to the channel-decoding; calculating an a-priori information based on the a-posteriori information and based on a residual redundancy of the data; and channel-decoding the given frame once more using the a-priori information.
In accordance with another mode of the invention, during an encoding of the data of the given frame, respective items of the data of the given frame are provided with different error protections depending on how important or a significant the respective items of the data are for a successful transmission.
In accordance with yet another mode of the invention, the a-priori information is determined by logically linking the a-posteriori information with source properties. The source properties may be determined by taking into account the last frames. The number of frames being taken into account may be based on at least one source property, such as a correlation and a sationarity.
In accordance with a further mode of the invention, in particular when the invention is used in a GSM system, the last 128 frames or the last 256 frames are taken into account when determining the source properties.
With the objects of the invention in view there is also provided, a channel-decoder configuration, including a channel-decoder decoding a given frame having data, the c
Hindelang Thomas
Ruscitto Alfredo
Xu Wen
Chase Shelly A
De'cady Albert
Greenberg Laurence A.
Lerner Herbert L.
Siemens Aktiengesellschaft
LandOfFree
Source controlled channel decoding using an intra-frame does not yet have a rating. At this time, there are no reviews or comments for this patent.
If you have personal experience with Source controlled channel decoding using an intra-frame, we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and Source controlled channel decoding using an intra-frame will most certainly appreciate the feedback.
Profile ID: LFUS-PAI-O-2535177