JP7233626B1 - Wireless communication system, wireless communication device, control circuit, storage medium, wireless communication method, and transmitter - Google Patents

Abstract

A wireless communication system (1) includes a pattern generation unit that generates a bit erasure pattern, an encoding unit that convolutionally encodes information bits to generate a first bit sequence, and a first bit sequence based on the bit erasure pattern. a transmitting device (100) comprising: a bit erasing unit for erasing bits and generating a second bit sequence; a modulation unit for generating symbols from a bit sequence obtained by dividing the second bit sequence; and a plurality of first reliabilities. a demodulation unit for calculating a plurality of combined bit strings, a likelihood combining unit for generating a plurality of second reliabilities by assigning a value obtained by combining a plurality of first reliabilities to a plurality of combined bit strings; and generating a plurality of third confidences; and creating a trellis diagram and assigning the plurality of third confidences to a plurality of branches of the trellis diagram. and a receiving device (200) comprising a decoding unit that decodes with.

Description

The present disclosure relates to a wireless communication system, wireless communication device, control circuit, storage medium, wireless communication method, and transmitting device that decode using a trellis diagram.

2. Description of the Related Art In recent years, devices remotely installed from a terminal are controlled, or data is collected from a terminal remotely installed. For these purposes, sensor networks using wireless communication or M2M (Machine-To-Machine) communication using wireless communication are becoming popular. In a sensor network or M2M communication, it is desirable to be able to communicate over a long distance between wireless communication terminals from the viewpoint of cost and flexibility of network construction.

In order to realize long-distance wireless communication, it is necessary to use a communication system with good sensitivity performance that can receive weak radio waves that have been attenuated by long-distance propagation. An orthogonal FSK (Frequency Shift Keying) system is given as a communication system with good sensitivity performance. The orthogonal FSK method is a method of modulating a signal by allocating data to mutually orthogonal carrier waves having different frequencies. In addition, in order to realize long-distance wireless communication, even if an error occurs in the data through a long-distance propagation path, the data is encoded with error correction code so that the erroneous data can be corrected. preferably sent. The error-correction-encoded data is decoded by a decoding device provided in the receiving device.

Patent Literature 1 discloses a technique for a decoding device that decodes data convolutionally coded as error correction coding and modulated by an M-ary FSK system in which M is the number of modulation levels. Conventionally, there are constraints on the relationship between the modulation multilevel number M and the coding rate R of the convolutional code. can be flexibly set.

Japanese Patent No. 6628945

However, according to the conventional technique described above, there is a possibility that the reliability assigned to the trellis diagram used in the decoding device will be the same or a similar value. In this case, there is a problem that it is difficult to discriminate between paths, leading to decoding errors.

The present disclosure has been made in view of the above, and an object of the present disclosure is to obtain a radio communication system capable of obtaining good decoding characteristics and flexibly setting a transmission rate.

In order to solve the above-described problems and achieve the object, the present disclosure is a wireless communication system including a transmitting device and a receiving device. The transmission device includes a pattern generator that generates a bit erasure pattern, an encoder that generates a first bit string by convolutionally encoding information bits based on an encoding rate, and a bit erasure pattern based on the A bit erasing unit for generating a second bit string based on the second number of bits by erasing one or more bits from the first bit string for each first number of bits; and a bit string obtained by dividing the second bit string. a modulator that generates symbols by modulating with . The receiving device includes a demodulator that calculates a plurality of first reliabilities that are the reliabilities of a plurality of bit strings that a symbol can take, and generates a plurality of composite bit strings that are bit strings composed of a second number of bits. a likelihood synthesis unit that assigns a value obtained by synthesizing a plurality of first reliabilities to the reliabilities of the plurality of synthesized bit strings to generate a plurality of second reliabilities that are the reliabilities of the plurality of synthesized bit strings; generating a plurality of extension bitstreams, which are bitstreams composed of bits of the number of bits of 1, assigning a plurality of duplicated second reliabilities to the reliabilities of the plurality of extension bitstreams, and assigning a plurality of reliabilities of the extension bitstreams; creating a trellis diagram using a likelihood replication unit that generates a plurality of third reliabilities, a coding rate and a plurality of extended bit sequences, and generating a plurality of third reliabilities in a plurality of branches of the trellis diagram; and a decoding unit that decodes by assigning to . The pattern generation unit performs bit erasure such that the number of branches to which reliability values equal to or less than a specified difference are assigned is less than or equal to a specified number between paths that compare probabilities in the trellis diagram used in the decoding unit. It is characterized by generating patterns .

The radio communication system according to the present disclosure has the effect of obtaining good decoding characteristics and being able to flexibly set the transmission rate.

A diagram showing a configuration example of a radio communication system according to Embodiment 1 A diagram showing a configuration example of a transmitting apparatus according to Embodiment 1 FIG. 4 is a diagram showing, in the frequency domain, signals with different frequencies that are associated with bit strings generated by dividing the second bit string by the modulating unit according to Embodiment 1; A diagram showing a configuration example of a receiving apparatus according to Embodiment 1 Flowchart showing a series of processes of the wireless communication system according to Embodiment 1 Diagram showing a general trellis line with a constraint length of 5 and a coding rate of R=1/2 FIG. 4 shows trellis lines used by the decoding unit according to Embodiment 1; FIG. 4 is a diagram showing a process of synthesizing two of four types of first reliability to generate 16 types of second reliability in a likelihood combining unit according to Embodiment 1; The likelihood synthesizing unit according to Embodiment 1 synthesizes a plurality of first reliabilitys to generate a second reliability, and the likelihood replication unit replicates the second reliability to generate a third reliability. Diagram showing trellis lines when applying a process that produces A diagram showing four shortest merging paths A to D for the trellis diagram shown in FIG. A diagram showing 22 paths that start from state number 0000 and merge again at state number 0000 through three state transitions with respect to the trellis diagram of FIG. Diagram showing the shortest 8 merging paths among the trellis lines when the constraint length is 7, the coding rate is R=1/2, and 6 coding bits are assigned to each branch. FIG. 4 is a diagram showing a configuration example of a processing circuit provided in the radio communication system according to Embodiment 1 when the processing circuit is realized by a processor and a memory; FIG. 4 is a diagram showing an example of a processing circuit provided in the wireless communication system according to Embodiment 1 when the processing circuit is configured by dedicated hardware; A diagram showing a configuration example of a radio communication system according to Embodiment 2 A likelihood combining unit according to Embodiment 3 combines a plurality of first reliabilitys to generate a second reliability, and a likelihood replication unit replicates the second reliability to generate a third reliability. In the trellis diagram when the process for generating is applied, bit erasure pattern [10111101] and bit erasure pattern [01111110 ] is applied to the information bit stream and the encoded bit stream

A radio communication system, a radio communication device, a control circuit, a storage medium, a radio communication method, and a transmission device according to embodiments of the present disclosure will be described below in detail with reference to the drawings.

Embodiment 1.
FIG. 1 is a diagram showing a configuration example of a radio communication system 1 according to Embodiment 1. As shown in FIG. A radio communication system 1 includes a transmitting device 100 and a receiving device 200 . The transmitting device 100 and the receiving device 200 perform wireless communication. Transmitting apparatus 100 convolutionally encodes data, modulates the convolutionally encoded data, and transmits the modulated data to receiving apparatus 200 . The receiving device 200 receives data transmitted from the transmitting device 100 .

The configuration and operation of transmitting apparatus 100 will be described. FIG. 2 is a diagram showing a configuration example of transmitting apparatus 100 according to Embodiment 1. As shown in FIG. Transmitting apparatus 100 includes pattern generating section 101 , encoding section 102 , bit erasing section 103 , modulating section 104 , transmission processing section 105 , and transmitting antenna 106 .

Pattern generating section 101 selects bits to be erased from the first bit string in bit erasing section 103 based on the coding rate, modulation multilevel number, likelihood combination number, error correction code structure, decoding unit, and the like. generates a bit erase pattern indicating the position of The pattern generator 101 outputs the generated bit erase pattern to the bit eraser 103 . A detailed operation of the pattern generation unit 101 will be described later.

Encoding section 102 generates a first bit string by convolutionally encoding information bits based on a predetermined encoding rate. Encoding section 102 outputs the generated first bit string to bit erasing section 103 . In this embodiment, a convolutional code with a constraint length of 5 and a coding rate of R=1/2 will be used as a specific example of the convolutional code. Since the coding rate is R=1/2, the coding section 102 outputs a coded bit string of 2 bits per 1 information bit.

