JPH10150384A - Spread spectrum communication system - Google Patents

Abstract

PROBLEM TO BE SOLVED: To provide a system enabling number of signal points to increase by decreasing an error rate in the case of data transmission in spread spectrum communication. SOLUTION: A convolution coding part 1 applies error correction coding processing to transmitted data to produce transmission data. A mapping part 2 arranges the transmission data to signal points corresponding to data values. A PN series generating part 3 generates a PN series based on a part of the transmission data. A spread part 4 multiplies a PN series generated by the PN series generating part 3 with an output of the mapping part 2 to make the transmission signal spread. An inverse spread part 5 detects the PN series used at a transmitter side to apply inverse spread to a reception signal by using the detected PN series. A re-mapping part 6 obtains the transmission data from the signal points where the reception signal is arranged. A supplement information processing part 7 obtains supplement information based on the detected PN series. A Viterbi decoding part 8 reproduces transmitted data from the transmission data while utilizing the supplement information.

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

The present invention relates to a spread spectrum communication system, and more particularly, to a spread spectrum communication system using an error correction coding process.

[0002]

2. Description of the Related Art Networking in offices and factories is rapidly spreading, and the number of users connected to the network is increasing, and the amount of data transferred on the network is also increasing. The increase in data amount is usually dealt with by improving the information transmission speed.

[0003] By the way, when transmitting information, usually,
The transmission data is being modulated. That is, the transmitting side modulates and transmits the data, and the receiving side decodes the modulated data. The modulation method is PSK (Phase Shift Keying)
And QAM (quadrature amplitude modulation) are known.

[0004] In the following, a technique using a spread spectrum communication system as a system for transmitting modulated data will be described, and a technique for speeding up information transmission will be described. In the spread spectrum communication, a DS (Direct Spreading) system, an FH (Frequency Hopping System) and the like are known depending on a difference in a spreading system. Here, the DS system will be described.

FIG. 17A shows a general model of a spread spectrum communication system. In the figure,
Only the baseband sections on the transmitting side and the receiving side are shown. On the transmission side, the modulation section 101 modulates information of transmission data. The modulation scheme is PSK, QAM, or the like. Spreading section 102 adds PN to the information-modulated data.
(Pseudorandom Noise). That is, the information modulated data is further spread modulated. The output of spreading section 102 is transmitted on a carrier wave.

On the receiving side, despreading section 111 applies a spreading section 1 to the received signal frequency-converted to the baseband.
02 is multiplied by the same PN sequence as the PN sequence used. That is, despreading section 111 despreads (respreads) the received signal to make it the same state as the output signal of modulating section 101. Demodulation unit 1
Reference numeral 12 reproduces transmission information by demodulating the signal despread by the despreading unit 111.

In order to improve the information transmission speed in the data transmission system, the number of information modulation levels (the number of signal points) is increased. Increasing the number of signal points used for data transmission increases the amount of data that can be transmitted in parallel, so that the information transmission speed can be improved without changing the bandwidth of the transmission path.

[0008] The technique of increasing the information transmission rate by increasing the number of levels of information modulation (the number of signal points) is not applied only to a system using a spread spectrum communication system, but is applied to general data transfer. It is also widely used in

[0009]

However, when the number of levels of information modulation (the number of signal points) is increased, as shown in FIGS. 18A to 18C, the distance between the signal points decreases. The distance between signal points is a measure of “error probability” when reproducing a digital signal, and the smaller the distance, the higher the error rate.

In order to prevent the error rate from deteriorating when the number of signal points is increased, it is necessary to improve the S / N ratio. For example, when comparing the 4PSK method and the 64QAM method, it is necessary to improve the S / N ratio by about 8 dB in order to achieve the same error rate of 10 ^-5 under a white Gaussian noise environment.

As a technique for improving the S / N ratio, a technique using an error correction code is known. The error correction code adds redundancy to enable correction of an error generated in a communication channel when transmitting digital information. FIG. 17B shows an example in which an error correction coding circuit is simply added to the configuration of FIG. 17A. Here, convolutional coding section 103 is provided on the transmission side before modulation section 101, and Viterbi decoding section 113 is provided on the reception side after demodulation section 112. With such a configuration, the error rate during data transmission can be reduced, and the number of signal points can be increased to some extent as compared with the configuration shown in FIG. As a result, the information transmission speed can be improved.

However, in general, when error encoding is performed, the amount of data to be transmitted is increased in order to impart redundancy to transmission data. Therefore, the data transmission rate must be increased accordingly. As a result, the signal band is expanded. On the other hand, the communication band is usually limited in wireless communication. As described above, when the error correction code is used, the bandwidth is expanded due to an increase in the amount of transmission information, but the signal bandwidth is limited, so that the signal is spread like a spread spectrum system. The spreading gain due to this may be limited by signal band spreading by error correction coding.

