WO1999000905A1 - Processeur et procede de traitement - Google Patents
Processeur et procede de traitement Download PDFInfo
- Publication number
- WO1999000905A1 WO1999000905A1 PCT/JP1998/002909 JP9802909W WO9900905A1 WO 1999000905 A1 WO1999000905 A1 WO 1999000905A1 JP 9802909 W JP9802909 W JP 9802909W WO 9900905 A1 WO9900905 A1 WO 9900905A1
- Authority
- WO
- WIPO (PCT)
- Prior art keywords
- arithmetic processing
- path
- metric
- processing device
- metrics
- Prior art date
Links
Classifications
-
- H—ELECTRICITY
- H03—ELECTRONIC CIRCUITRY
- H03M—CODING; DECODING; CODE CONVERSION IN GENERAL
- H03M13/00—Coding, decoding or code conversion, for error detection or error correction; Coding theory basic assumptions; Coding bounds; Error probability evaluation methods; Channel models; Simulation or testing of codes
- H03M13/65—Purpose and implementation aspects
- H03M13/6502—Reduction of hardware complexity or efficient processing
-
- H—ELECTRICITY
- H03—ELECTRONIC CIRCUITRY
- H03M—CODING; DECODING; CODE CONVERSION IN GENERAL
- H03M13/00—Coding, decoding or code conversion, for error detection or error correction; Coding theory basic assumptions; Coding bounds; Error probability evaluation methods; Channel models; Simulation or testing of codes
- H03M13/37—Decoding methods or techniques, not specific to the particular type of coding provided for in groups H03M13/03 - H03M13/35
- H03M13/39—Sequence estimation, i.e. using statistical methods for the reconstruction of the original codes
- H03M13/3961—Arrangements of methods for branch or transition metric calculation
-
- H—ELECTRICITY
- H03—ELECTRONIC CIRCUITRY
- H03M—CODING; DECODING; CODE CONVERSION IN GENERAL
- H03M13/00—Coding, decoding or code conversion, for error detection or error correction; Coding theory basic assumptions; Coding bounds; Error probability evaluation methods; Channel models; Simulation or testing of codes
- H03M13/37—Decoding methods or techniques, not specific to the particular type of coding provided for in groups H03M13/03 - H03M13/35
- H03M13/39—Sequence estimation, i.e. using statistical methods for the reconstruction of the original codes
- H03M13/41—Sequence estimation, i.e. using statistical methods for the reconstruction of the original codes using the Viterbi algorithm or Viterbi processors
- H03M13/4107—Sequence estimation, i.e. using statistical methods for the reconstruction of the original codes using the Viterbi algorithm or Viterbi processors implementing add, compare, select [ACS] operations
-
- H—ELECTRICITY
- H03—ELECTRONIC CIRCUITRY
- H03M—CODING; DECODING; CODE CONVERSION IN GENERAL
- H03M13/00—Coding, decoding or code conversion, for error detection or error correction; Coding theory basic assumptions; Coding bounds; Error probability evaluation methods; Channel models; Simulation or testing of codes
- H03M13/37—Decoding methods or techniques, not specific to the particular type of coding provided for in groups H03M13/03 - H03M13/35
- H03M13/39—Sequence estimation, i.e. using statistical methods for the reconstruction of the original codes
- H03M13/41—Sequence estimation, i.e. using statistical methods for the reconstruction of the original codes using the Viterbi algorithm or Viterbi processors
- H03M13/4161—Sequence estimation, i.e. using statistical methods for the reconstruction of the original codes using the Viterbi algorithm or Viterbi processors implementing path management
- H03M13/4169—Sequence estimation, i.e. using statistical methods for the reconstruction of the original codes using the Viterbi algorithm or Viterbi processors implementing path management using traceback
-
- H—ELECTRICITY
- H03—ELECTRONIC CIRCUITRY
- H03M—CODING; DECODING; CODE CONVERSION IN GENERAL
- H03M13/00—Coding, decoding or code conversion, for error detection or error correction; Coding theory basic assumptions; Coding bounds; Error probability evaluation methods; Channel models; Simulation or testing of codes
- H03M13/65—Purpose and implementation aspects
- H03M13/6561—Parallelized implementations
-
- H—ELECTRICITY
- H03—ELECTRONIC CIRCUITRY
- H03M—CODING; DECODING; CODE CONVERSION IN GENERAL
- H03M13/00—Coding, decoding or code conversion, for error detection or error correction; Coding theory basic assumptions; Coding bounds; Error probability evaluation methods; Channel models; Simulation or testing of codes
- H03M13/65—Purpose and implementation aspects
- H03M13/6563—Implementations using multi-port memories
-
- H—ELECTRICITY
- H03—ELECTRONIC CIRCUITRY
- H03M—CODING; DECODING; CODE CONVERSION IN GENERAL
- H03M13/00—Coding, decoding or code conversion, for error detection or error correction; Coding theory basic assumptions; Coding bounds; Error probability evaluation methods; Channel models; Simulation or testing of codes
- H03M13/65—Purpose and implementation aspects
- H03M13/6569—Implementation on processors, e.g. DSPs, or software implementations
Definitions
- the present invention relates to an arithmetic processing device that is incorporated in a mobile communication device or the like, and particularly performs an ACS (addition, comparison, selection) operation of Viterbi decoding.
- ACS addition, comparison, selection
- DSP digital signal processor
- Viterbi decoding achieves maximum likelihood decoding of convolutional codes by repeating simple processing of addition, comparison, and selection, and finally by a traceback operation to decode data.
- a convolutional code is generated by mod 2 addition of an input bit and a certain number of bits preceding the input bit, and a plurality of encoded data are generated corresponding to one input bit.
- the number of input information bits that affect the encoded data is called the constraint length (K), and the number is equal to the number of shift registers used for mod 2 addition.
- the coded data is determined by the input bits and the state of the preceding (K-1) input bits, and transitions to a new state when a new information bit is input.
- the state to which the coded data transits depends on whether the new input bit is 0 or 1. Since each of the (K-1) bits can be 1 or 0, the number of states to which the coded data transitions is 2 ( ⁇ _1 ) .
- the received encoded data sequence is observed, and the most probable state is estimated from all possible state transitions. Therefore, each time encoded data (received data sequence) corresponding to one information bit is obtained, the signal distance (metric) of the path to each state at that time is calculated, and the path of the path reaching the same state is calculated.
- the path metric a is the distance between signals (branch metric) X between the output symbol of the path input to the state S [2 n] and the received data sequence (branch metric), and the surviving path up to the previous state S [n]. This is the sum with the path metric A, which is the sum of the branch metrics.
- the path metric b is the distance (branch metric) y between the output symbol of the path input to the state S [2 n] and the received data sequence (branch metric), and the state S [n + 2 ( ⁇ — 2) ] is the sum with the path metric ⁇ , which is the sum of the branch metrics of the surviving paths up to and including].
- the path metrics a and b input to the state S [2 ⁇ ] are compared, and the smaller path is selected as the surviving path.
- a conventional arithmetic processing device for Viterbi decoding will be described with reference to an example of a general-purpose arithmetic processing device TMS320C54x (manufactured by TEXAS INSTRUMENTS, hereinafter referred to as "C54x").
- TMS320C54x manufactured by TEXAS INSTRUMENTS, hereinafter referred to as "C54x"
- C54x the following equation (1) is used as a convolutional code.
- G 2 (D) 1 + D + D 3 + D4 (1)
- the branch metric is defined by the following equation (2).
- the butterfly structure is processed at high speed by setting the Arithmetic Logic Unit (ALU) in a dual 16-bit mode.
- ALU Arithmetic Logic Unit
- To determine the new path metric (J), two path metrics, 2 * J and 2 * J + 1, and a branch metric (M and —M) are calculated in parallel with the DADST instruction and compared with the CMP S instruction. I do.
- To determine the new path metric (J + 8), the two path metrics and the branch metric (M and 1M) are calculated in parallel using the DSAD instruction. The operation results are stored in the upper and lower parts of the double-precision accumulator respectively.
- the new path metric is determined by the CMP S instruction.
- the CMPS instruction compares the upper and lower accumulators and stores the larger of these values in memory. Each time a comparison is made, the 16-bit transition and the register (TRN) indicate which is selected. The contents written to TRN are stored in memory after each symbol processing. The information stored in memory is used to find the best path during the traceback process.
- Figure 4 shows a macro program for butterfly computation for Viterbi decoding. The value of the branch metric is stored in the T registry before the macro is called.
- Figure 5 shows an example of path metric memory mapping.
- the above is the ACS operation mechanism of the general-purpose DSP C54X.
- the C54x realizes updating of two path metrics in four machine cycles.
- demand for non-voice communication which requires higher transmission quality with a lower bit error rate than voice communication, is expected to increase.