The bit eraser 103 erases one or more bits for each first number of consecutive bits from the first bit string generated by the encoder 102 based on the bit erase pattern generated by the pattern generator 101. By doing so, a second bit string is generated. Bit erasing section 103 outputs the generated second bit string to modulating section 104 . By periodically erasing bits by the bit erasing unit 103, the first bit string is converted to a second bit string having a new coding rate V higher than the coding rate R used in the coding unit 102. bit string. For example, the bit erasing unit 103 erases 2 bits in the first bit string of continuous 6 bits and outputs the second bit string, thereby creating the first bit string with the coding rate R=1/2. Convert to a second bit string with a coding rate of V=3/4.

Specifically, the bit erasure pattern generated by the pattern generation unit 101 is set to [101101], and the bit corresponding to the position of [0] in the bit erasure pattern is set as an erasure bit. [101101] indicates the position of the bit to be erased by the bit eraser 103 . In this case, if a continuous 6-bit encoded bit string is [a, b, c, d, e, f], the bit erasing unit 103 removes 2 bits of [a, b, c, d, e, f] The 4th bit b and the 5th bit e are erased, and 4 bits of [a, c, d, f] are output to the modulation section 104 . Bit erasing section 103 periodically erases bits at predetermined positions within one period of the encoded bit string, thereby increasing data transmission efficiency per unit time and per unit frequency compared to the case where bits are not erased. can do. The first bit string is a bit string for one cycle, and represented by [a, b, c, d, e, f] using the above specific example. The second bit string is a bit string for one cycle obtained by erasing one or more bits from the first bit string. be. The encoded bit string is an encoded bit string composed of a first number of bits.

Modulating section 104 uses one or more of the second bit sequences generated by bit erasing section 103 to perform modulation processing using orthogonal signals to generate a plurality of symbols. As an example, a case where modulation section 104 generates two symbols from one period of the second bit string will be described. Assuming that the second bit string is represented by [a, c, d, f], the two symbols generated by modulation section 104 are a symbol representing [a, c] corresponding to the front two bits and a symbol representing the rear two bits. It becomes a symbol indicating [d, f] corresponding to a bit. Modulation section 104 generates symbols by modulating the second bit sequence using the divided bit sequences.

As an example of modulation processing using orthogonal signals performed by the modulation section 104, for example, modulation processing is performed using orthogonal 4FSK using four mutually orthogonal signals with different frequencies. Modulation section 104 establishes a one-to-one correspondence between the four bit patterns that can be taken by the two bits forming the divided second bit string and the four orthogonal signals with different frequencies. As a combination of , four combinations are determined in advance. Modulating section 104 selects signals of frequencies corresponding to a plurality of bit strings generated by dividing the second bit string obtained from bit erasing section 103 using four predetermined combinations, and modulates the selected frequencies. A signal is output to the transmission processing unit 105 . The number of bits transmitted by one symbol generated by modulating section 104 is 2, which is the same as the 2 bits forming the bit string generated by dividing the second bit string. The result of multiplying the number of 2-bit bit strings generated by dividing the second bit string in modulation section 104 by the number of bits transmitted by one symbol is the same 4 bits as the 4 bits forming the second bit string. be. The multiplication result described above is 4 bits, which is the same as the bit number of the bit erasure period generated by the pattern generation unit 101 .

FIG. 3 is a diagram showing, in the frequency domain, signals with different frequencies that are associated with bit strings generated by dividing the second bit string by modulation section 104 according to Embodiment 1. FIG. In FIG. 3, the horizontal axis represents frequency, and the vertical axis represents signal strength. Signals 11 to 14 are mutually orthogonal signals with different frequencies. For example, modulation section 104 associates bit string [0, 0] with signal 11 . Similarly, modulating section 104 associates bit string [0, 1] with signal 12 . Modulation section 104 associates bit string [1, 0] with signal 13 . Modulation section 104 associates bit string [1, 1] with signal 14 . Modulating section 104 outputs any one of signals 11 to 14 corresponding to the bit pattern of the bit string generated by dividing the second bit string obtained from bit erasing section 103 to transmission processing section 105. . A symbol generated for a bit string by modulation section 104 is any one of signals 11 to 14 .

The processing of the modulating unit 104 for two cycles of the second bit string when the bit erasing unit 103 generates the second bit string from the first bit string as one cycle will be described. As described above, bit erasing section 103 outputs to modulating section 104 a 4-bit second bit string obtained by erasing 2 bits from the 6-bit first bit string. The processing of this bit erasing unit 103 is periodically performed on the continuous first bit string. Modulating section 104 associates the 2-bit signal periodically output from bit erasing section 103 with any one of signals 11-14. For example, two 12-bit first bit strings [a, b, c, d, e, f, g, h, i, j, k, m] are obtained from the encoding unit 102, and the bit erasing unit 103 , based on the bit erase pattern [1, 0, 1, 1, 0, 1] generated by the pattern generation unit 101, bits are generated as [a, c, d, f, g, i, j, m] is periodically erased to generate two second bit strings. This processing of the bit erasing unit 103 corresponds to the bit erasing unit 103 performing two cycles of processing for erasing the bits.

In this case, modulating section 104 modulates any one of signals 11 to 14 corresponding to [a, c] corresponding to the first two bits of [a, c, d, f] corresponding to the processing result of the first period. to generate one symbol at the frequency of the selected signal. Further, modulation section 104 selects any one of signals 11 to 14 corresponding to [d, f] corresponding to the last two bits of [a, c, d, f], and modulates the frequency of the selected signal. generates one symbol of Similarly, modulation section 104 modulates signals 11 to 14 corresponding to [g, i] corresponding to the preceding two bits for [g, i, j, m] corresponding to the processing result of the second period. , one of the signals 11 to 14 corresponding to [j, m] corresponding to the last two bits is selected, and a symbol of the frequency of each selected signal is generated.

Transmission processing section 105 uses the signal of the frequency selected by modulation section 104 to perform predetermined waveform shaping processing, DA (Digital to Analog) conversion processing, up-conversion processing, power amplification processing, and the like. A high frequency analog signal is generated using the carrier frequency. Transmission processing section 105 outputs the generated high-frequency analog signal to transmission antenna 106 . A transmission antenna 106 transmits the high-frequency analog signal generated by the transmission processing unit 105 . In the following description, high-frequency analog signals may be simply referred to as signals.

The configuration and operation of receiving device 200 will be described. FIG. 4 is a diagram showing a configuration example of receiving apparatus 200 according to Embodiment 1. As shown in FIG. Reception apparatus 200 includes reception antenna 201 , reception processing section 202 , demodulation section 203 , likelihood synthesis section 204 , likelihood duplication section 205 and decoding section 206 .

Receiving antenna 201 receives a signal transmitted by transmitting apparatus 100 . Reception processing section 202 performs filter processing, down-conversion processing, AD (Analog to Digital) processing, and the like on the reception signal received by reception antenna 201, and converts the signal into a baseband signal. Reception processing section 202 outputs the converted baseband signal to demodulation section 203 .

Demodulation section 203 uses the received baseband signal to calculate a plurality of first reliabilities, which are the reliabilities of a plurality of bit strings that the symbols of the signal received by receiving apparatus 200 can have. Demodulation section 203 holds, for example, information on the same signal as the four signals with different frequencies used in modulation section 104 of transmitting apparatus 100, and performs correlation processing on the baseband signal with the four signals with different frequencies. The correlation power value calculated by performing the above is taken as the reliability. Since the four signals with different frequencies are orthogonal to each other, the correlation power value of the signal whose frequency matches the frequency of the signal output from modulation section 104 is , and the three correlation power values that do not match the frequency of the signal output from modulation section 104 are expected to take small values close to noise. The processing of demodulation section 203 is performed on baseband signals corresponding to the number of symbols transmitted from transmitting apparatus 100 . Demodulation section 203 outputs a plurality of calculated four degrees of reliability per symbol to likelihood combining section 204 . The four reliability levels calculated by the demodulator 203 are called a plurality of first reliability levels.

A likelihood synthesizing unit 204 generates a plurality of synthesized bit strings, which are bit strings composed of bits of a second number of bits, and assigns a value obtained by synthesizing a plurality of first reliabilities to the reliabilities of the plurality of synthesized bit strings. , to generate a plurality of second reliabilities that are reliabilities of the plurality of composite bitstreams. Likelihood duplicating section 205 generates a plurality of extended bit strings that are bit strings composed of bits of the first number of bits, assigns a plurality of duplicated second reliabilities to the reliabilities of the plurality of extended bit strings, and which is the reliability of the extended bit string, and is necessary when the decoding unit 206 performs the process of decoding the convolutional code. Detailed operations of likelihood combining section 204 and likelihood duplicating section 205 will be described later together with detailed operations of pattern generating section 101 of transmitting apparatus 100 and decoding section 206 of receiving apparatus 200 .

