EXPLOITING KNOWN PADDING DATA TO IMPROVE BLOCK DECODE SUCCESS RATE
A method and system of decoding a ccnvolutionally encoded data block having known padding bits. A Viterbi decoder is constrained to a state corresponding to k−1 padding bits immediately adjacent to data bits of the data block, where k is a constraint length of a convolution encoder used to encode the data block. Symbols of the encoded data block that have influence only from the padding bits are discarded.
Latest RESEARCH IN MOTION LIMITED Patents:
- Aligning timing for direct communications
- MANAGING SHORT RANGE WIRELESS DATA TRANSMISSIONS
- METHODS AND SYSTEMS FOR CONTROLLING NFC-CAPABLE MOBILE COMMUNICATIONS DEVICES
- IMAGING COVER FOR A MOBILE COMMUNICATION DEVICE
- MOBILE WIRELESS COMMUNICATIONS DEVICE PROVIDING NEAR FIELD COMMUNICATION (NFC) UNLOCK AND TAG DATA CHANGE FEATURES AND RELATED METHODS
This is the first application filed for the present technique.
MICROFICHE APPENDIXNot Applicable.
TECHNICAL FIELDThis application relates to communications networks in which known padding data are used to rate adapt data to frames/blocks of different sizes, and in particular to methods and systems for exploiting the known padding data to improve decode success rate of convolutionally encoded blocks.
BACKGROUNDWithin the modern network space, it is frequently required to map data from one transmission protocol (or format) to another. For example, within the Enhanced Data for Global Evolution (EDGE) protocol, data can be mapped to/from any of nine different channel coding schemes, depending on the radio link quality. Each coding scheme provides a respective different data throughput (or rate). Consequently, in order to perform a mapping between any two coding schemes, the data must be rate (or size) adapted. In order to do this, padding data are used to match the size of a data block (of the source coding scheme) to the block size of the destination coding scheme.
For example,
Typically, the padding data is provided as either a data fill of binary 1's or 0's. In all cases, conventional forward error correction (FEC) schemes are unaware of padding bits within a data block. The entire data block will be encoded (e.g. using convolutional encoding) and decoded (e.g. using a Viterbi decoder) without reference to the presence or absence of padding bits within the data block. Since the padding bits are known in advance, this leads to redundant encoding and decoding operations, and increases the probability of errored data decoding.
It would be desirable to enhance forward error correction of encoded data containing known padding bits by exploiting the fact that these padding bits are known and not only need not be decoded, but can also be used to improve the probability of decode success of the data.
Further features and advantages of the present technique will become apparent from the following detailed description, taken in combination with the appended drawings, in which:
It will be noted that throughout the appended drawings, like features are identified by like reference numerals.
DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTThe present disclosure provides methods and apparatus for enhancing forward error correction of convolutionally encoded data containing known padding bits at the beginning and/or at the end of a block. Embodiments are described below, by way of example only, with reference to
Referring to
As is well known in the art, convolutionally encoded symbols are decoded using a Viterbi decoder 18 which may, for example, be implemented in a terminal device 4. If desired, the Viterbi decoder 18 may be implemented as part of the RF transceiver 8 as shown in
Normally, the decoder is assumed to start at state 0 and is forcibly terminated, through the appending of k−1 tail bits, at state 0. This provides a convenient mechanism for converging both ends of the trellis so that a unique survivor path representing the most likely decode of the received symbols can be chosen.
In the case of rate adaptation between any two of the various coding schemes of the EDGE protocol, for example, the number and location(s) of padding bits within the data block are known, as is the value of each padding bit. With this information, it is possible to skip over decoding of the known padding bits, and constrain the start state and/or the end state of the decoder according to the known values of the padding bits to force the padding bits to be decoded to their known values and thereby improve accuracy of the decoding operation.
It will be appreciated that the methods described above with reference to
By constraining the decoder in the above-noted manner, the decoder only decodes bits corresponding to actual data. This improves decoder performance by minimizing the number of bits that need to be decoded in order to recover the data from a data block. In addition, decoding accuracy of the data is improved, because selection of the highest probability path for traceback and decoding is not perturbed by (possibly erroneously) decoded padding bits.
The conventional approach to using known data in Viterbi decoding involves trellis pruning to exclude non-candidate paths. However, implementation of a Viterbi decoder that supports arbitrary trellis pruning is complicated. The present technique is much more useful in practice because it is easy to implement a Viterbi decoder with selectable start and/or end states.
The embodiment(s) described above is(are) intended to be representative only. The scope of the present application is therefore intended to be limited solely by the scope of the appended claims.
Claims
1-12. (canceled)
13. A method of decoding a convolutionally encoded data block having a predetermined number of known padding bits, the method comprising:
- constraining a Viterbi decoder to a state corresponding to k−1 padding bits immediately adjacent data bits of the data block, where k is a constraint length of a convolutional encoder used to encode the data block, and wherein the k−1 padding bits immediately adjacent data bits of the data block comprise less than the predetermined number of known padding bits; and
- discarding symbols of the encoded data block that have influence only from padding bits other than the k−1 padding bits immediately adjacent data bits of the data block.