- the number of path metrics (the number of states) doubles, and the amount of computation in Viterbi decoding by the DSP doubles.
- non-voice communication has a larger amount of information than voice communication, and as the amount of information increases, the amount of operation in Viterbi decoding including ACS operation also increases.
- the amount of computation by the DSP increases, it becomes difficult to maintain the battery of the mobile terminal for a long time.
- a first object of the present invention is to provide an arithmetic processing device that efficiently processes ACS operation of video decoding by a DSP with a small investment of hardware. This object is achieved by arranging two sets of the comparison unit, the addition unit, and the storage unit for storing the comparison result in the processing unit, and executing the ACS operation in parallel.
- FIG. 1 is a trellis diagram showing a state transition path of a convolutional encoder in video decoding.
- FIG. 2 is a schematic diagram showing a butterfly structure of a trellis diagram
- FIG. 3 is a schematic diagram showing an example of a generated code by a convolutional encoder
- FIG. 4 is a program diagram showing an example of a video operation for channel coding.
- Fig. 5 is a schematic diagram showing an example of storing pointer control and path metrics.
- FIG. 6 is a block diagram showing the configuration of the arithmetic processing unit according to Embodiment 1 of the present invention.
- FIG. 7 is a block diagram illustrating an example of a convolutional encoder having a coding rate of 1/2
- FIG. 9 is a block diagram illustrating a configuration of an arithmetic processing device according to Embodiment 2 of the present invention.
- FIG. 10 is a timing chart illustrating a pipeline operation of the arithmetic processing unit according to Embodiment 2 of the present invention.
- FIG. 11 is a schematic diagram showing an operation example of RAM memory access according to the second embodiment of the present invention.
- FIG. 12 is a block diagram illustrating a configuration of an arithmetic processing device according to Embodiment 3 of the present invention.
- FIG. 13 is a schematic diagram showing an operation example of memory access of a dual-port RAM according to the third embodiment of the present invention.
- FIG. 14 is a block diagram illustrating a configuration of an arithmetic processing device according to Embodiment 4 of the present invention.
- FIG. 15 is a timing chart illustrating a pipeline operation of the arithmetic processing unit according to Embodiment 4 of the present invention.
- FIG. 16 is a block diagram illustrating a configuration of an arithmetic processing device according to Embodiment 5 of the present invention.
- FIG. 17 is a diagram showing an ACS operation result of the arithmetic processing unit according to Embodiment 6 of the present invention.
- FIG. 18 is a block diagram illustrating a configuration of an arithmetic processing device according to Embodiment 6 of the present invention.
- FIG. 19 is a block diagram showing a configuration of an arithmetic processing unit according to Embodiment 7 of the present invention.
- FIG. 20 is a block diagram showing a configuration of an arithmetic processing unit according to Embodiment 8 of the present invention.
- FIG. 21 shows an input / output diagram of 4: 2 COMPRESOR in Embodiment 8 of the present invention.
- FIG. 22 is a block diagram illustrating a configuration of an arithmetic processing device according to Embodiment 9 of the present invention.
- FIG. 23 shows the carry control of the double precision AU.
- FIG. 24 is a block diagram showing a configuration of the arithmetic processing device according to Embodiment 10 of the present invention.
- FIG. 25 is a block diagram showing a configuration of the arithmetic processing device according to Embodiment 11 of the present invention.
- FIG. 26 is a block diagram showing a configuration of a mobile station device according to Embodiment 12 of the present invention.
- FIG. 27 is a block diagram illustrating a configuration of a mobile station device according to Embodiment 13 of the present invention.
- FIG. 28 is a block diagram showing a configuration of a base station apparatus according to Embodiment 14 of the present invention.
- FIG. 29 is a block diagram showing a configuration of a base station apparatus according to Embodiment 15 of the present invention.
- FIG. 6 is a block diagram showing a configuration of the arithmetic processing device according to the first embodiment.
- a path metric storage unit 1 stores a path metric, and supplies data and transfers a calculation result via a bus 2.
- the branch metric storage unit 3 stores the branch metric and supplies data via the bus 4.
- the comparison unit 5 and the comparison unit 9 compare the data inputted from the path metric storage unit 1 and the branch metric storage unit 3 via the bus 2 and the bus 4, respectively.
- the adder 6 and the adder 10 add the data read from the path metric storage 1 and the branch metric storage 3 via the buses 2 and 4, respectively.
- the comparison result storage unit 7 stores the comparison result of the comparison unit 5, and the comparison result storage unit 11 stores the comparison result of the comparison unit 9. Then, the comparison result storage unit 7 and the comparison result storage unit 11 transfer the comparison result to the path metric storage unit 1 via the bus 2.
- the selection unit 8 receives the addition result of the addition unit 6, determines the output based on the comparison result of the comparison unit 5, and the selection unit 12 receives the addition result of the addition unit 10 and compares it with the comparison unit 9. The output is determined based on the result. Then, the selection units 8 and 12 transfer the selection result determined via the bus 13 to the path metric storage unit 1.
- the ACS calculation processing of the calculation processing device will be described with reference to the drawings.
- the path metric and branch metric data types are both single-precision data.
- the double precision data is (X, ⁇ )
- the upper side of the double precision data is represented by X
- the lower side of the double precision data is represented by ⁇ .
- node NO and node N1 are node N'0 and node N'4.
- the branch metric when transitioning from node NO to node N '0 is BM0
- the branch metric when transitioning from node N 1 to node N' 0 is BM1
- the branch metric when transitioning from node NO to node N '4 The metric is BM1
- the branch metric when transitioning from node N1 to node N'4 is BM0.
- branch metrics BM0 and BM1 are exchanged and added to the path metric ⁇ 40 of the node NO and the path metric PM1 of the node N1, respectively.
- a path metric of 4 results in PM'4.
- the ACS operation from the first node N ′ 0 to the node N ′ 3 is performed by the comparison unit 5, the addition unit 6, the comparison result storage unit 7 and the selection unit 8, and in parallel with this, the second half of the node N ′ 4 From the node NO and the node N1 to the node N'0 and the node N'4 from the comparison unit 9, the addition unit 10, the comparison result storage unit 11 and the selection unit 12.
- the ACS operation of will be described in detail.
- two path metrics (PM1, PM0) from the path metric storage unit 1 are output to the bus 2
- two branch metrics (BM1, BM0) from the branch metric storage unit 3 are output to the bus 4.
- the comparison unit 5 inputs two path metrics (PM1, PM0) from the bus 2 and inputs two branch metrics (BM1, BM0) from the bus 4, and calculates PM1 + BM1-PM0-BM0. .
- the adder 6 inputs two path metrics (PM1, PM0) from the bus 2 and inputs two branch metrics (BM1, BM0) from the bus 4, and outputs PM1 + BM1 and PMO + BM0. Is calculated and output to the selection unit 8 as (PM1 + BM1, PM0 + BM0).
- the selecting unit 8 inputs the most significant bit (Most Significant Bit: hereinafter, referred to as “MSB”), which is the sign bit of the comparison result P Ml + BM 1 -PMO—BM 0 of the comparing unit 5, Select whether to output PM1 + BM1 or output lower PM0 + BM0 according to the MSB value.
- MSB most significant bit
- MSB which is the comparison result of the comparison unit 5, is simultaneously stored in the comparison result storage unit 7.
- the comparison unit 9 receives two path metrics (PM1, PM0) from the bus 2 and two branch metrics (BM1, BM0) from the bus 4, and calculates PM1 + BM0—PM0—BM1.
- the adder 10 inputs two path metrics (PM1, PMO) from the bus 2 and two branch metrics (BM1, BM0) from the bus 4, and outputs PM1 + BM0 and PMO + BM1. Is calculated and output to the selection unit 12 as (PM1 + BM0, PM0 + BM1).
- the selection unit 12 calculates the comparison result of the comparison unit 9 PM 1 + BM0—PM0—BM1 Input B and select whether to output upper PM1 + BM0 or lower PM0 + BM1 according to the value of MSB.
- the MSB which is the comparison result of the comparison unit 9, is stored in the comparison result storage unit 11 at the same time.
- the ACS operation of video decoding by DSP can be executed in parallel, and the arithmetic processing can be performed at high speed with a relatively small amount of processing. it can.
- FIG. 9 is a block diagram illustrating a configuration of the arithmetic processing device according to the second embodiment.
- the same parts as those of the arithmetic processing device shown in FIG. 9 the same parts as those of the arithmetic processing device shown in FIG.
- the storage unit for storing the path metrics is configured by the RAM 14 composed of four banks.
- the arithmetic processing device shown in FIG. 9 is suitable for the arithmetic processing of the pipeline structure shown in FIG.
- RAMI 4 power Double-precision RAM that can read even addresses and odd addresses consecutively Assuming that the RAM is readable, if the following conditions (a) and (b) are satisfied, calculations can be performed simply by specifying the even addresses. You can read out the two path metrics used for
- a one-state path metric is stored in an even address and an odd address in consecutive addresses.