The decoding unit 206 performs processing for decoding the convolutional code with the constraint length of 5 and the coding rate R=1/2 used in the encoding by the encoding unit 102 . As described above, the bit eraser 103 erases 2 bits from the first bit string based on the bit erase pattern generated by the pattern generator 101 . On the other hand, the first reliability output from demodulation section 203 to likelihood combining section 204 is modulated using a 2-bit bit string generated by dividing the 4-bit second bit string after erasing 2 bits. Confidence for the symbols processed. Therefore, even if the first reliability is used as it is, the decoding unit 206 cannot perform a decoding process that considers the division of the erased 2 bits and the bit string.

Decoding section 206 determines the difference between the first number of bits used by bit erasing section 103 when erasing bits from the first bit string and the reciprocal 1/R of coding rate R used by coding section 102. Calculate the lowest common multiple X. The decoding unit 206 performs decoding using a trellis diagram configured so that the least common multiple X and the number of bits of the encoded bit string per branch match. In the case of this embodiment, the first number of bits used by bit erasing section 103 is 6, the coding rate R used by coding section 102 is 1/2, and the reciprocal 1/R of coding rate R is Since it is 2, the lowest common multiple X is 6. Therefore, decoding section 206 configures the trellis diagram so that the number of bits of the encoded bit string per branch is six. A coded bit string per branch is also called an extended bit string. Decoding section 206 can generate information bits used by encoding section 102 by performing Viterbi decoding using the configured trellis diagram. The decoding unit 206 creates a trellis diagram using the coding rate and the plurality of extended bit strings, and performs decoding by assigning a plurality of third reliabilities to a plurality of branches of the trellis diagram.

FIG. 5 is a flow chart showing a series of processes of the radio communication system 1 according to the first embodiment. The pattern generation unit 101 generates a bit erasure pattern based on the coding rate, the number of modulation levels, the likelihood combination number, the structure of the error correction code, the unit of decoding, and the like (step S1). The encoding unit 102 generates a first bit string by convolutionally encoding information bits based on a predetermined encoding rate (step S2). The bit eraser 103 generates the first bit erased pattern generated by the encoder 102 based on the bit erasure pattern generated by the pattern generator 101 for each number of consecutive first bits determined by the pattern generator 101 . Bits are erased from the bit string to generate a second bit string (step S3). Modulation section 104 uses the second bit sequence to perform modulation processing using orthogonal signals to generate symbols (step S4). The transmission processing unit 105 uses the symbols to generate high-frequency analog signals (step S5). The transmitting antenna 106 transmits the high frequency analog signal to the receiving antenna 201 (step S6).

The receiving antenna 201 receives the signal transmitted by the transmitting antenna 106 (step S7). The reception processing unit 202 converts the received signal into a baseband signal (step S8). The demodulator 203 uses the baseband signal to calculate a plurality of first degrees of reliability (step S9). The likelihood synthesizing unit 204 calculates a plurality of second reliabilities using the plurality of first reliabilities (step S10). The likelihood replicating unit 205 calculates a plurality of third reliabilities using the plurality of second reliabilities (step S11). The decoding unit 206 constructs a trellis diagram using a plurality of third reliabilities, and performs decoding using Viterbi decoding (step S12).

Detailed operations of likelihood synthesis section 204, likelihood duplication section 205 and decoding section 206 of receiving apparatus 200, and pattern generation section 101 of transmitting apparatus 100 will be described.

FIG. 6 shows a typical trellis line with a constraint length of 5 and a coding rate of R=1/2. FIG. 6 shows a general trellis diagram corresponding to three state transitions. State numbers 401 to 416 in FIG. 6 indicate state numbers “0000”, “0001”, . . . , “1111” in the trellis diagram. Note that the numerical values 0 or 1 added to the branches in FIG. 6 indicate correspondence between the information bits and the encoded bit strings. For example, in a branch labeled “0/00”, bit 0 is input to encoding section 102 as an information bit, and 2 bits whose bits are 00 are output from encoding section 102 as the first bit string. indicates that Note that the trellis diagram in FIG. 6 explains a general trellis diagram, and is not the trellis diagram used by decoding section 206 . Also, due to space limitations, only some branches in FIG. 6 show correspondence between information bits and coded bit strings. Correspondence with bit strings is made.

FIG. 7 is a diagram showing trellis lines used by decoding section 206 according to the first embodiment. FIG. 7 corresponds to one state transition in which three state transitions in FIG. 6 are put together. In FIG. 7, state numbers are attached only to the left end of the figure. The decoding process performed by the decoding unit 206 is performed by state transition along branches connecting state numbers from the left end to the right end of FIG. In FIG. 7, state numbers 501 to 516 indicate state numbers in the trellis diagram as in FIG. For this reason, one information bit and two coding bits are assigned to each branch per state transition in FIG. 6, whereas three information bits are assigned per state transition in FIG. , and 6 coding bits are assigned to each branch. In other words, FIG. 7 is a trellis diagram configured so that the lowest common multiple X used by the decoding unit 206 and the number of bits of the encoded bit string per branch match.

As shown in the trellis diagram of FIG. 7, the decoding unit 206 requires the reliability of the encoded bit string of 6 bits for each state transition. That is, the decoding unit 206 requires 64 types of reliability per state transition. On the other hand, the demodulator 203 can only output four reliabilities corresponding to symbols that are divided into 2 bits after 2 bits are erased by the bit eraser 103 and modulated. Therefore, in likelihood combining section 204 and likelihood duplicating section 205, receiving apparatus 200 uses the four first reliabilities output from demodulating section 203 to obtain the 64 values required by decoding section 206 as follows. A process for generating the third reliability of is performed.

The likelihood synthesizing section 204 generates 16 different second reliabilities using the second bit string and the two first reliabilities. The second bit string [a, c, d, f] obtained from the bit erasing unit 103 by the modulating unit 104 in the transmitting apparatus 100 described above is divided into two bits each of [a, c] and [d, f]. A case will be described as an example.

The second bit string [a, c, d, f] output from bit erasing section 103 is divided into 2 bits [a, c] and 2 bits [d, f] in modulating section 104. is converted into symbols for two symbols. On the other hand, demodulation section 203 sets four types of first reliability for each of the symbols generated based on [a, c] and the symbols generated based on [d, f]. Calculate That is, at the time demodulator 203 outputs the four types of first reliability to likelihood synthesizer 204, of the encoded bit string of 6 bits, which is the first number of bits used in bit eraser 103, This means that the reliability corresponding to only two bits made up of symbols is obtained.

A likelihood combining unit 204 combines a plurality of first degrees of reliability. The likelihood synthesizing unit 204 generates a plurality of synthesized bit strings, which are bit strings composed of bits of the second number of bits, and assigns the reliability of the plurality of synthesized bit strings to the reliability of the plurality of synthesized bit strings, thereby obtaining a plurality of synthesized bit strings. generate a plurality of second confidences that are confidences of The second bit number is 4, and the composite bit string having the same configuration as the second bit string is [a, c, d, f]. Therefore, the likelihood synthesis unit 204 adds [a, c, d, f]=[0, 0, 0, 0] to the reliability of [a, c, d, f]=[0, 0, 0, 0] and [a, c]=[0 , 0] and the first reliability of [d, f]=[0, 0] are assigned, and output to likelihood replication section 205 as the second reliability.

FIG. 8 is a diagram showing a process of synthesizing two of four types of first reliability to generate 16 types of second reliability in likelihood combining section 204 according to Embodiment 1. In FIG. In FIG. 8, for example, among the four types of first reliability calculated by the demodulator 203, the reliability corresponding to [a, c]=[0, 0] and the reliability corresponding to [d, f]=[0 , 0] is the highest, the reliability corresponding to [a, c, d, f]=[0, 0, 0, 0] is the highest among the 16 types of second reliability. (hereafter referred to as maximum likelihood). Furthermore, [a,c,d,f]=[0,0, 0,1], [0,0,1,0], [0,0,1,1], and a first confidence corresponding to [d,f]=[0,0]. corresponding to [a, c, d, f] = [0, 1, 0, 0], [1, 0, 0, 0], [1, 1, 0, 0] generated with the confidence of The second reliability to be obtained also has a large value (hereinafter referred to as quasi-maximum likelihood).

A likelihood duplication unit 205 duplicates a plurality of second degrees of reliability. Further, likelihood replicating section 205 generates a plurality of extended bit strings, which are bit strings configured with the first number of bits. The likelihood replication unit 205 assigns a plurality of second reliabilities replicated based on the bit erasure pattern generated by the pattern generation unit 101 to the reliabilities of the generated extension bit strings, thereby generating a plurality of extension bit strings. Generate a plurality of third confidences that are confidences of The first number of bits is 6 and the extended bit string is [a, b, c, d, e, f]. Therefore, the likelihood replication unit 205, for example, [a, b, c, d, e, f]=[0, 0, 0, 0, 0, 0], [0, 0, 0, 0, 1, 0], [0, 1, 0, 0, 0, 0], [0, 1, 0, 0, 1, 0] and [a, c, d, f ]=[0, 0, 0, 0] to generate a third reliability and output it to the decoding unit 206 . In addition, in the present embodiment, the bit erasure pattern used in likelihood replicating section 205 is shared between transmitting apparatus 100 and receiving apparatus 200 .

