Unified technique for multi-rate trellis coding and decoding
원문보기
IPC분류정보
국가/구분
United States(US) Patent
등록
국제특허분류(IPC7판)
H03M-013/25
H03M-013/41
출원번호
US-0639714
(2000-08-14)
발명자
/ 주소
Halder, Bijit
출원인 / 주소
Virata Corporation
대리인 / 주소
Burns, Doane, Swecker & Mathis, LLP
인용정보
피인용 횟수 :
7인용 특허 :
19
초록▼
The present invention, generally speaking, provides efficient multi-rate trellis encoder and decoder structures. The trellis encoder allows for a variable number of uncoded bits to be represented in a transmit symbol. The decoder maps received symbols to a smaller constellation by dropping selected
The present invention, generally speaking, provides efficient multi-rate trellis encoder and decoder structures. The trellis encoder allows for a variable number of uncoded bits to be represented in a transmit symbol. The decoder maps received symbols to a smaller constellation by dropping selected symbol bits, whereby, for each of multiple cosets, points within that coset are mapped to a fewer number of points. Substantial simplification of the decoder structure is therefore achieved.
대표청구항▼
1. In a digital transmission system in which data symbols are transmitted using a convolutional encode channel coding arrangement and a one-dimensional modulation coding arrangement, a method of varying data transmission rate, comprising:within a framework of the one-dimensional modulation coding ar
1. In a digital transmission system in which data symbols are transmitted using a convolutional encode channel coding arrangement and a one-dimensional modulation coding arrangement, a method of varying data transmission rate, comprising:within a framework of the one-dimensional modulation coding arrangement, identifying a set of modulated values;for each of multiple different data rates, performing convolutional coding by operating on m data bits to produce n data bits where n>m, the particular m data bits being operated upon being the same for each different data rate;for one data rate, mapping data symbols including a portion of said m data bits to a complete set of said modulated values; andfor another different data rate, mapping data symbols including a portion of said m data bits to a subset of said modulated values. 2. The method of claim 1, wherein the number of modulated values is 2 k where k≧2. 3. The method of claim 2, wherein k=4. 4. The method of claim 2, wherein the digital data transmission system is an xDSL system. 5. The method of claim 4, wherein the one-dimensional modulation coding arrangement is Pulse Amplitude Modulation. 6. The method of claim 5, wherein the multiple different data rates correspond to at least one of 16 PAM and 32 PAM. 7. The method of claim 6, wherein one or more additional data rates of the one-dimensional modulation coding arrangement correspond to one or more of the following: 4 PAM and 8 PAM. 8. A multi-rate digital transmission system in which data symbols are transmitted using a convolutional encode channel coding arrangement and a one-dimensional modulation coding arrangement, a set of modulated values having been identified within a framework of the one-dimensional modulation coding arrangement, comprising:a data-converter for receiving a serial data input signal and responsive to a control signal producing a variable-width parallel data output signal wherein, in at least one mode of operation of the data converter, at least one bit of the variable-width parallel data output signal is zero-filled;a convolutional encoder responsive to a subset of one or more bits of the variable-width parallel data output signal for producing a number of encoded bits greater than the number of said one or more bits; anda constellation mapper responsive to a control signal and to data symbols for producing a modulated value, the data symbols including at least a portion of the variable-width parallel data output signal and said encoded bits;wherein, for one data rate, data symbols are mapped to a complete set of said modulated values, and for another different data rate, data symbols are mapped to a subset of said modulated values. 9. The apparatus of claim 8, wherein the number of modulated values is 2 k where k≧2. 10. The apparatus of claim 9, wherein k=4. 11. The apparatus of claim 9, wherein the digital data transmission system is an xDSL system. 12. The apparatus of claim 9, wherein the one-dimensional modulation coding arrangement is Pulse Amplitude Modulation. 13. The apparatus of claim 12, wherein the multiple different data rates correspond to at least 16 PAM and 32 PAM. 14. The apparatus of claim 13, wherein one or more additional data rates of the one-dimensional modulation coding arrangement correspond to one or more of the following: 4 PAM and 8 PAM. 15. In a digital transmission system in which data symbols are transmitted using a convolutional channel coding arrangement and a modulation coding arrangement, a method of decoding a transmitted data signal, comprising:identifying within a modulation constellation multiple cosets of constellation points;mapping received symbols to a smaller constellation by dropping selected symbol bits, whereby, for each of multiple cosets, points within that coset are mapped to a fewer number of points;generating a reduced number of branch metric values using said smaller constellation; andperforming add-compare-select operations based on said reduced number of branch metric values. 