- One-state path metrics are divided into the first half and the second half, and each is stored in a separate bank.
- bank 0 of RAM 14 stores the path metrics of the first half of the old state (PMO, PM1, PM2, PM3 in Fig. 8), and bank 1 stores the path metrics of the second half of the old state (PM4, PM in Fig. 8).
- PMO path metrics of the first half of the old state
- PM4 path metrics of the second half of the old state
- two path metrics are generated by executing one cycle of the ACS operation, and these are stored in bank 2 and bank 3 via the bus 13, respectively.
- the double-precision data is transferred from the bus 13, the path metrics from the node N'0 to the node N'3 are stored in the bank 2, and the path metrics from the node N 'to the node N'7 are stored in the bank 3. Is stored.
- FIG. 11 is a schematic diagram showing an operation example of memory access of the RAM 14 corresponding to FIG.
- the storage unit for storing the path metric uses the RAM 14 composed of four banks, and the pair of the bank for reading the path metric and the pair of the bank for storing the storage each time the one-state ACS operation is completed. By switching, the ACS calculation of Viterbi decoding by DSP can be executed in parallel.
- bank 0 and bank 1 and bank 2 and bank 3 Is a pair of banks, but other combinations can be implemented in the same way by simply changing the address supplied at the memory access stage and the address at the time of storage.
- the RAM 14 is configured by four banks, but the same can be implemented as long as the number of banks is four or more.
- FIG. 12 is a block diagram showing a configuration of the arithmetic processing device according to the third embodiment.
- the same parts as those of the arithmetic processing device shown in FIG. 12 the same parts as those of the arithmetic processing device shown in FIG. 12
- a storage unit for storing a path metric is configured by a dual-port RAM 15 composed of three banks.
- the arithmetic processing device shown in FIG. 12 is suitable for the arithmetic processing of the pipeline structure shown in FIG.
- reading and writing to the same bank can be designated by one instruction.
- instruction 1 to execute the ACS operation in the operation execution stage of the (n + 1) th cycle, the address of the path metric to be read in the memory access stage in the nth cycle and the address of the path metric to be written to the dual port RAMI 5 Supply.
- instruction 1 to execute the ACS operation in the operation execution stage of the (n + 1) th cycle, the address of the path metric to be read in the memory access stage in the nth cycle and the address of the path metric to be written to the dual port RAMI 5 Supply.
- the (n + 1) th cycle even and odd addresses can be read consecutively from the dual port RAMI 5, an ACS operation is performed, and one path metric can be written to the same bank.
- the arithmetic processing device can read out two path metrics used for the operation only by specifying an even address.
- a one-state path metric is stored in an even address and an odd address in consecutive addresses.
- the one-state path metric is divided into the first half and the second half, and each is stored in a separate bank.
- the first half path metric of the old state (PM0, PM1, PM2, PM3 in Fig. 8) is stored in bank 0 of dual port RAMI 5, and the second half of the old state path metric (PM4 in Fig. 8).
- PM5, PM6, PM7) are stored
- two path metrics are generated in one cycle of execution of the ACS operation, and these are stored in bank 0 and bank 2 via bus 13.
- the bus 13 transfers double-precision data
- the path metrics from node N'0 to node N'3 are stored in bank 0, and the bank 2 stores the path metrics from node N'4 to node N'7. Is stored.
- FIG. 13 is a schematic diagram showing an operation example of memory access of the dual-port RAM 15 corresponding to FIG.
- the arithmetic processing unit shown in FIG. 12 performs only switching between bank 1 and bank 2 and, without switching bank 0, performs the ACS operation for Viterbi decoding by DSP in parallel. Can be performed.
- the dual-port RAMI 5 is configured with three banks, but the same can be implemented as long as the number of banks is three or more.
- FIG. 14 is a block diagram showing a configuration of the arithmetic processing device according to the fourth embodiment.
- the same portions as those of the arithmetic processing device shown in FIG. 6 are denoted by the same reference numerals, and description thereof will be omitted.
- the arithmetic processing unit shown in FIG. 14 includes an input register 16 and an input register 17 for inputting data from the bus 2 and outputting data to the comparing unit 5, the comparing unit 9, the adding unit 6 and the adding unit 10.
- the arithmetic processing device shown in FIG. 14 is suitable for the arithmetic processing of the pipeline structure shown in FIG.
- the ACS operation is executed in the operation execution stage of the (n + 2) th cycle of the instruction 1
- the path metric address to be read in the memory access stage of the nth cycle is supplied to the RAM 14 in advance, and the (n + 1) th cycle is executed.
- the data output from the RAM 14 is latched via the bus 2 into the input register 16 and the input register 17.
- one stage of overnight transfer is inserted in the pipeline stage shown in FIG. 10 before the operation execution stage. That is, at the beginning of the operation execution stage, the data output from the RAM 14 is the input register before each of the calculators (comparator 5, comparator 9, adder 6, and adder 10). Since it is determined, the time required for data transfer from the RAM 14 can be saved.
- FIG. 16 is a block diagram showing a configuration of the arithmetic processing device according to the fifth embodiment.
- the same parts as those of the arithmetic processing device shown in FIG. 16 the same parts as those of the arithmetic processing device shown in FIG.
- the arithmetic processing device shown in FIG. 16 is different from the arithmetic processing device in FIG. 14 in that a swap circuit 18 for outputting data input from the branch metric storage unit 3 as it is or for exchanging upper and lower order and outputting the data is added. .
- the arithmetic processing device shown in FIG. 16 is suitable for the arithmetic processing of the pipeline structure shown in FIG.
- the branch circuit 18 when the branch circuit 18 receives double-precision data from the branch metric storage unit 3 in the form of ⁇ BM1, BM0 ⁇ , it outputs the values of the two branch metrics as ⁇ BM1, BM0 ⁇ as they are.
- Upper and lower It has a function to switch between output and output as ⁇ BM0, BM1 ⁇ by an instruction or the like.
- both ACS operations use a common branch metric BM0 and BM1 and BM0 and BM1 It is a skipped relationship.
- the ACS operation from the node NO and the node N1 to the node N'0 and the ACS operation from the node N6 and the node N7 to the node N'3 are performed by the comparing unit 5 and the adding unit 6.
- the ACS operation from the node NO and the node N1 to the nodes N and 4 and the ACS operation from the node N6 and the node N7 to the node N'7 are performed by the comparing unit 9 and the adding unit 10.
- the swap circuit 18 solves such redundancy. For example, only the form of ⁇ BM0, BM1 ⁇ is stored in the branch metric storage unit 3, and ⁇ BM0, BM1 ⁇ is stored in the swap circuit 18. By inputting BM 1 ⁇ and switching and outputting ⁇ BM0, BM 1 ⁇ or ⁇ BM1, BM0 ⁇ by an instruction or the like, the redundancy of the branch metric storage unit 3 can be omitted.
- the above relationship also holds for N3, node N4, and node N5,
- the present invention can be implemented in the same manner, because the above holds true for other combinations of the constraint length K and the coding rate. Further, the same can be implemented by using a dual-port RAM as a storage unit for storing a path metric.
- FIG. 18 is a block diagram showing a configuration of the arithmetic processing device according to the sixth embodiment.
- the same parts as those of the arithmetic processing device shown in FIG. 16 are denoted by the same reference numerals, and description thereof will be omitted.
- the arithmetic processing unit shown in FIG. 18 is different from the arithmetic processing unit in FIG. 16 in that the comparison unit 5 is composed of an adder 19, an adder 20 and a comparator 21, and the addition unit 6 is an adder. It consists of 22 and adder 23. Further, the comparing section 9 is constituted by an adder 24, an adder 25 and a comparator 26, and the adding section 10 is constituted by an adder 27 and an adder 28.
- adders 19 and 20 input data from bus 4 and input register 16 and add them.
- the comparator 21 receives the addition results from the adders 19 and 20 and compares them, and outputs the comparison results to the comparison result storage unit 7 and the selection unit 8.
- the adders 22 and 23 input the data from the bus 4 and the input register 16 and add them together, and output the addition result to the selector 8.
- the adders 24 and 25 receive data from the bus 4 and the input register 17 and add the data.
- the comparator 26 receives the addition results from the adders 24 and 25 and compares them, and outputs the comparison results to the comparison result storage unit 11 and the selection unit 12.
- the adder 27 and the adder 28 input the data from the bus 4 and the input register 17 and add them, and output the addition result to the selector 12.
- the arithmetic processing device shown in FIG. 18 is suitable for the arithmetic processing of the pipeline structure shown in FIG.
- the comparator 21 inputs the addition result ⁇ A + C ⁇ of the adder 19 and the addition result ⁇ B + D ⁇ of the adder 20 and compares ⁇ A + C— (B + D) ⁇ . Output the MSB of the result.