FIG. 9 shows that likelihood synthesizing section 204 according to Embodiment 1 synthesizes a plurality of first reliabilities to generate a second reliability, and likelihood replication section 205 replicates the second reliability. FIG. 10 is a diagram showing a trellis line when applying a process of generating a third reliability by In FIG. 9, eight branches per one state transition indicated by dashed lines are the reliability for [a, c]=[0, 0] calculated by demodulator 203 and [d, f]=[0 , 0] indicates the branch to which the combined value is replicated. In FIG. 9, the 48 branches per state transition indicated by the dotted line are the reliability for [a, c]=[0, 0] calculated by demodulation section 203, or the reliability calculated by demodulation section 203. [d, f]=[0, 0] is replicated to the branch.

The decoding unit 206 creates a trellis diagram using the coding rate and a plurality of extended bit strings, assigns 64 types of third reliability to the corresponding branches, and operates the Viterbi algorithm to perform the decoding process. to generate information bits used by encoding section 102 .

Features of the wireless communication system 1 according to this embodiment will be described in detail. A factor that influences the decoding performance of convolutional codes is the relationship between paths to be merged on a trellis diagram. In the Viterbi algorithm, the reliability of a path is calculated as the sum of the reliability of each branch that composes the path. Also, a plurality of paths that start from the same state number, go through different state transitions, and go through transitions that merge again with the same state number are one of the combinations of paths that are likely to cause decoding errors. In this embodiment, likelihood synthesizing section 204 and likelihood duplicating section 205 synthesize and duplicate the four types of reliability calculated by demodulating section 203, and perform processing to generate 64 types of reliability. there is Specifically, the likelihood synthesizing unit 204 synthesizes the two first reliabilitys to generate a second reliability, and the likelihood replication unit 205 replicates the second reliability to generate a third reliability. yields a confidence of Therefore, in the decoding unit 206, many branches are generated to which reliability values equal to or less than the specified difference are assigned. The reliability of a value equal to or less than a specified difference is a reliability having a small difference between two or more reliabilitys, that is, having values close to each other. The decoding unit 206 cannot discriminate the likelihood between branches to which the same or a reliability value equal to or less than a specified difference is assigned. As a result, decoding section 206 may have more decoding errors than a decoding method that does not combine and duplicate reliability. To which branch in the trellis diagram of the decoding unit 206 the first reliability generated in the demodulation unit 203 is assigned depends on the bit erasure pattern generated in the pattern generation unit 101 and the error correction used in the encoding unit 102. It depends on the structure of the code and the number of divisions of the second bit string to generate one modulation symbol in modulation section 104 , that is, the number of symbols to be combined in likelihood combining section 204 .

Therefore, in the trellis diagram used in the decoding unit 206, the pattern generation unit 101 has a specified number of branches to which the same or a specified difference or less in reliability is assigned between the paths whose probabilities are compared. A bit erasure pattern is generated that reduces the number of bits. FIG. 10 is a diagram showing four shortest merging paths A to D for the trellis diagram shown in FIG. In FIG. 10, a branch 701 and a branch 702 indicate branches forming path A among the four shortest merging paths AD. Similarly, branches 703 and 704 denote the branches forming path B, branches 705 and 706 denote the branches forming path C, and branches 707 and 708 denote the branches forming path D. Paths A to D start from state number 0000 at the left end, go through two state transitions, and merge at state number 0000 again. Therefore, paths A to D are one of the combinations of paths in which decoding errors are likely to occur. In FIG. 10, the reliability of path A is obtained as the sum of the reliability of branch 701 corresponding to coded bit string 000000 and the reliability of branch 702 corresponding to coded bit string 000000. FIG. Furthermore, the reliability of branch 701 and branch 702 is a third reliability obtained by replicating the second reliability generated by synthesizing the two first reliabilitys in likelihood synthesizing section 204 by likelihood replicating section 205 . Confidence is assigned.

As described above, the first reliability assigned to a branch in the trellis diagram is obtained by combining the first reliability in likelihood combining section 204 and replicating the second reliability in likelihood replicating section 205. Based on the combination of extended bit strings, in other words, the bit erasing pattern generated by pattern generating section 101 used in bit erasing section 103, and the number of symbols generated by dividing one second bit string in modulating section 104. It is determined. In the present embodiment, the bit erasing section 103 erases 2 bits from the first bit string consisting of 6 bits to generate the second bit string consisting of 4 bits. There are 15 bit erase patterns: [0, 0, 1, 1, 1, 1], [0, 1, 0, 1, 1, 1], ..., [1, 1, 1, 1, 0, 0] exist. Further, in the present embodiment, modulation section 104 divides the second bit string composed of 4 bits into two, the symbol corresponding to the leading 2 bits of the second bit string and the symbol corresponding to the trailing bit string of the second bit string. Generate a symbol corresponding to 2 bits.

Therefore, pattern generation section 101 applies a bit erasure pattern to the encoded bits corresponding to the branches in the trellis diagram assigned based on the structure of the error correction code used in encoding section 102, and converts the extracted 4 bits to As in the modulation section 104, when divided into 2 bits in the front and 2 bits in the rear, the number of branches to which the same first reliability is assigned between the paths for which the probabilities are compared in the trellis diagram is small, and each path is Bit erasure patterns are selected that contain different numbers of branches that are assigned the same first reliability. For example, consider the case where the bit erase pattern [1, 1, 0, 0, 1, 1] is applied to the shortest merge path shown in FIG. Assuming that the first bit string is [a, b, c, d, e, f], the second bit string is [a, b, e, f]. The two bit strings are [a,b] and [e,f]. At this time, since the corresponding coded bit string is 000000, the branch 701 has the reliability for [a, b]=[0, 0] and the reliability for [e, f]=[0, 0] among the first reliability. ] is assigned.

Similarly, branch 702 is assigned a combined value of the first confidence for [a,b]=[0,0] and the confidence for [e,f]=[0,0]. be done. Branch 703 is assigned a value obtained by combining the reliability of [a, b]=[1, 1] and the reliability of [e, f]=[1, 0] among the first reliability. A branch 704 is assigned a combined value of the reliability for [a,b]=[0,1] and the reliability for [e,f]=[0,0] among the first reliability. Branch 705 is assigned a value obtained by combining the reliability of [a, b]=[0, 0] and the reliability of [e, f]=[1, 0] among the first reliability. Branch 706 is assigned a value obtained by combining the reliability for [a,b]=[1,0] and the reliability for [e,f]=[1,1] among the first reliability. Branch 707 is assigned a value obtained by combining the reliability for [a, b]=[1, 1] and the reliability for [e, f]=[0, 0] among the first reliability. Branch 708 is assigned a combined value of the first reliability for [a,b]=[1,1] and the reliability for [e,f]=[1,1].

Here, assuming an ideal state without noise, when the second bit string used in modulation section 104 is [a, b, e, f]=[0, 0, 0, 0], the first , the reliability corresponding to [a, b]=[0, 0] and the reliability corresponding to [e, f]=[0, 0] are maximum. Therefore, branch 701 and branch 702 is the maximum likelihood branch with the highest reliability among all branches. Furthermore, the reliability of path A generated by summing the reliability assigned to branch 701 and the reliability assigned to branch 702 is the maximum among all paths, and path A becomes the maximum likelihood path. Also, a branch 704 to which a reliability corresponding to [a, b]=[0, 0] or a reliability corresponding to [e, f]=[0, 0] among the first reliability is assigned; 705 and branch 707 are quasi-maximum likelihood branches having the second highest reliability after the maximum likelihood branch, and paths BD including these branches are quasi-maximum likelihood paths. Therefore, when comparing the probabilities among the four shortest merge paths shown in FIG. 10, paths A to D all have large values, making it difficult to distinguish between them, and decoding section 206 may increase decoding errors.