16. The method of claim 15, further comprising:providing a path table having a distinct row for each state of a convolutional encoder and having multiple columns in accordance with a history depth of the decoder;storing within the path table binary values representing paths within a trellis corresponding to the convolutional encoder; andperforming a path update operation by, for each of multiple rows of the path table, rewriting the row with values from a different row. 17. The method of claim 16, wherein add-compare-select operations comprise computing a path-weight vector of a dimension comparable to a dimension of a column of the path table, further comprising:identifying an index value corresponding to a minimum path weight; andoutputting as a decoded bit a value stored within a specified column of the path table at a location corresponding to said index value. 18. The method of claim 17, further comprising updating the path table by dropping said specified column and adding a new column at a different location within the path table, wherein the specified column and the new column are separated by a number of columns substantially equal to the history depth of the decoder. 19. A decoder for decoding data symbols transmitted using a convolutional channel coding arrangement and a modulation coding arrangement, wherein a combination of encoded bits produced by the channel coding arrangement and uncoded bits obtained directly from a data source are modulation encoded, comprising:a convolutional decoder responsive to a received data symbol for decoding said encoded bits to produce decoded bits; anda decision block responsive to at least a portion of the received data symbol and to said decoded bits for producing estimates of said uncoded bits;wherein the decision block has multiple modes and the number of uncoded bits produced is different in different modes. 20. The apparatus of claim 19, wherein the channel coding arrangement and modulation coding arrangement are in accordance with an HDSL2/G.SHDSL standard. 21. The apparatus of claim 20, wherein the number of uncoded bits is greater than two. 22. A encoder for coding data symbols to be transmitted using a convolutional channel coding arrangement and a modulation coding arrangement, wherein a combination of encoded bits produced by the channel coding arrangement and uncoded bits obtained directly from a data source are modulation encoded, comprising:a convolutional encoder by operating on m data bits to produce n encoded bits where n>m; anda bits-to-symbol mapper responsive to said encoded bits and to data bits other than said m data bits for producing a data symbol to be transmitted;wherein the bits-to-symbol mapper has multiple modes, and the number of uncoded bits used by the bits-to-symbol mapper is different in different modes. 23. A decoder for decoding data symbols transmitted using a convolutional channel coding arrangement and a modulation coding arrangement, comprising:a branch metric generator for mapping received symbols to a smaller constellation by dropping selected symbol bits, whereby, for each of multiple cosets, points within that coset are mapped to a fewer number of points, and for generating a reduced number of branch metric values using said smaller constellation; andand add-compare-select block for performing add-compare-select operations based on said reduced number of branch metric values. 24. The method of claim 15, further comprising:a path table having a distinct row for each state of a convolutional encoder and having multiple columns in accordance with a history depth of the decoder, the add-compare-select block storing within the path table binary values representing paths within a trellis corresponding to the convolutional encoder; anda path update block for performing a path update operation by, for each of multiple rows of the path table, rewriting the row with values from a differen t row. 25. The method of claim 16, wherein:the add-compare-select block computes a path-weight vector of a dimensional comparable to a dimension of a column of the path table, identifies an index value corresponding to a minimum path weight, and applies the index value to the path update block; andthe path update block outputs as a decoded bit a value stored within a specified column of the path table at a location corresponding to said index value. 26. The method of claim 17, wherein the path update block updates the path table by dropping said specified column and adding a new column at a different location within the path table, wherein the specified column and the new column are separated by a number of columns substantially equal to the history depth of the decoder.
Rhee Dojun ; Souvanthong Chanthachith,FRX, Apparatus and method for recovering information bits from a 64/256-quadrature amplitude modulation treliss coded modulation decoder.
Choi Hyung-Jin,KRX ; Cho Sung-Bae,KRX ; Jung Suk-Jin,KRX ; Lee Hyung-Kil,KRX, Decoding method and apparatus using trace deletion for Viterbi algorithm.
Zhong Yan ; Yang Lin ; Rafie Manouchehr, Method and apparatus for generating branch metrics and branch indices for convolutional code Viterbi decoders.
Cumberton John (Danbury CT) Foster Alun E. (Groot-Bijgaarden BEX) Rae Alexander S. (Farnham GB2) Fernandes Vernon J. (Tilehurst GB2) Carter Michael R. (Maidenhead GB2), Soft decision convolutional code transmission systems.
※ AI-Helper는 부적절한 답변을 할 수 있습니다.