- the adder 22 inputs the path metric ⁇ A ⁇ and the branch metric ⁇ C ⁇ and outputs an addition result ⁇ A + C ⁇ , and the adder 23 inputs the path metric ⁇ B ⁇ and the branch metric ⁇ D ⁇ . And outputs the addition result ⁇ B + D ⁇ .
- the adder 24 inputs the path metric ⁇ A ⁇ and the branch metric ⁇ D ⁇ , outputs the addition result ⁇ A + D ⁇ , and the adder 25 outputs the path metric ⁇ B ⁇ and the branch metric ⁇ C ⁇ . Input and output the addition result ⁇ B + C ⁇ .
- the comparator 26 inputs the addition result ⁇ A + D ⁇ of the adder 24 and the addition result ⁇ B + C ⁇ of the adder 25, performs a comparison of ⁇ A + D— (B + C) ⁇ , and performs a comparison. Output the MSB of the result.
- the adder 27 inputs the path metric ⁇ A ⁇ and the branch metric ⁇ D ⁇ , and outputs the addition result ⁇ A + D ⁇ .
- the adder 28 inputs the path metric ⁇ B ⁇ and the branch metric ⁇ C ⁇ . And outputs the addition result ⁇ B + C ⁇ .
- two path metrics can be updated in one machine cycle by pipeline operation by DSP.
- the above relationship holds for the nodes N3, N4, and N5, and also holds for other combinations of the constraint length K and the coding rate.
- the present invention can be similarly implemented by using a dual port RAM as the storage unit for storing the path metric.
- FIG. 19 is a block diagram showing a configuration of the arithmetic processing device according to the seventh embodiment.
- the same portions as those of the arithmetic processing device shown in FIG. 19 the same portions as those of the arithmetic processing device shown in FIG.
- the arithmetic processing unit shown in FIG. 19 uses an arithmetic logic unit (hereinafter, referred to as “ALU”) 29 instead of the comparator 21 in comparison with the arithmetic processing unit in FIG. It has an input register 30, an input register 31, a bus 32, a node 33, a bus 37, a bus 38, and selectors 34 and 35.
- ALU arithmetic logic unit
- the input register 30 inputs data from the RAM 14 via the bus 37
- the input register 31 inputs data from the RAM 14 via the bus 38.
- the bus 32 and the bus 33 receive data from the registry file 36.
- the selector 34 selects the output of the data input from the bus 32, the adder 19, and the input register 30, and the selector 35 selects the output of the data input from the bus 33, the adder 20, and the input register 31.
- ALU 29 inputs data from selectors 34 and 35 and performs arithmetic and logic operations on bus 13
- the result of the arithmetic and logic operation is output, and the MSB of the result of the arithmetic and logic operation is output to the comparison result storage unit 7 and the selection unit 8.
- the arithmetic processing device shown in FIG. 19 is suitable for the arithmetic processing of the pipeline structure shown in FIG.
- the selector 34 selects the output of the adder 19 and inputs it to the ALU 29, and the selector 35 selects the output of the adder 20 and inputs it to the ALU 29. Then, the ALU 29 subtracts the input two data, and outputs the subtraction result MSB to the comparison result storage unit 7 and the selection unit 8.
- the selector 34 and the selector 35 select the bus 32 and the bus 33, respectively, and the data output to the bus 32 and the bus 33 from the register evening file 36. To ALU 29.
- the selector 34 and the selector 35 select the bus 32 and the input register 31, respectively, and the data output from the register file 36 to the bus 32,
- the data output from the RA 114 to the input register 31 via the bus 38 is input to the ALU 29.
- the selector 34 and the selector 35 select the input register 30 and the bus 33, respectively, and output the input register 30 from the RAMI 4 via the bus 37.
- the input data and the data output from the register file 36 to the bus 33 are input to the ALU 29.
- the selector 34 and the selector 35 select the input register 30 and the input register 31 respectively, and the input register is transferred from the RAM 14 via the bus 37 and the bus 38. Input the data entered in evening 30 and input register evening 31 to ALU29.
- the chip area can be reduced and the cost can be reduced. It should be noted that the same operation can be performed by using a dual-port RAM as a means for storing the path metric.
- FIG. 20 is a block diagram showing a configuration of the arithmetic processing device according to the eighth embodiment.
- the same portions as those of the arithmetic processing device shown in FIG. 19 are denoted by the same reference numerals, and description thereof will be omitted.
- the arithmetic processing unit shown in FIG. 20 is different from the arithmetic processing unit in FIG. 19 in that two adders 19 and an adder 20 are configured by 4: 2 COMPRES SOR 39, and two adders 24 and an adder are provided. 25 is composed of 4: 2 COMPRES S OR 40.
- the 4: 2 COMPRES SOR 39 and 4: 2C ⁇ MPRES S ⁇ R40 are connected in series by the number of single-precision bits shown in Figure 21 and perform addition processing faster than a normal full adder.
- the 4: 2 COMPRES SOR 39 inputs data from the bus 4 and the input register 16 and outputs the operation result to the selector 34 and the selector 35.
- 4: 2 COMPRES SOR40 inputs data from bus 4 and input register 17 and outputs the operation result to comparator 26.
- the arithmetic processing device shown in FIG. 20 is suitable for the arithmetic processing of the pipeline structure shown in FIG.
- two path metrics are output as ⁇ A, B ⁇ from the input register 16 and the input register 17, and two branch metrics are output as ⁇ C, D ⁇ from the swap circuit 18.
- 2 COMPRES SOR 39 has a path metric ⁇ A ⁇ Launch metric ⁇ C ⁇ , path metric ⁇ B ⁇ inversion ⁇ — B ⁇ , and branch metric ⁇ D ⁇ inversion ⁇ — D ⁇ are input, and ⁇ A + C ⁇ and ⁇ B + D ⁇ are output.
- Two outputs ⁇ A + C ⁇ and ⁇ B + D ⁇ of 2COMPRES SOR39 are input to ALU29 via selector 34 and selector 35 and added. However, in this case, in order to realize the two's complement of ⁇ B ⁇ and ⁇ D ⁇ ,
- the path metric ⁇ A ⁇ and the branch metric ⁇ C ⁇ are input to the adder 22, and the addition result ⁇ A + C ⁇ is output.
- the metric ⁇ B ⁇ and the branch metric ⁇ D ⁇ are input to the adder 23, and the addition result ⁇ B + D ⁇ is output.
- the 4: 2 COMPRES SOR40 receives the path metric ⁇ A ⁇ , the branch metric ⁇ D ⁇ , the inversion ⁇ —B ⁇ of the path metric ⁇ B ⁇ , and the inversion ⁇ —C ⁇ of the branch metric ⁇ C ⁇ .
- ⁇ A + C ⁇ and ⁇ B + D ⁇ are output.
- two outputs ⁇ A + C ⁇ and ⁇ B + D- ⁇ of the 4: 2 COMPRES SOR39 are input and added.
- "1" is input to the 4: 2 COMPRES SOR40 and the least significant carry input of the comparator 26.
- ⁇ A + D— (B + C) ⁇ is obtained, and the MSB is output from the comparator 26.
- the path metric ⁇ A ⁇ and the branch metric ⁇ D ⁇ are input to the adder 27, and the addition result ⁇ A + D ⁇ is output.
- the path metric ⁇ B ⁇ and the branch metric ⁇ C ⁇ are input to the adder 28, and the addition result ⁇ B + C ⁇ is output.
- the two path metrics ⁇ A, B ⁇ of input register 16 and input register 17 are set to ⁇ PM1, PM0 ⁇ , and the swap circuit 1 PP
- the eighth embodiment by applying the 4: 2CO MPRESS OR to the comparison unit that performs the ACS operation, it is possible to realize a higher-speed operation than when two adders are used.
- the constraint length K 4, the coding rate 1/2, and the description using the nodes ⁇ ⁇ 0, Nl, ⁇ 6, and ⁇ 7 in the old state, but the nodes ⁇ 2 and ⁇ 3 , Node # 4 and Node # 5, the above relationship holds, and also holds for other combinations of the constraint length ⁇ and the coding rate.
- the present invention can be similarly implemented by using a dual port RAM as a means for storing a path metric.
- FIG. 22 is a block diagram showing a configuration of the arithmetic processing device according to the ninth embodiment.
- the same portions as those of the arithmetic processing device shown in FIG. 20 are denoted by the same reference numerals, and description thereof will be omitted.
- the arithmetic processing device shown in FIG. 22 uses double-precision adders 41 and 42 as adders, and at least one uses a double-precision AU41 as compared with the arithmetic processing device in FIG.
- double-precision AU41 inputs double-precision data from input register 16 and bus 4 and performs double-precision arithmetic operation. Evening data is input from bus 17 and bus 4 to perform double-precision addition.