On the other hand, when the bit erasure pattern is [1, 0, 1, 1, 1, 0], the second bit string is [a, c, d, e], which is used by modulation section 104 for symbol generation. Bit strings obtained by dividing the second bit string are [a, c] and [d, e]. At this time, among the four shortest merge paths shown in FIG. 701 and 702, resulting in a maximum likelihood path. However, paths BD do not contain any maximum-likelihood or quasi-maximum-likelihood branches. Therefore, when decoding section 206 discriminates between the four shortest merge paths shown in FIG. 10, errors are less likely to occur. It can be seen that deterioration in decoding performance due to bit erasure can be suppressed more when the bit erasure pattern is set to [1, 0, 1, 1, 1, 0]. Therefore, in the present embodiment, pattern generation section 101 generates [1, 0, 1, 1, 1, 0] as a bit erasure pattern and outputs it to bit erasure section 103 . In the trellis diagram used in the decoding unit 206, the pattern generation unit 101 selects a branch to which a reliability value equal to or less than a specified difference is assigned among the shortest paths to be merged among the paths whose probabilities are compared. A bit erasure pattern is generated that is less than or equal to the specified number.

In this embodiment, the case where the bit erase pattern is selected based on the four shortest merge paths shown in FIG. 10 has been exemplified, but the present invention is not limited to this. For example, while the second shortest merge path, that is, the shortest merge path shown in FIG. effect is obtained.

FIG. 11 is a diagram showing 22 paths that start from state number 0000 and merge again at state number 0000 through three state transitions in the trellis diagram of FIG. As in the case of the shortest merge paths shown in FIG. 10, pattern generator 101 applies possible bit erasure patterns to the coded bits corresponding to each branch included in these paths, and forwards the extracted 4 bits. Two bit strings per branch obtained by dividing into 2 bits and rear 2 bits are calculated. Then, pattern generation section 101 selects a bit erasure pattern in which the number of maximum likelihood branches and quasi-maximum likelihood branches included in 22 paths passing through 3 state transitions is small. For example, when the pattern generation unit 101 applies the bit erasure pattern [1, 1, 0, 0, 1, 1], 20 of the 22 merge paths shown in FIG. Since branches are included, there is a risk that decoding errors will increase in decoding section 206 . On the other hand, when the pattern generation unit 101 applies the bit erasure pattern [1, 0, 1, 1, 1, 0], among the 22 merge paths shown in FIG. Since the number of paths is reduced to 14, it can be seen that decoding errors can be suppressed in decoding section 206 . Therefore, the pattern generator 101 generates [1, 0, 1, 1, 1, 0] as the bit erase pattern and outputs it to the bit eraser 103 . Pattern generation section 101 generates a bit erasure pattern based on the structure of the error correction code used in encoding section 102 and the number of first reliability levels combined in likelihood combining section 204 .

Further, in the present embodiment, the pattern generating section 101 selects a bit erasure pattern with a small number of paths including maximum likelihood branches or quasi-maximum likelihood branches in the case of the constraint length 5 shown in FIG. , but not limited to. For example, pattern generation section 101 selects a bit erasure pattern in which the number of paths including maximum likelihood branches or quasi-maximum likelihood branches is small and the number of maximum likelihood branches or quasi-maximum likelihood branches included in each path varies. The same effect can be obtained with

FIG. 12 is a diagram showing eight shortest merging paths among the trellis lines when the constraint length is 7, the coding rate is R=1/2, and 6 coding bits are assigned to each branch. . The eight merge paths in FIG. 12 start from state number 000000 at the left end, and merge again at state number 000000 through three state transitions. As in the case of the constraint length 5 shown in FIG. 10, the pattern generator 101 applies possible bit erasure patterns to the coded bits corresponding to each branch included in these paths, and modulates the extracted bit string. A plurality of bit strings obtained by division in the same manner as in section 104 are calculated. For example, as in FIG. 10, there are 15 possible bit patterns when the coding rate is V=3/4. A bit erasure pattern is selected that has a small number of maximum likelihood branches and that the number of maximum likelihood branches and quasi-maximum likelihood branches included in each pass varies.

For example, when the pattern generation unit 101 applies the bit erase pattern [1, 0, 1, 1, 0, 1] and when the bit erase pattern [0, 1, 1, 1, 1, 0] is applied , the number of paths containing maximum-likelihood and quasi-maximum-likelihood branches is the same as 4 out of 8. However, each path contains a different number of maximum-likelihood or quasi-maximum-likelihood branches. When the pattern generation unit 101 applies the bit erasure pattern [1, 0, 1, 1, 0, 1], one path includes two maximum likelihood branches and one path includes three quasi-maximum likelihood branches. , one path including two quasi-maximum likelihood branches, and one path including one quasi-maximum likelihood branch. On the other hand, when the pattern generation unit 101 applies the bit erasure pattern [0, 1, 1, 1, 1, 0], there is one path including two maximum likelihood branches and a path including two quasi-maximum likelihood branches. is composed of three. Compared to the case where the pattern generation unit 101 applies the bit erasure pattern [0, 1, 1, 1, 1, 0], the pattern generation unit 101 applies the bit erasure pattern [1, 0, 1, 1, 0, 1]. ] is applied, it is easy to distinguish between the shortest merge paths shown in FIG. Therefore, the pattern generator 101 generates [1, 0, 1, 1, 0, 1] as the bit erase pattern and outputs it to the bit eraser 103 .

Further, in the present embodiment, the case where the bit erasing section 103 erases 2 bits out of every 6 bits to set the coding rate to V=3/4 has been exemplified, but the present invention is not limited to this. This embodiment can also be applied, for example, to a case where the bit erasing section 103 erases 2 bits out of every 8 bits so that the coding rate is V=2/3. In this case, modulating section 104 outputs symbols corresponding to 2-bit bit strings obtained by dividing the 6-bit second bit string into three. Further, the decoding unit 206 obtains the least common multiple X=8 of the first number of bits 8 used in the bit erasing unit 103 and the reciprocal 2/1 of the encoding rate used in the encoding unit 102, and matches it. Decoding is performed using a trellis diagram configured so that the number of bits in the encoded bit string per branch is eight. Therefore, based on these pieces of information, the pattern generation unit 101 generates branches to which reliability values equal to or less than a specified difference are assigned between paths for which probabilities are compared in the trellis diagram used by the decoding unit 206. is less than or equal to a specified number, that is, a bit erase pattern is generated.

For example, when the pattern generation unit 101 generates the bit erasure pattern [1, 0, 1, 0, 1, 1, 1, 1], the bit erasure unit 103 generates the output [a, b, c, d, e, f, g, h] to erase 2 bits [b, d] and output a bit string of [a, c, e, f, g, h]. Modulating section 104 divides the 6-bit bit string output from bit erasing section 103 into three, and each 2 bits of [a, c], [e, f], and [g, h] becomes one symbol. A symbol is constructed using 4FSK. Demodulation section 203 calculates four types of first reliability corresponding to 4FSK and outputs them to likelihood combining section 204 . Likelihood synthesizing section 204 synthesizes the four first reliabilities for three symbols to generate 64 second reliabilities, and outputs the generated second reliabilities to likelihood duplicating section 205 . Likelihood duplicating section 205 duplicates 256 types of third reliability from 64 types of second reliability and outputs them to decoding section 206 . Decoding section 206 performs decoding using 256 third reliabilities output from likelihood replicating section 205 based on the trellis diagram in which 8 bits are allocated per branch.

Furthermore, in this embodiment, the modulation method of modulation section 104 is FSK, but the present invention is not limited to this, and the method is not limited to FSK as long as it is a method that can calculate reliability in units of symbols. This embodiment is preferably applied to a system that generates a modulated signal based on an orthogonal signal, such as an FSK or M-ary transmission system, in which it is particularly difficult to accurately calculate the reliability per bit. As a method for generating modulated signals, for example, a case where an M-ary transmission method (M=4) using length-4 Hadamard sequences is applied instead of 4FSK used in this embodiment will be described. When applying the M-ary transmission scheme, modulation section 104 generates symbols using four mutually orthogonal sequences generated from Hadamard sequences instead of four mutually orthogonal signals with different frequencies. In addition, as in the case of using 4FSK, modulation section 104 determines in advance the correspondence between the two leading bits or the two trailing bits of the second bit string, and converts the orthogonal sequence according to the input second bit string. Select and output. Demodulation section 203 holds the same four orthogonal sequences used in modulation section 104, and performs correlation processing on the baseband signal with the four orthogonal sequences to calculate a correlation power value. Also, demodulation section 203 outputs the correlation power value as reliability, as in the case of 4FSK. Other processing is the same as the processing in the case of 4FSK.

Further, in the present embodiment, transmitting apparatus 100 includes pattern generating section 101 and shares the bit erasure pattern generated by pattern generating section 101 with receiving apparatus 200 . A similar effect can also be obtained by adopting a configuration in which the bit erasure pattern is shared with the transmission device 100 . Furthermore, the pattern generation unit 101 may generate a bit erase pattern for each operation of the bit erase unit 103 once or a plurality of times, or may generate a bit erase pattern only when the bit erase unit 103 operates for the first time. good.