14. The method as claimed in claim 13, wherein the known padding bits comprise n known padding bits prepended to the data bits, and wherein constraining the Viterbi decoder comprises constraining a Start state of the decoder to a state corresponding to the last k−1 of the n known padding bits prepending the data bits.
15. The method as claimed in claim 14, wherein discarding symbols of the encoded data block comprises:
- disabling operation of the Viterbi decoder while latching the first n bits of the data block into the shift register of the Viterbi decoder.
16. The method as claimed in claim 13, wherein the known padding bits comprise n known padding bits appended to the data bits, and wherein constraining the Viterbi decoder comprises constraining an End state of the decoder to a state corresponding to the first k−1 of the known padding bits appending the data bits.
17. The method as claimed in claim 16, wherein discarding symbols of the encoded data block comprises:
- discarding any remaining bits of the encoded data block after processing the first e=(N−n)+(k−1) bits of the encoded data block through the Viterbi decoder, where N is the total length of the data block.
18. The method as claimed in claim 13, wherein the convolutionally encoded data block is a data block of an Enhanced Data for Global Evolution, EDGE, data frame.
19. A system for decoding a convolutionally encoded data block having a predetermined number of known padding bits, the system comprising:
- a Viterbi decoder; and
- a controller adapted to: constrain the Viterbi decoder to a state corresponding to k−1 padding bits immediately adjacent data bits of the data block, where k is a constraint length of a convolution encoder used to encode the data block, and wherein the k−1 padding bits immediately adjacent data bits of the data block comprise less than the predetermined number of known padding bits; and discard symbols of the encoded data block that have influence only from padding bits other than the k−1 padding bits immediately adjacent data bits of the data block.
20. The system as claimed in claim 19, wherein the known padding bits comprise n known padding bits prepended to the data bits, and wherein the controller is adapted to constrain a start state of the decoder to a state corresponding to the last k−1 of the known padding bits prepending the data bits.
21. The system as claimed in claim 20, wherein the controller is adapted to discard symbols of the encoded data block including:
- disabling operation of the Viterbi decoder while latching the first n bits of the data block into the shift register of the Viterbi decoder.
22. The system as claimed in claim 19, wherein the known padding bits comprise n known padding bits appended to the data bits, and wherein the controller is adapted to constrain an End state of the decoder to a state corresponding to the first k−1 of the n known padding bits appending the data bits.
23. The system as claimed in claim 22, wherein the controller is adapted to discard symbols of the encoded data block including:
- discarding any remaining bits of the encoded data block after processing and the first e=(N−n)+(k−1) bits of the encoded data block through the Viterbi decoder, where N is the total length of the data block.
24. The system as claimed in claim 19, wherein the convolutionally encoded data block is a data block of an Enhanced Data for Global Evolution, EDGE, data frame.
25. A computer program product for decoding a convolutionally encoded data block having known padding bits, the computer program product comprising a machine readable medium tangibly embodying instructions to cause a computing device to:
- constrain a Viterbi decoder to a state corresponding to k−1 padding bits immediately adjacent data bits of the data block, where k is a constraint length of a convolutional encoder used to encode the data block, and wherein the k−1 padding bits immediately adjacent data bits of the data block comprise less than the predetermined number of known padding bits; and
- discard symbols of the encoded data block that have influence only from padding bits other than the k−1 padding bits immediately adjacent data bits of the data block.
26. The computer program product as claimed in claim 25, wherein the known padding bits comprise n known padding bits prepended to the data bits, and wherein the instructions further cause the computing device to constrain a Start state of the decoder to a state corresponding to the last k−1 of the n known padding bits prepending the data bits.
27. The computer program product as claimed in claim 26, wherein the instructions further cause the computing device to discard symbols of the encoded data block including:
- disabling operation of the Viterbi decoder while latching the first n bits of the data block into the shift register of the Viterbi decoder.
28. The computer program product as claimed in claim 25, wherein the known padding bits comprise n known padding bits appended to the data bits, and wherein the instructions further cause the computing device to constrain an End state of the decoder to a state corresponding to the first k−1 of the known padding bits appending the data bits.
29. The computer program product as claimed in claim 28, wherein the instructions further cause the computing device to discard symbols of the encoded data block including:
- discarding any remaining bits of the encoded data block after processing the first e=(N−n)+(k−1) bits of the encoded data block through the Viterbi decoder, where N is the total length of the data block.
30. The computer program product as claimed in claim 25, wherein the convolutionally encoded data block is a data block of an Enhanced Data for Global Evolution, EDGE, data frame.
Type: Application
Filed: Sep 13, 2012
Publication Date: Jan 10, 2013
Applicant: RESEARCH IN MOTION LIMITED (Waterloo)
Inventor: Phat TRAN (Waterloo)
Application Number: 13/613,632
International Classification: H03M 13/23 (20060101); H03M 13/41 (20060101);