- the double precision AU 41 outputs the calculation result to the selection unit 8 and the bus 13, and the output of the double precision adder 42 outputs the calculation result to the selection unit 12.
- the arithmetic processing device shown in FIG. 22 is suitable for the arithmetic processing of the pipeline structure shown in FIG.
- the double-precision AU 41 When performing the ACS operation in the ninth embodiment, the double-precision AU 41 inputs two path metrics from the input register 16 as ⁇ A, B ⁇ in double-precision format, and from the swap circuit 18 via the bus 4 Input the two branch metrics as ⁇ C, D ⁇ in double precision format and add double precision.
- the double-precision AU4 1 forces the carry from the bit position of the single-precision MSB to the next stage to be zero, and adds the two path metrics and the branch metric ⁇ A + C, B + D ⁇ at the same time.
- the double-precision adder 42 inputs two path metrics from the input register 17 as ⁇ A, B ⁇ in double-precision format, and double-metrics the two branch metrics from the swap circuit 18 via the bus 4. Enter ⁇ D, C ⁇ in the format. Then, the double-precision adder 42 forcibly sets the carry from the bit position of the single-precision MSB to the next stage to zero, and adds the two path metrics and the branch metric ⁇ A + D, B + C ⁇ . At the same time.
- double-precision AU 41 is used in the adder that performs ACS operation, and during ACS operation, the carry from the bit position of the single-precision MSB to the next stage is forcibly set to zero.
- double-precision arithmetic operations other than the ACS operation by adding control to propagate the carry, for example, it is possible to double as a double-precision accumulator for multiply-accumulate operations.
- the chip area can be further reduced, and the cost can be reduced. It should be noted that the same can be implemented by using a dual port RAM as a means for storing the path metric. (Embodiment 10)
- FIG. 24 is a block diagram showing a configuration of the arithmetic processing device according to the tenth embodiment.
- the same portions as those of the arithmetic processing device shown in FIG. 22 are denoted by the same reference numerals, and description thereof will be omitted.
- the arithmetic processing device shown in FIG. 24 uses a shift register 43 and a shift register 44 as a comparison result storage unit, as compared with the arithmetic processing device of FIG.
- shift register 43 inputs the MSB of the operation result of ALU 29 and outputs data to bus 2.
- the shift register 44 inputs the MSB of the operation result of the comparator 26 and outputs data to the bus 2.
- the arithmetic processing device shown in FIG. 24 is suitable for the arithmetic processing of the pipeline structure shown in FIG.
- the MSB of the comparison result by ALU 29 is shifted into shift register 43 as needed, and the MSB of the comparison result by comparator 26 is shifted into shift register 44 as needed.
- the path select signal can be stored in RAM 14.
- the path select signal indicates which of the two paths is selected, and is a signal used when performing traceback after the completion of the ACS operation.
- a pass select signal can be stored by performing the ACS operation for the number of single precision bits.
- the shift register as the storage unit for storing the comparison result for performing the ACS operation, for example, it is possible to share the operation instruction using the shift register of the division system. If it is possible to use an arithmetic processing unit as an LSI, the chip area can be further reduced and the cost can be reduced. Note that the same operation can be performed even when a dual-port RAM is used as a means for storing a path metric. (Embodiment 11)
- FIG. 25 is a block diagram showing a configuration of the arithmetic processing unit according to Embodiment 11.
- the same portions as those in the arithmetic processing device shown in FIG. 24 are denoted by the same reference numerals, and description thereof will be omitted.
- the arithmetic processing unit shown in FIG. 25 is different from the arithmetic processing unit in FIG. 24 in that the input register 17 always swaps and inputs path metric data from the bus 2 and the 4: 2 COMPRES SOR 40 is connected to the swap circuit 18 from the swap circuit 18.
- the output branch metric data is input without being swapped, and the negate value of the comparison result of the comparator 26 is shifted into the shift register 44.
- the arithmetic processing device shown in FIG. 25 is suitable for the arithmetic processing of the pipeline structure shown in FIG.
- two path metrics ⁇ A, B ⁇ are input to input register 16 directly as ⁇ A, B ⁇ , and are constantly swapped to input register 17 ⁇ B, A ⁇ Is entered as Then, in the 4: 2 COMPRES SOR40, the two branch metrics from the swap circuit 18 are ⁇ C ⁇ and ⁇ —D ⁇ , and the two path metrics from the input register 17 are ⁇ B ⁇ and ⁇ —A ⁇ . Input and ⁇ A + D ⁇ and ⁇ B + C ⁇ are output. Then, the comparator 26 receives two outputs ⁇ A + D ⁇ and ⁇ B + C ⁇ of the 4: 2 COMPRES SOR 40 and calculates ⁇ A + D—B—C ⁇ .
- the double-precision adder 42 receives two branch metrics from the swap circuit 18 as ⁇ C, D ⁇ , and two path metrics from the input register as ⁇ B, A ⁇ .
- ⁇ And ⁇ A + D ⁇ are processed in parallel at the same time, and output to the selection unit 12 in the form of ⁇ B + C, A + D ⁇ .
- the comparator 26 outputs the MSB of the comparison result to the selector 12, and outputs the MSB of the negation value of the comparison result to the shift register 44.
- an input for storing two path metrics is provided.
- By swapping and inputting one of the two digits it is possible to eliminate the need for swapping between the 4: 2 COMPRESSOR 40 and the input of the double-precision adder 42 in the execution (EX) stage, resulting in higher speed.
- ACS operation can be performed. Note that the same can be implemented by using a dual-port RAM as a means for storing a path metric.
- FIG. 26 is a block diagram showing a configuration of a mobile station device according to Embodiment 12.
- a mobile station device 45 shown in FIG. 26 includes an antenna unit 46 for transmission and reception, a radio unit 47 including a reception unit 48 and a transmission unit 49, signal modulation and demodulation, encoding and decoding.
- a display unit 61 for displaying, an operation unit 62 such as a numeric keypad, and a control unit 63 for controlling each unit are provided.
- the baseband signal processing section 50 includes a demodulation section 51 for demodulating a reception signal, a modulation section 52 for modulating a transmission signal, and a one-chip DSP 53.
- the DSP 53 includes a Viterbi decoding unit 55 composed of the arithmetic processing device according to any one of Embodiments 1 to 11, a convolutional encoding unit 56 for convolutionally encoding a transmission signal, and a code of an audio signal.
- Speech codec unit 57 for decoding, timing for transmitting the received signal from demodulation unit 51 to Viterbi decoding unit 55, and timing for controlling the timing for transmitting the transmission signal from convolutional coding unit 56 to modulation unit 52
- the control units 54 and are respectively formed by software.
- the control unit 63 for example, displays a signal input from the operation unit 62 on the display unit 61, receives a signal input from the operation unit 62, and performs an outgoing / incoming call operation according to a communication sequence.
- the control signal is output to the antenna unit 46, the radio unit 47, the baseband signal processing unit 50, and the like.
- the audio signal When audio is transmitted from the mobile station device 45, the audio signal input from the microphone 59 The signal is AD-converted by an AD converter (not shown), encoded by a voice codec 57 of the DSP 53, and input to a convolutional encoder 56. When data is transmitted, externally input data is input to the convolutional encoder 56 via the data input / output unit 60.
- the data input to the convolutional coding unit 56 is convolutionally coded, the data is rearranged and the transmission output timing is adjusted by the evening imaging control unit 54, and is output to the modulation unit 52.
- the data input to the modulation section 52 is digitally modulated, D / A converted, and output to the transmission section 49 of the radio section 47.
- the data input to the transmitting section 49 is converted into a radio signal and transmitted from the antenna section 46 as a radio wave.
- the radio wave received by the antenna unit 46 is received by the receiving unit 48 of the radio unit 47, AD-converted, and subjected to baseband signal processing.
- the signal is output to the demodulation unit 51 of the unit 50.
- the data demodulated by the demodulation unit 51 is rearranged by the timing control unit 54, and then decoded by the video decoding unit 55.
- the decoded data is voice-decoded by the voice codec unit 57, D / A converted, and output from the speaker 58 as voice.
- the decoded data is output to the outside via the data input / output unit 60, and the data decoded by the Viterbi decoding unit 55.
- the mobile station apparatus 45 includes a Viterbi decoding unit 55, a convolutional coding unit 56, an audio codec unit 57, and a timing control unit 54, each of which is a single-chip DSP 5. Since it is formed with the software of No. 3, it can be assembled with a small number of parts. Further, since this Viterbi decoding unit 55 is formed by any one of the arithmetic processing units according to the first to eleventh embodiments, the pipeline processing by the DSP 53 causes two path metrics per machine cycle. Rick update can be realized, which enables high-speed Viterbi decoding by DSP 53 with a relatively small amount of processing ACS operation can be realized.