Further, in the present embodiment, decoding section 206 obtains the least common multiple X between the first number of bits in bit erasing section 103 and the reciprocal 1/R of coding rate R used in encoding section 102, and obtains the least common multiple Although the trellis diagram of branches in which X and the number of bits per branch are the same, the present invention is not limited to this. The decoding unit 206 can obtain the same effect even if the number of bits per branch is a common multiple other than the lowest common multiple X. FIG.

In this embodiment, demodulation section 203 is configured to use the correlation power value between the baseband signal and the frequency signal as a method for generating the first reliability, but the present invention is not limited to this. For example, the demodulator 203 calculates a complex correlation value between the baseband signal and the frequency signal, and outputs the real component of the calculated complex correlation value as the first reliability. The process of calculating the first reliability using this complex correlation value corresponds to synchronous detection FSK processing, and the demodulator 203 generates reliability that is more resistant to noise than when using the correlation power value. can be done.

Next, the hardware configuration of the radio communication system 1 will be explained. In radio communication system 1, transmitting antenna 106 and receiving antenna 201 are antenna elements. Pattern generation section 101, encoding section 102, bit elimination section 103, modulation section 104, transmission processing section 105, reception processing section 202, demodulation section 203, likelihood synthesis section 204, likelihood duplication section 205, and decoding section 206 , is implemented by a processing circuit. The processing circuitry may be a processor and memory executing programs stored in the memory, or may be dedicated hardware. Processing circuitry is also called control circuitry.

FIG. 13 is a diagram showing a configuration example of the processing circuit 90 when the processing circuit included in the wireless communication system 1 according to the first embodiment is realized by the processor 91 and the memory 92. As shown in FIG. A processing circuit 90 shown in FIG. 13 is a control circuit and includes a processor 91 and a memory 92 . When the processing circuit 90 is composed of the processor 91 and the memory 92, each function of the processing circuit 90 is implemented by software, firmware, or a combination of software and firmware. Software or firmware is written as a program and stored in memory 92 . In the processing circuit 90, each function is realized by the processor 91 reading and executing the program stored in the memory 92. FIG. That is, the processing circuitry 90 includes a memory 92 for storing programs that result in the processing of the wireless communication system 1 being executed. This program can also be said to be a program for causing the wireless communication system 1 to execute each function realized by the processing circuit 90 . This program may be provided by a storage medium storing the program, or may be provided by other means such as a communication medium.

In the above program, in the transmitting device 100, the pattern generating section 101 generates a bit erasure pattern in a first step, and the coding section 102 convolutionally codes the information bits based on the coding rate. a second step of generating a first bit string; a third step of generating a second bit string based on the number of bits of the modulation unit 104, a fourth step of generating a symbol by modulating the second bit string using the divided bit string, and a receiving device In 200, a fifth step in which demodulation section 203 calculates a plurality of first reliabilities that are reliabilities of a plurality of bit strings that a symbol can take; and assigning a value obtained by combining the plurality of first reliabilities to the reliabilities of the plurality of synthesized bit strings, and a plurality of second and a likelihood replication unit 205 generates a plurality of extended bit strings that are bit sequences composed of bits of the first number of bits, and replicates the reliability of the plurality of extended bit strings. a seventh step of assigning a plurality of second reliabilities and generating a plurality of third reliabilities that are the reliabilities of the plurality of extended bit strings; and an eighth step of decoding by creating a trellis diagram using and assigning a plurality of third reliability levels to a plurality of branches of the trellis diagram. .

Here, the processor 91 is, for example, a CPU (Central Processing Unit), a processing device, an arithmetic device, a microprocessor, a microcomputer, or a DSP (Digital Signal Processor). The memory 92 is a non-volatile or volatile memory such as RAM (Random Access Memory), ROM (Read Only Memory), flash memory, EPROM (Erasable Programmable ROM), EEPROM (registered trademark) (Electrically EPROM). A semiconductor memory, a magnetic disk, a flexible disk, an optical disk, a compact disk, a mini disk, or a DVD (Digital Versatile Disc) is applicable.

FIG. 14 is a diagram showing an example of the processing circuit 93 when the processing circuit included in the wireless communication system 1 according to Embodiment 1 is configured by dedicated hardware. The processing circuit 93 shown in FIG. 14 is, for example, a single circuit, a composite circuit, a programmed processor, a parallel programmed processor, an ASIC (Application Specific Integrated Circuit), an FPGA (Field Programmable Gate Array), or a combination thereof. thing applies. The processing circuit may be partly implemented by dedicated hardware and partly implemented by software or firmware. Thus, the processing circuitry may implement each of the functions described above through dedicated hardware, software, firmware, or a combination thereof.

As described above, according to the present embodiment, in transmitting apparatus 100, pattern generation section 101 generates a to generate a bit erase pattern. The bit eraser 103 erases at least one bit from the first bit string based on the bit erase pattern to generate a second bit string. Modulation section 104 generates symbols using bit strings obtained by dividing the second bit string. In receiving apparatus 200 , demodulation section 203 obtains the first reliability for the symbols output from modulation section 104 . A likelihood synthesizing unit 204 synthesizes a plurality of first reliabilitys to obtain a second reliability. The likelihood replication unit 205 uses the second reliability to generate a third reliability that the decoding unit 206 uses in the decoding process. The decoding unit 206 uses the third reliability to create a trellis diagram in which the number of bits of the encoded bit string per branch matches the least common multiple X, and decodes the convolutional code using the trellis diagram. process.

As a result, in the radio communication system 1, the transmitting apparatus 100 encodes at a coding rate such that the reciprocal of the coding rate is not a natural number, such as the coding rate R=3/4, and further modulates the multilevel number and the bit Even if the erasure period does not match, decoding using symbol-by-symbol reliability makes it possible to achieve good error rate performance. Therefore, receiving apparatus 200 can flexibly set the transmission rate. In other words, the radio communication system 1 can achieve good decoding characteristics and flexible transmission rate settings at the same time.

Embodiment 2.
In Embodiment 1, radio communication system 1 is configured such that transmitting apparatus 100 and receiving apparatus 200 perform radio communication. In Embodiment 2, a case will be described in which a wireless communication device includes a transmitting device 100 and a receiving device 200, and wireless communication is performed between the wireless communication devices.

FIG. 15 is a diagram showing a configuration example of a radio communication system 1a according to the second embodiment. The radio communication system 1 a includes two radio communication devices 300 . The wireless communication device 300 can perform two-way wireless communication with another wireless communication device 300 . In radio communication apparatus 300, the configurations and operations of transmitting apparatus 100 and receiving apparatus 200 are the same as those of transmitting apparatus 100 and receiving apparatus 200 in the first embodiment.

Embodiment 3.
In Embodiment 1, the pattern generation unit 101 of the transmission device 100 generates the same value or a value equal to or less than a specified difference between the paths for which the probability is compared in the trellis diagram used in the decoding unit 206 of the reception device 200. The case of generating a bit erasure pattern in which the number of branches to which a reliability of . In the third embodiment, the pattern generation unit 101 detects the difference between the information bit string and the maximum likelihood bit string associated with the branch to which the reliability of the same or less than the specified difference is assigned among the bit erasure patterns described above. A case of generating a bit erase pattern in which is less than or equal to a prescribed difference will be described.

In the Viterbi algorithm, when a branch selection error occurs, a decoding error occurs in units of branches to which extension bit strings are assigned, that is, in units of symbols. Therefore, by using the bit erasure pattern described in Embodiment 1, it is expected that the decoding error rate in symbol units, which is an index for evaluating the decoding performance of convolutional codes, can be improved.

Another index for evaluating the decoding performance of a convolutional code is the bitwise decoding error rate. The decoding error rate in bit units is caused by the information bit string assigned to the erroneously selected branch in the Viterbi algorithm. For example, if the information bit string corresponding to the first transition in the trellis diagram shown in FIG. , a symbol error occurs. At this time, even if any branch other than the branch input to the state number 000000 is selected, the number of error symbols is two symbols constituting the reliability assigned to the branch. However, since the information bit strings assigned to each branch are different at the first transition of the trellis diagram shown in FIG. 12, the number of error bits caused by the selected branches is also different. For example, when a branch input to state number 001000 is selected, the assigned information bit string is [001], so the number of error bits is 1 bit. On the other hand, when the branch input to the state number 111000 is selected, the assigned information bit string is [111], so the number of error bits is 3 bits. decoding error rate may be degraded.

Therefore, in the third embodiment, the pattern generation unit 101 generates a bit erasure pattern in which the number of branches to which reliability values equal to or less than a specified difference are assigned is less than or equal to a specified number between paths whose probabilities are compared. Among them, a bit erasure pattern is generated such that the difference between the information bit corresponding to the branch to which the reliability of the same value or a value less than the specified difference is assigned and the maximum likelihood bit is less than the specified difference, that is, less. . That is, the pattern generation unit 101 specifies the number of bits different from the maximum likelihood bit string in the information bit string corresponding to the branch to which the reliability of the same or less than the specified difference is assigned between the paths whose likelihoods are compared. Generates a bit erase pattern that is less than or equal to the specified number.

FIG. 16 shows that likelihood synthesizing section 204 according to Embodiment 3 synthesizes a plurality of first reliabilities to generate a second reliability, and likelihood replication section 205 replicates the second reliability. In the trellis diagram when the process for generating the third reliability is applied, the bit erasure pattern [10111101] and bit erasure pattern [01111110] are applied. FIG. FIG. 16 is a trellis diagram with a constraint length of 9, a coding rate of R=1/2, and 8 coded bits per branch. and an encoded bit string extracted when the bit erasure pattern is [10111101] and when the bit erasure pattern is [01111110]. The 16 merge paths in FIG. 16 start from state number 00000000 and merge again at state number 00000000 through three state transitions.

In Embodiment 3, the bit erasing section 103 converts the coding rate to V=2/3, and the likelihood combining section 204 combines three of the four types of first reliability to obtain 64 types of second reliability. generate degrees. At this time, the extracted 6-bit encoded bit string shown in FIG. 16 is associated with three symbols composed of 2-bit units. Therefore, for example, when the reliability corresponding to the symbol [00] is the highest, if [00] is included in the extracted coded bit string, the reliability of the corresponding branch increases, and between the paths that compare the probability Selection errors may occur. The bit erasure pattern [10111101] and the bit erasure pattern [01111110] shown in FIG. 16 have the same number of branches including [00], that is, the number of branches to which reliability values equal to or less than the specified difference are assigned. Therefore, it is assumed that these symbol error rates are the same.

On the other hand, in the third embodiment, the pattern generation unit 101 generates the information bit string and the maximum likelihood bit string [0000] assigned to each branch of the bit erase pattern [10111101] and the bit erase pattern [01111110]. A bit erase pattern is generated such that the difference between is less than or equal to the specified difference. From FIG. 16, when the bit erasure pattern [10111101] is applied, in the information bit string corresponding to the path including [00] in the encoded bit string, the difference from the maximum likelihood bit string [0000], that is, the number of 1s included in the information bit string Calculating the number results in 26 bits. On the other hand, when the bit erase pattern [01111110] is applied, the number of 1's included in the information bit string is calculated to be 27 bits. Therefore, the bit erasure pattern [10111101] can suppress decoding errors in bit units. Therefore, in the third embodiment, the pattern generator 101 generates [10111101] as the bit erase pattern and outputs it to the bit eraser 103 .

The configurations shown in the above embodiments are only examples, and can be combined with other known techniques, or can be combined with other embodiments without departing from the scope of the invention. It is also possible to omit or change part of the configuration.

1, 1a wireless communication system, 100 transmitter, 101 pattern generator, 102 encoder, 103 bit eraser, 104 modulator, 105 transmission processor, 106 transmitter antenna, 200 receiver, 201 receiver antenna, 202 receiver process section, 203 demodulation section, 204 likelihood synthesis section, 205 likelihood duplication section, 206 decoding section, 300 radio communication apparatus.

Claims (16)