- the demodulation unit 51 and the modulation unit 52 are shown separately from the DSP 53, but they may be configured by DSP 53 software. Further, it is also possible to use the DSP of the sixth embodiment as the DSP, and to configure the convolution coding unit 56, the audio codec unit 57 and the timing control unit 54 with different components.
- FIG. 27 is a block diagram showing a configuration of the mobile station apparatus according to Embodiment 13.
- the mobile station device 45A shown in FIG. 27 the same parts as those of the mobile station device 45 shown in FIG.
- the mobile station apparatus 45 A shown in FIG. 27 has a spreading section 65 in the modulation section 52 A and a despread section in the demodulation section 51 A, as compared with the mobile station apparatus 45 in FIG. 6 4 is provided as the baseband signal processing unit 5 OA for the CDMA communication system.
- the timing control unit 54 may include a RAKE receiving unit that combines a plurality of fingers selected from a delay pull file or the like.
- mobile station apparatus 45 A in Embodiment 13 provides CDMA communication by providing despreading section 64 in demodulating section 51 A and spreading section 65 in modulating section 52 A. Can be applied.
- FIG. 28 is a block diagram showing a configuration of the base station apparatus according to Embodiment 14.
- base station apparatus 68 in Embodiment 14 includes an antenna section 46 including a receiving antenna 66 and a transmitting antenna 67, a receiving section 48 and a transmitting section 49. , A baseband signal processing unit 69 that modulates and demodulates signals and encodes and decodes signals, and a data input / output unit that inputs and outputs data to be transmitted and received to and from a wired line 60 and a control unit 63 for controlling each unit.
- the baseband signal processing unit 69 includes a demodulation unit 51 for demodulating a received signal, a modulation unit 52 for modulating a transmission signal, and a single-chip DSP 53A.
- a Viterbi decoding unit 55 composed of the arithmetic processing device according to any one of Embodiments 1 to 11, a convolutional encoding unit 56 for convolutionally encoding a transmission signal, and a demodulation unit 51 for demodulating a reception signal
- a timing control unit 54 for controlling the timing for transmitting the transmission signal from the convolutional coding unit 56 to the modulation unit 52 from the convolution encoding unit 56 to the Viterbi decoding unit 55, respectively, by software.
- the base station device 68 When the base station device 68 receives data from a wired line, the data is input to a convolutional coding unit 56 via a data input / output unit 60, convolutionally coded, and a timing control unit 5 Rearrangement of input data and adjustment of transmission output timing input at 4 are performed and output to modulation section 52.
- the data input to the modulation section 52 is digitally modulated, D / A converted, converted to a radio signal by the transmission section 49, and transmitted as a radio wave from the antenna section 46.
- the base station device 68 receives data from a wireless channel
- the radio wave received by the antenna unit 46 is AD-converted by the receiving unit 48 and the baseband signal processing unit 69
- the demodulation unit 51 demodulates the signal.
- the demodulated data is subjected to data rearrangement and the like at the evening control unit 54, decoded at the video decoding unit 55, and output to the wired line via the data input / output unit 60.
- the base station device 68 in the embodiment 13 includes the Viterbi decoding unit 55, the convolutional coding unit 56, and the timing control unit 54 formed by the software of the DSP 53A of one chip. As a result, it can be assembled with a small number of parts. Also, since this Viterbi decoding unit 55 forms the arithmetic processing unit according to any one of the first to eleventh embodiments, two path metrics can be calculated in one machine cycle by pipeline processing by the DSP 53A. The update can be realized, and the ACS operation of Viterbi decoding by DSP 53 A can be performed at high speed with a relatively small amount of processing. realizable.
- the demodulation unit 51 and the modulation unit 52 are shown separately from the DSP 53A, but they may be configured by software of the DSP 53A. Further, it is also possible to use the DSP of the sixth embodiment as DSP 53A, and to configure convolution coding section 56, audio codec section 57, and evening control section 54 as separate components.
- FIG. 29 is a block diagram showing a configuration of a base station mobile station apparatus according to Embodiment 15. 29.
- the base station device 68A shown in FIG. 29 the same parts as those of the mobile station device 45 shown in FIG.
- the mobile station device 45A shown in FIG. 27 is different from the mobile station device 45 shown in FIG. 26 in that the modulation section 52A is provided with a spreading section 65, and the demodulation section 51A is provided with a despreading section 64.
- the timing control unit 54 may include a RAKE receiving unit that combines a plurality of fingers selected from a delay pull file or the like.
- base station apparatus 68A in Embodiment 15 can be applied to CDMA communication by providing despreading section 64 in demodulation section 51A and spreading section 65 in modulation section 52A. it can.
- ACS calculation of Viterbi decoding by DSP can be realized at high speed with relatively small processing amount, Smaller size, lighter weight, lower price, and longer battery life.
Landscapes
- Physics & Mathematics (AREA)
- Probability & Statistics with Applications (AREA)
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Error Detection And Correction (AREA)
- Detection And Correction Of Errors (AREA)
- Detection And Prevention Of Errors In Transmission (AREA)
Description
Claims
Priority Applications (6)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
US09/147,663 US6330684B1 (en) | 1997-06-30 | 1998-06-29 | Processor and processing method |
DE69827915T DE69827915T2 (de) | 1997-06-30 | 1998-06-29 | Verarbeitungsverfahren und -vorrichtung |
EP98929741A EP0923197B1 (en) | 1997-06-30 | 1998-06-29 | Processor and processing method |
US10/748,242 US7139968B2 (en) | 1997-06-30 | 2003-12-31 | Processing unit and processing method |
US11/022,811 US7325184B2 (en) | 1997-06-30 | 2004-12-28 | Communications digital signal processor and digital signal processing method |
US11/929,126 US20080072129A1 (en) | 1997-06-30 | 2007-10-30 | Communications digital signal processor and digital signal processing method |
Applications Claiming Priority (4)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
JP9/173878 | 1997-06-30 | ||
JP17387897 | 1997-06-30 | ||
JP10/168567 | 1998-06-16 | ||
JP16856798A JP3338374B2 (ja) | 1997-06-30 | 1998-06-16 | 演算処理方法および装置 |
Related Child Applications (3)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
US09/147,663 A-371-Of-International US6330684B1 (en) | 1997-06-30 | 1998-06-29 | Processor and processing method |
US09147663 A-371-Of-International | 1998-06-29 | ||
US09/974,807 Division US6477661B2 (en) | 1997-06-30 | 2001-10-12 | Processing unit and processing method |
Publications (1)
Publication Number | Publication Date |
---|---|
WO1999000905A1 true WO1999000905A1 (fr) | 1999-01-07 |