a pattern generator that generates a bit erase pattern;
an encoding unit that generates a first bit string by convolutionally encoding information bits based on an encoding rate;
a bit eraser that generates a second bit string based on a second number of bits by erasing one or more bits from the first bit string for each first number of bits based on the bit erase pattern;
a modulation unit that generates a symbol by modulating the second bit string using a bit string obtained by dividing the second bit string;
a transmitting device comprising
a demodulator that calculates a plurality of first reliabilities that are reliabilities of a plurality of bit strings that the symbol can take;
generating a plurality of synthesized bit strings that are bit strings composed of bits of the second number of bits; assigning a value obtained by synthesizing the plurality of first reliabilities to reliabilities of the plurality of synthesized bit strings; a likelihood combining unit that generates a plurality of second reliabilities that are reliabilities of the combined bit string;
generating a plurality of extended bit strings that are bit strings composed of bits of the first number of bits, assigning the plurality of second reliabilities that are duplicated to the reliabilities of the plurality of extended bit strings, and assigning the plurality of extended bit strings to the plurality of extended bit strings a likelihood replication unit that generates a plurality of third confidences that are confidences of
a decoding unit for decoding by creating a trellis diagram using the coding rate and the plurality of extended bit sequences and assigning the plurality of third reliabilities to a plurality of branches of the trellis diagram;
a receiving device comprising
with
The pattern generation unit reduces the number of branches to which reliability values equal to or less than a specified difference are assigned between paths for which probabilities are compared in a trellis diagram used in the decoding unit is equal to or less than a specified number. generating the bit erase pattern;
A wireless communication system characterized by: In the trellis diagram used in the decoding unit, the pattern generation unit selects a branch to which a reliability value equal to or less than a specified difference is assigned among the shortest paths to be merged among the paths for which the probability is compared. generating the bit erase pattern that is equal to or less than a specified number;
The wireless communication system according to claim 1 , characterized by: The pattern generation unit generates the bit erasure pattern based on the structure of the error correction code used in the encoding unit and the number of the first degrees of reliability synthesized by the likelihood synthesis unit.
The wireless communication system according to claim 1 or 2 , characterized by: In the trellis diagram used in the decoding unit provided in the receiving device, bit erasure in which the number of branches to which reliability values equal to or less than a specified difference are assigned is less than or equal to a specified number between paths for which likelihoods are compared. a pattern generator that generates a pattern;
an encoding unit that generates a first bit string by convolutionally encoding information bits based on an encoding rate;
a bit eraser that generates a second bit string based on a second number of bits by erasing one or more bits for each first number of bits from the first bit string based on the bit erase pattern; ,
A transmitting device comprising: In the trellis diagram used in the decoding unit provided in the receiving device, the maximum likelihood bit string in the information bit string corresponding to the branch to which the reliability of the same or less than the specified difference is assigned between the paths for which the probability is compared a pattern generator that generates a bit erase pattern in which the number of bits different from the
an encoding unit that generates a first bit string by convolutionally encoding information bits based on an encoding rate;
a bit eraser that generates a second bit string based on a second number of bits by erasing one or more bits for each first number of bits from the first bit string based on the bit erase pattern; ,
A transmitting device comprising: A transmitter according to claim 4 or 5 ;
a demodulator that calculates a plurality of first reliabilities that are reliabilities of a plurality of bit strings that a symbol can take;
generating a plurality of synthesized bit strings that are bit strings composed of bits of a second number of bits, assigning a value obtained by synthesizing the plurality of first reliabilities to reliabilities of the plurality of synthesized bit strings, and combining the plurality of synthesized bit strings; a likelihood synthesizer that generates a plurality of second reliabilities that are reliabilities of bit strings;
generating a plurality of extended bit strings that are bit strings composed of bits of a first number of bits, assigning the plurality of second reliabilities that are duplicated to the reliabilities of the plurality of extended bit strings; a likelihood replication unit that generates a plurality of third confidences that are confidences;
a decoding unit for decoding by creating a trellis diagram using the coding rate and the plurality of extended bit sequences and assigning the plurality of third reliabilities to a plurality of branches of the trellis diagram;
a receiving device comprising
A wireless communication system comprising: A transmitter according to claim 4 or 5 ;
a demodulator that calculates a plurality of first reliabilities that are reliabilities of a plurality of bit strings that a symbol can take;
generating a plurality of synthesized bit strings that are bit strings composed of bits of a second number of bits, assigning a value obtained by synthesizing the plurality of first reliabilities to reliabilities of the plurality of synthesized bit strings, and combining the plurality of synthesized bit strings; a likelihood synthesizer that generates a plurality of second reliabilities that are reliabilities of bit strings;
generating a plurality of extended bit strings that are bit strings composed of bits of a first number of bits, assigning the plurality of second reliabilities that are duplicated to the reliabilities of the plurality of extended bit strings; a likelihood replication unit that generates a plurality of third confidences that are confidences;
a decoding unit for decoding by creating a trellis diagram using the coding rate and the plurality of extended bit sequences and assigning the plurality of third reliabilities to a plurality of branches of the trellis diagram;
a receiving device comprising
A wireless communication device comprising: A control circuit for controlling a wireless communication system,
In a trellis diagram used in decoding, generating a bit erasure pattern in which the number of branches to which reliability values equal to or less than a specified difference are assigned between paths for which probabilities are compared is less than or equal to a specified number;
generating a first bit string by convolutionally encoding the information bits based on the coding rate;
generating a second bit string based on a second number of bits by erasing one or more bits for each first number of bits from the first bit string based on the bit erase pattern;
generating a symbol by modulating the second bit string using a bit string obtained by dividing the second bit string;
A control circuit, characterized in that it causes the wireless communication system to implement: A control circuit for controlling a wireless communication system,
In the trellis diagram used in decoding, the number of bits different from the maximum likelihood bitstring in the information bitstring corresponding to the branch assigned with the same or less than the specified difference between the paths whose likelihoods are compared. Generates a bit erase pattern in which is less than or equal to a specified number,
generating a first bit string by convolutionally encoding the information bits based on the coding rate;
generating a second bit string based on a second number of bits by erasing one or more bits for each first number of bits from the first bit string based on the bit erase pattern;
generating a symbol by modulating the second bit string using a bit string obtained by dividing the second bit string;
A control circuit, characterized in that it causes the wireless communication system to implement: A control circuit for controlling a wireless communication system,
Generate a bit erasure pattern in which, in the trellis diagram used in decoding, there are no more than a specified number of branches that are assigned confidence values equal to or less than a specified difference between paths for which probabilities are compared, or , in the trellis diagram used in decoding, in the information bit string corresponding to the branch to which the reliability of the same or less than the specified difference is assigned between the paths for which the probability is compared, the maximum likelihood bit string and the number of bits different from the Generate a bit erase pattern whose number is less than or equal to the specified number,
generating a first bit string by convolutionally encoding the information bits based on the coding rate;
generating a second bit string based on a second number of bits by erasing one or more bits for each first number of bits from the first bit string based on the bit erase pattern;
generating a symbol by modulating the second bit string using a bit string obtained by dividing the second bit string;
calculating a plurality of first reliabilities that are reliabilities of a plurality of bit strings that the symbol can take;
generating a plurality of synthesized bit strings that are bit strings composed of bits of the second number of bits; assigning a value obtained by synthesizing the plurality of first reliabilities to reliabilities of the plurality of synthesized bit strings; generating a plurality of second confidences that are confidences of the composite bitstream;
generating a plurality of extended bit strings that are bit strings composed of bits of the first number of bits, assigning the plurality of second reliabilities that are duplicated to the reliabilities of the plurality of extended bit strings, and assigning the plurality of extended bit strings to the plurality of extended bit strings generate a plurality of third confidences that are the confidences of
Decoding by creating a trellis diagram using the coding rate and the plurality of extended bit sequences, and assigning the plurality of third reliabilities to a plurality of branches of the trellis diagram;
A control circuit, characterized in that it causes the wireless communication system to implement: A storage medium storing a program for controlling a wireless communication system,
Said program
In a trellis diagram used in decoding, generating a bit erasure pattern in which the number of branches to which reliability values equal to or less than a specified difference are assigned between paths for which probabilities are compared is less than or equal to a specified number;
generating a first bit string by convolutionally encoding the information bits based on the coding rate;
generating a second bit string based on a second number of bits by erasing one or more bits for each first number of bits from the first bit string based on the bit erase pattern;
generating a symbol by modulating the second bit string using a bit string obtained by dividing the second bit string;
A storage medium, characterized in that it causes the wireless communication system to implement: A storage medium storing a program for controlling a wireless communication system,
Said program
In the trellis diagram used in decoding, the number of bits different from the maximum likelihood bitstring in the information bitstring corresponding to the branch assigned with the same or less than the specified difference between the paths whose likelihoods are compared. Generates a bit erase pattern in which is less than or equal to a specified number,
generating a first bit string by convolutionally encoding the information bits based on the coding rate;
generating a second bit string based on a second number of bits by erasing one or more bits for each first number of bits from the first bit string based on the bit erase pattern;
generating a symbol by modulating the second bit string using a bit string obtained by dividing the second bit string;
A storage medium, characterized in that it causes the wireless communication system to implement: A storage medium storing a program for controlling a wireless communication system,
Said program
Generate a bit erasure pattern in which, in the trellis diagram used in decoding, there are no more than a specified number of branches that are assigned confidence values equal to or less than a specified difference between paths for which probabilities are compared, or , in the trellis diagram used in decoding, in the information bit string corresponding to the branch to which the reliability of the same or less than the specified difference is assigned between the paths for which the probability is compared, the maximum likelihood bit string and the number of bits different from the Generate a bit erase pattern whose number is less than or equal to the specified number,
generating a first bit string by convolutionally encoding the information bits based on the coding rate;
generating a second bit string based on a second number of bits by erasing one or more bits for each first number of bits from the first bit string based on the bit erase pattern;
generating a symbol by modulating the second bit string using a bit string obtained by dividing the second bit string;
calculating a plurality of first reliabilities that are reliabilities of a plurality of bit strings that the symbol can take;
generating a plurality of synthesized bit strings that are bit strings composed of bits of the second number of bits; assigning a value obtained by synthesizing the plurality of first reliabilities to reliabilities of the plurality of synthesized bit strings; generating a plurality of second confidences that are confidences of the composite bitstream;
generating a plurality of extended bit strings that are bit strings composed of bits of the first number of bits, assigning the plurality of second reliabilities that are duplicated to the reliabilities of the plurality of extended bit strings, and assigning the plurality of extended bit strings to the plurality of extended bit strings generate a plurality of third confidences that are the confidences of
Decoding by creating a trellis diagram using the coding rate and the plurality of extended bit sequences, and assigning the plurality of third reliabilities to a plurality of branches of the trellis diagram;
A storage medium, characterized in that it causes the wireless communication system to implement: A wireless communication method in a wireless communication system, comprising:
at the transmitting device,
Bit erasure in which the number of branches to which reliability values equal to or less than a specified difference are assigned is less than or equal to a specified number between paths for which the pattern generator compares probabilities in a trellis diagram used in decoding. a first step of generating a pattern;
a second step in which the encoding unit generates a first bit string by convolutionally encoding the information bits based on the encoding rate;
A bit eraser generates a second bit string based on a second number of bits by erasing one or more bits for each first number of bits from the first bit string based on the bit erase pattern. a third step;
a fourth step in which the modulating unit modulates the second bit string using the divided bit strings to generate symbols;
A wireless communication method comprising: A wireless communication method in a wireless communication system, comprising:
at the transmitting device,
In the trellis diagram used in decoding, the pattern generation unit, between the paths for which the probability is compared, the maximum likelihood bit string in the information bit string corresponding to the branch to which the reliability value equal to or less than the specified difference is assigned. a first step of generating a bit erasure pattern in which the number of bits different from the
a second step in which the encoding unit generates a first bit string by convolutionally encoding the information bits based on the encoding rate;
A bit eraser generates a second bit string based on a second number of bits by erasing one or more bits for each first number of bits from the first bit string based on the bit erase pattern. a third step;
a fourth step in which the modulating unit modulates the second bit string using the divided bit strings to generate symbols;
A wireless communication method comprising: A wireless communication method in a wireless communication system, comprising:
at the transmitting device,
Bit erasure in which the number of branches to which reliability values equal to or less than a specified difference are assigned is less than or equal to a specified number between paths for which the pattern generator compares probabilities in a trellis diagram used in decoding. In the trellis diagram used for pattern generation or decoding, the probability is the same or less than the specified difference between the paths for which the probability is compared. a first step of generating a bit erasure pattern in which the number of bits different from the likelihood bit string is equal to or less than a specified number;
a second step in which the encoding unit generates a first bit string by convolutionally encoding the information bits based on the encoding rate;
A bit eraser generates a second bit string based on a second number of bits by erasing one or more bits for each first number of bits from the first bit string based on the bit erase pattern. a third step;
a fourth step in which the modulating unit modulates the second bit string using the divided bit strings to generate symbols;
at the receiving device,
a fifth step in which the demodulator calculates a plurality of first reliabilities, which are reliabilities of a plurality of bit sequences that the symbol can take;
A likelihood synthesizing unit generates a plurality of synthesized bit strings that are bit strings composed of bits of the second number of bits, and a value obtained by synthesizing the reliabilities of the plurality of synthesized bit strings with the plurality of first reliabilities. and generating a plurality of second reliabilities that are reliabilities of the plurality of synthesized bitstreams;
A likelihood replicating unit generates a plurality of extended bit sequences that are bit sequences composed of bits of the first number of bits, and assigns the plurality of replicated second reliabilities to reliabilities of the plurality of extended bit sequences. , a seventh step of generating a plurality of third reliabilities that are reliabilities of the plurality of extension bitstreams;
Eighth, wherein the decoding unit creates a trellis diagram using the coding rate and the plurality of extended bit sequences, and decodes by assigning the plurality of third reliabilities to a plurality of branches of the trellis diagram a step of
A wireless communication method comprising:

Images