Family
ID=26492229
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
PCT/JP1998/002909 WO1999000905A1 (fr) | 1997-06-30 | 1998-06-29 | Processeur et procede de traitement |
Country Status (6)
Country | Link |
---|---|
US (6) | US6330684B1 (ja) |
EP (1) | EP0923197B1 (ja) |
JP (1) | JP3338374B2 (ja) |
CN (3) | CN1118140C (ja) |
DE (1) | DE69827915T2 (ja) |
WO (1) | WO1999000905A1 (ja) |
Families Citing this family (39)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
JP3338374B2 (ja) * | 1997-06-30 | 2002-10-28 | 松下電器産業株式会社 | 演算処理方法および装置 |
US6473088B1 (en) * | 1998-06-16 | 2002-10-29 | Canon Kabushiki Kaisha | System for displaying multiple images and display method therefor |
US6647071B2 (en) * | 1998-11-06 | 2003-11-11 | Texas Instruments Incorporated | Method and apparatus for equalization and tracking of coded digital communications signals |
JP3683128B2 (ja) * | 1999-07-06 | 2005-08-17 | 日本電気株式会社 | 無線通信機および無線通信機の消費電力制御方法 |
WO2001069411A2 (en) | 2000-03-10 | 2001-09-20 | Arc International Plc | Memory interface and method of interfacing between functional entities |
US6760385B1 (en) * | 2000-05-30 | 2004-07-06 | Adtran, Inc. | Universal parallel processing decoder |
JP2001352254A (ja) * | 2000-06-08 | 2001-12-21 | Sony Corp | 復号装置及び復号方法 |
US7234100B1 (en) * | 2000-09-28 | 2007-06-19 | Intel Corporation | Decoder for trellis-based channel encoding |
US6904105B1 (en) * | 2000-10-27 | 2005-06-07 | Intel Corporation | Method and implemention of a traceback-free parallel viterbi decoder |
JP3984790B2 (ja) * | 2001-01-15 | 2007-10-03 | 日本電気株式会社 | ビタビ復号処理装置 |
US6848074B2 (en) * | 2001-06-21 | 2005-01-25 | Arc International | Method and apparatus for implementing a single cycle operation in a data processing system |
US7020830B2 (en) * | 2001-12-24 | 2006-03-28 | Agere Systems Inc. | High speed add-compare-select operations for use in viterbi decoders |
US7127667B2 (en) * | 2002-04-15 | 2006-10-24 | Mediatek Inc. | ACS circuit and viterbi decoder with the circuit |
US7089481B2 (en) * | 2002-07-22 | 2006-08-08 | Agere Systems Inc. | High speed arithmetic operations for use in turbo decoders |
US7359464B2 (en) * | 2003-12-31 | 2008-04-15 | Intel Corporation | Trellis decoder and method of decoding |
US20050157823A1 (en) * | 2004-01-20 | 2005-07-21 | Raghavan Sudhakar | Technique for improving viterbi decoder performance |
JP4230953B2 (ja) * | 2004-03-31 | 2009-02-25 | 株式会社ケンウッド | ベースバンド信号生成装置、ベースバンド信号生成方法及びプログラム |
US7779339B2 (en) * | 2004-05-14 | 2010-08-17 | Panasonic Corporation | ACS circuit |
CN101160729B (zh) * | 2004-06-23 | 2011-09-14 | Nxp股份有限公司 | 用于并行处理递归数据的定址体系结构 |
US7231586B2 (en) * | 2004-07-21 | 2007-06-12 | Freescale Semiconductor, Inc. | Multi-rate viterbi decoder |
TWI255622B (en) * | 2004-10-21 | 2006-05-21 | Mediatek Inc | Method of computing path metrics in a high-speed Viterbi detector and related apparatus thereof |
US7607072B2 (en) * | 2005-01-28 | 2009-10-20 | Agere Systems Inc. | Method and apparatus for-soft-output viterbi detection using a multiple-step trellis |
EP1746756B1 (en) * | 2005-07-21 | 2013-01-16 | STMicroelectronics Srl | A method and system for decoding signals, corresponding receiver and computer program product |
US8025655B2 (en) | 2005-09-12 | 2011-09-27 | Bridgepoint Medical, Inc. | Endovascular devices and methods |
KR100729619B1 (ko) * | 2005-11-07 | 2007-06-19 | 삼성전자주식회사 | 고속 데이터 전송을 위한 비터비 디코딩 방법 및 장치 |
CN101346695A (zh) * | 2005-12-27 | 2009-01-14 | 松下电器产业株式会社 | 具有可重构的运算器的处理器 |
CN101361039A (zh) * | 2006-01-20 | 2009-02-04 | 松下电器产业株式会社 | 处理器 |
US20070239182A1 (en) * | 2006-04-03 | 2007-10-11 | Boston Scientific Scimed, Inc. | Thrombus removal device |
JP4303256B2 (ja) * | 2006-04-04 | 2009-07-29 | パナソニック株式会社 | デジタル信号受信装置およびデジタル信号受信方法 |
US20080005842A1 (en) * | 2006-07-07 | 2008-01-10 | Ferno-Washington, Inc. | Locking mechanism for a roll-in cot |
US20080270658A1 (en) * | 2007-04-27 | 2008-10-30 | Matsushita Electric Industrial Co., Ltd. | Processor system, bus controlling method, and semiconductor device |
US8140949B2 (en) * | 2007-11-13 | 2012-03-20 | Alpha Imaging Technology Corp. | ACS unit of a Viterbi decoder and method for calculating a bit error rate before a Viterbi decoder |
CA2621147C (en) * | 2008-02-15 | 2013-10-08 | Connotech Experts-Conseils Inc. | Method of bootstrapping an authenticated data session configuration |
KR101410697B1 (ko) | 2008-03-20 | 2014-07-02 | 삼성전자주식회사 | 수신기 및 상기 수신기의 동작 제어 방법 |
US8943392B2 (en) * | 2012-11-06 | 2015-01-27 | Texas Instruments Incorporated | Viterbi butterfly operations |
US20150170067A1 (en) * | 2013-12-17 | 2015-06-18 | International Business Machines Corporation | Determining analysis recommendations based on data analysis context |
KR102319779B1 (ko) * | 2014-02-25 | 2021-11-01 | 도레이 카부시키가이샤 | 해도 복합 섬유, 복합 극세 섬유 및 섬유 제품 |
DE102015107509A1 (de) * | 2014-10-14 | 2016-04-14 | Infineon Technologies Ag | Chip und Verfahren zum Betreiben einer Verarbeitungsschaltung |
JP7007115B2 (ja) * | 2017-06-01 | 2022-01-24 | Necプラットフォームズ株式会社 | ビタビ復号装置、及び、ビタビ復号方法 |
Citations (3)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
JPS63215227A (ja) * | 1987-03-04 | 1988-09-07 | Toshiba Corp | ビタビ復号回路 |
US5633897A (en) * | 1995-11-16 | 1997-05-27 | Atmel Corporation | Digital signal processor optimized for decoding a signal encoded in accordance with a Viterbi algorithm |
JPH09148943A (ja) * | 1995-11-17 | 1997-06-06 | Nec Corp | ビタビ復号装置 |
Family Cites Families (40)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
JPS6081925A (ja) * | 1983-10-12 | 1985-05-10 | Nec Corp | 誤り訂正装置 |
JPS60173930A (ja) * | 1984-02-20 | 1985-09-07 | Fujitsu Ltd | パイプライン処理ビタビ復号器 |
JPS6468853A (en) | 1987-09-10 | 1989-03-14 | Fujitsu Ltd | Memory control system |
JPH01102541A (ja) * | 1987-10-16 | 1989-04-20 | Toshiba Corp | 放射線画像装置 |
JPH03128552A (ja) * | 1989-07-31 | 1991-05-31 | Nec Corp | データ伝送システムのための変復調装置 |
US5027374A (en) * | 1990-03-26 | 1991-06-25 | Motorola, Inc. | Bit serial Viterbi decoder add/compare/select array |
US5375129A (en) * | 1990-07-19 | 1994-12-20 | Technophone Limited | Maximum likelihood sequence detector |
US5220570A (en) * | 1990-11-30 | 1993-06-15 | The Board Of Trustees Of The Leland Stanford Junior University | Programmable viterbi signal processor |
US5502735A (en) * | 1991-07-16 | 1996-03-26 | Nokia Mobile Phones (U.K.) Limited | Maximum likelihood sequence detector |
JP2702831B2 (ja) * | 1991-08-28 | 1998-01-26 | 松下電送株式会社 | ヴィタビ復号法 |
US5327440A (en) * | 1991-10-15 | 1994-07-05 | International Business Machines Corporation | Viterbi trellis coding methods and apparatus for a direct access storage device |
US5280489A (en) * | 1992-04-15 | 1994-01-18 | International Business Machines Corporation | Time-varying Viterbi detector for control of error event length |
JPH05335972A (ja) * | 1992-05-27 | 1993-12-17 | Nec Corp | ビタビ復号器 |
JPH0722969A (ja) | 1993-06-16 | 1995-01-24 | Matsushita Electric Ind Co Ltd | 演算装置 |
AU652896B2 (en) | 1992-09-29 | 1994-09-08 | Matsushita Electric Industrial Co., Ltd. | Arithmetic apparatus |
JPH06164423A (ja) | 1992-11-20 | 1994-06-10 | Murata Mach Ltd | ビタビ復号器 |
EP0644661B1 (en) * | 1993-09-20 | 2000-06-14 | Canon Kabushiki Kaisha | Signal processing apparatus |
ZA947317B (en) * | 1993-09-24 | 1995-05-10 | Qualcomm Inc | Multirate serial viterbi decoder for code division multiple access system applications |
US5414738A (en) * | 1993-11-09 | 1995-05-09 | Motorola, Inc. | Maximum likelihood paths comparison decoder |
US5550870A (en) * | 1994-03-02 | 1996-08-27 | Lucent Technologies Inc. | Viterbi processor |
JP3241210B2 (ja) * | 1994-06-23 | 2001-12-25 | 沖電気工業株式会社 | ビタビ復号方法及びビタビ復号回路 |
FR2724273B1 (fr) * | 1994-09-05 | 1997-01-03 | Sgs Thomson Microelectronics | Circuit de traitement de signal pour mettre en oeuvre un algorithme de viterbi |
JP3304631B2 (ja) | 1994-09-14 | 2002-07-22 | ソニー株式会社 | ビタビ復号方法及びビタビ復号装置 |
JPH0946240A (ja) | 1995-07-27 | 1997-02-14 | Mitsubishi Electric Corp | ビタビ復号機能を有するデータ処理装置 |
KR0144505B1 (ko) * | 1995-09-18 | 1998-08-17 | 구자홍 | 영상표시기기의 화면 자동 조정장치 및 방법 |
US5742621A (en) * | 1995-11-02 | 1998-04-21 | Motorola Inc. | Method for implementing an add-compare-select butterfly operation in a data processing system and instruction therefor |
JPH09232973A (ja) * | 1996-02-28 | 1997-09-05 | Sony Corp | ビタビ復号器 |
KR100195745B1 (ko) * | 1996-08-23 | 1999-06-15 | 전주범 | 비터비 복호화기의 가산 비교 선택 장치 |
JPH10107651A (ja) | 1996-09-27 | 1998-04-24 | Nec Corp | ビタビ復号装置 |
EP0851591B1 (en) | 1996-12-24 | 2001-09-12 | Matsushita Electric Industrial Co., Ltd. | Data processor and data processing method |
JPH10209882A (ja) | 1997-01-24 | 1998-08-07 | Japan Radio Co Ltd | ビタビ復号方法 |
JP3242059B2 (ja) | 1997-06-04 | 2001-12-25 | 沖電気工業株式会社 | ビタビ復号器 |
JP3338374B2 (ja) * | 1997-06-30 | 2002-10-28 | 松下電器産業株式会社 | 演算処理方法および装置 |
US6257756B1 (en) * | 1997-07-16 | 2001-07-10 | Motorola, Inc. | Apparatus and method for implementing viterbi butterflies |
JPH11112361A (ja) | 1997-10-01 | 1999-04-23 | Matsushita Electric Ind Co Ltd | データ復号装置及びデータ復号方法 |
US5987490A (en) * | 1997-11-14 | 1999-11-16 | Lucent Technologies Inc. | Mac processor with efficient Viterbi ACS operation and automatic traceback store |
US5912908A (en) * | 1997-11-21 | 1999-06-15 | Lucent Technologies Inc. | Method of efficient branch metric computation for a Viterbi convolutional decoder |
US6195782B1 (en) * | 1998-05-28 | 2001-02-27 | Advanced Micro Devices, Inc. | MLSE implementation using a general purpose DSP and shared hardware for a GSM application |
JP3403382B2 (ja) * | 1999-10-29 | 2003-05-06 | 松下電器産業株式会社 | トレリス復号装置及びその方法 |
US20080077451A1 (en) * | 2006-09-22 | 2008-03-27 | Hartford Fire Insurance Company | System for synergistic data processing |
-
1998
- 1998-06-16 JP JP16856798A patent/JP3338374B2/ja not_active Expired - Fee Related
- 1998-06-29 DE DE69827915T patent/DE69827915T2/de not_active Expired - Lifetime
- 1998-06-29 US US09/147,663 patent/US6330684B1/en not_active Expired - Lifetime
- 1998-06-29 EP EP98929741A patent/EP0923197B1/en not_active Expired - Lifetime
- 1998-06-29 WO PCT/JP1998/002909 patent/WO1999000905A1/ja active IP Right Grant
- 1998-06-29 CN CN98800905A patent/CN1118140C/zh not_active Expired - Fee Related
- 1998-06-29 CN CNB031451616A patent/CN100512075C/zh not_active Expired - Fee Related
- 1998-06-29 CN CN2006101003464A patent/CN101018103B/zh not_active Expired - Fee Related
-
2001
- 2001-10-12 US US09/974,807 patent/US6477661B2/en not_active Expired - Lifetime
-
2002
- 2002-09-24 US US10/252,394 patent/US6735714B2/en not_active Expired - Lifetime
-
2003
- 2003-12-31 US US10/748,242 patent/US7139968B2/en not_active Expired - Fee Related
-
2004
- 2004-12-28 US US11/022,811 patent/US7325184B2/en not_active Expired - Fee Related
-
2007
- 2007-10-30 US US11/929,126 patent/US20080072129A1/en not_active Abandoned
Patent Citations (3)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
JPS63215227A (ja) * | 1987-03-04 | 1988-09-07 | Toshiba Corp | ビタビ復号回路 |
US5633897A (en) * | 1995-11-16 | 1997-05-27 | Atmel Corporation | Digital signal processor optimized for decoding a signal encoded in accordance with a Viterbi algorithm |
JPH09148943A (ja) * | 1995-11-17 | 1997-06-06 | Nec Corp | ビタビ復号装置 |
Non-Patent Citations (3)
Title |
---|
LOU H: "VITERBI DECODER DESIGN FOR THE IS-95 CDMA FORWARD LINK", 1996 IEEE 46TH. VEHICULAR TECHNOLOGY CONFERENCE. MOBILE TECHNOLOGY FOR THE HUMAN RACE. ATLANTA, APR. 28 - MAY 1, 1996., NEW YORK, IEEE., US, vol. 02, no. 02, 1 January 1996 (1996-01-01), US, pages 1346 - 1350, XP002913962, ISBN: 978-0-7803-3158-7 * |
See also references of EP0923197A4 * |
UEDA K, ET AL.: "A 16-BIT DIGITAL SIGNAL PROCESSOR WITH SPECIALLY ARRANGED MULTIPLY-ACCUMULATOR FOR LOW POWER CONSUMPTION", IEICE TRANSACTIONS ON ELECTRONICS, INSTITUTE OF ELECTRONICS, TOKYO, JP., vol. E78-C, no. 12, 1 December 1995 (1995-12-01), Tokyo, JP., pages 1709 - 1716, XP002913961, ISSN: 0916-8524 * |
Also Published As
Publication number | Publication date |
---|---|
US6330684B1 (en) | 2001-12-11 |
EP0923197B1 (en) | 2004-12-01 |
CN101018103A (zh) | 2007-08-15 |
US7325184B2 (en) | 2008-01-29 |
JP3338374B2 (ja) | 2002-10-28 |
US6735714B2 (en) | 2004-05-11 |
US20030066022A1 (en) | 2003-04-03 |
CN1516381A (zh) | 2004-07-28 |
US20050163233A1 (en) | 2005-07-28 |
DE69827915T2 (de) | 2005-04-07 |
CN100512075C (zh) | 2009-07-08 |
CN1118140C (zh) | 2003-08-13 |
US6477661B2 (en) | 2002-11-05 |
DE69827915D1 (de) | 2005-01-05 |
US20020016946A1 (en) | 2002-02-07 |
US7139968B2 (en) | 2006-11-21 |
US20080072129A1 (en) | 2008-03-20 |
CN1231083A (zh) | 1999-10-06 |
EP0923197A1 (en) | 1999-06-16 |
CN101018103B (zh) | 2013-02-27 |
EP0923197A4 (en) | 2000-10-04 |
US20040177313A1 (en) | 2004-09-09 |
JPH1174801A (ja) | 1999-03-16 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
WO1999000905A1 (fr) | Processeur et procede de traitement | |
JP3241210B2 (ja) | ビタビ復号方法及びビタビ復号回路 | |
KR100439211B1 (ko) | 연산처리장치 | |
WO2005011129A1 (ja) | ビタビ復号器 | |
EP1111798B1 (en) | Digital signal processor with co-processor for Viterbi decoding | |
US20060143554A1 (en) | Scalable traceback technique for channel decoder | |
JP2798123B2 (ja) | ビタビ復号装置 | |
US7661059B2 (en) | High performance turbo and Viterbi channel decoding in digital signal processors | |
JP3996858B2 (ja) | 演算処理装置 | |
JP3634333B2 (ja) | ディジタル信号処理プロセッサ | |
JP3383661B2 (ja) | 演算処理装置 | |
US6125153A (en) | Data processor and data processing method | |
US7496827B2 (en) | System and method for multi-mode multi-state path metric addressing | |
JP3253906B2 (ja) | データ処理装置及びデータ処理方法 | |
JP2002217747A (ja) | ビタビ復号処理装置 | |
JPH0730438A (ja) | ビタビ復号方法 | |
SC140 | How to Implement a Viterbi Decoder on the | |
JPH10178356A (ja) | 演算処理装置及びそれを用いた無線局装置 | |
JP2002185337A (ja) | 基地局装置 | |
KR20030082320A (ko) | 비터비 디코더의 브렌치 매트릭 연산장치 | |
WO2007000708A1 (en) | Viterbi decoder and decoding method thereof |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
WWE | Wipo information: entry into national phase |
Ref document number: 98800905.6 Country of ref document: CN |
|
AK | Designated states |
Kind code of ref document: A1 Designated state(s): CN US |
|
AL | Designated countries for regional patents |
Kind code of ref document: A1 Designated state(s): AT BE CH CY DE DK ES FI FR GB GR IE IT LU MC NL PT SE |
|
WWE | Wipo information: entry into national phase |
Ref document number: 09147663 Country of ref document: US |
|
WWE | Wipo information: entry into national phase |
Ref document number: 1998929741 Country of ref document: EP |
|
121 | Ep: the epo has been informed by wipo that ep was designated in this application | ||
WWP | Wipo information: published in national office |
Ref document number: 1998929741 Country of ref document: EP |
|
WWG | Wipo information: grant in national office |
Ref document number: 1998929741 Country of ref document: EP |