FIG. 17 (c) is a modification of the configuration of FIG. 17 (b), in which a mapping section 104 is provided in place of the modulation section 101 and a remapping section 114 is provided in place of the demodulation section 112. Configuration. The mapping unit 104
The input signal pattern and the phase plane information used when transmitting the signal of the pattern are stored in association with each other, and the phase plane information is extracted by using the input signal (output of the convolutional coding unit 103) as a key, and the phase is extracted. The transmission data is transmitted based on the plane information. Remapping section 114 performs processing corresponding to mapping section 104 on the receiving side to reproduce transmission data. This configuration is sometimes called a coded modulation scheme.

In the coded modulation system shown in FIG.
The error rate in data transmission can be reduced without expanding the signal band. However, considering the poor communication environment in which multipath fading and the like occur, it cannot be said that resistance to the minimum inter-signal-point distance is high, and there is a limit to improving the information transmission speed by increasing the number of signal points. . The configuration shown in FIG. 17B also has a limit in improving the information transmission speed by increasing the number of signal points in consideration of a poor communication environment in which multipath fading or the like occurs.

The present invention is to solve the above-mentioned problem in the spread spectrum communication system, and provides a system capable of increasing the number of signal points by reducing an error rate in data transmission. Another object of the present invention is to simplify the processing of the decoding device of the error correction coding system.

[0016]

The spread spectrum communication system according to the present invention has a configuration in which data in N bits is arranged and transmitted at 2 ^N signal points.

The following means are provided on the transmitting side. The error correction coding means performs error correction coding processing on transmission data in M (M ≦ N) bits to generate transmission data in N bits. The spreading sequence storing means stores a plurality of spreading sequences. The extracting means extracts a spread sequence from the spread sequence storing means based on at least a part of the N-bit transmission data error-correction-coded by the error correction coding means. The mapping means includes an error-correction-coded N
The bit-wise transmission data is arranged at a signal point corresponding to the data value. The spreading means multiplies the output of the mapping means by the spreading sequence extracted by the extracting means. This spread signal is transmitted to the receiving side.

The receiving unit is provided with the following units. The spreading sequence detecting means detects the spreading sequence used by the spreading means. The remapping unit performs despreading processing on the received signal using the spread sequence detected by the spread sequence detection unit, and generates N-bit data based on a signal obtained as a result. The supplementary information processing means obtains at least a part of the N-bit unit transmission data based on the spread sequence detected by the spread sequence detection means.
The error correction decoding unit executes an error correction decoding process based on the N-bit unit data generated by the remapping unit and at least a part of the N-bit unit transmission data obtained by the supplementary information processing unit. The M-bit transmission data is reproduced.

According to the above arrangement, the transmission side spreads the transmission data using a spreading sequence corresponding to a part of the transmission data, and the reception side detects the spreading sequence.
The receiving side recognizes a part of the transmission data. That is,
A part of the transmission data is transmitted using the spreading sequence. Also, when transmitting transmission data by arranging it at predetermined signal points,
A part of transmission data corresponding to the spread sequence is associated with a signal point used for the data transmission. And
When the encoded transmission data is decoded on the reception side, it is determined at which signal point the transmission data is arranged and transmitted using a part of the transmission data recognized on the reception side. As described above, the process of using a part of the transmission data transmitted using the spread sequence to assist the process of detecting the signal point where the transmission data is arranged is assisted.

[0020]

FIG. 1 is a block diagram illustrating a basic configuration of a spread spectrum communication system according to the present embodiment. FIG. 1 shows only the configuration of the baseband unit.

The convolutional encoder 1 performs error correction encoding on transmission data. Mapping section 2 arranges output data of convolutional encoding section 1 at a signal point corresponding to the output data. The output of the mapping unit 2 is information indicating a predetermined position on the phase plane, and the I-phase data and Q
Consists of phase data. The PN sequence generating unit 3 holds a plurality of types of PN sequences (spreading sequences), selects one PN sequence based on a part of the output of the convolutional coding unit 1, and supplies the selected PN sequence to the spreading unit 4. Spreading section 4 multiplies the output of mapping section 2 by the PN sequence supplied from PN sequence generating section 3 to spread the transmission signal. Although not shown, the signal spread in the spreading section 4 is transmitted on a carrier in a radio frequency band.

On the receiving side, after the carrier component is removed, that is, after the frequency conversion from the radio frequency band to the baseband is performed, the received signal is input to the despreading unit 5. The despreading unit 5 detects a PN sequence used (selected) on the transmitting side. Then, it multiplies the received signal by the detected PN sequence, and transfers the despread signal to the remapping unit 6. The despread signal transferred to the remapping unit 6 is data representing a signal point used for transmitting data. Further, the despreading unit 5 passes the detected PN sequence (or information for identifying the detected PN sequence) to the supplementary information processing unit 7.

The remapping unit 6 performs a process opposite to the process performed by the mapping unit 2 on the output signal of the despreading unit 5. That is, the mapping unit 2 obtains signal points from transmission data, whereas the remapping unit 6 obtains transmission data from signal points. The output of the remapping unit 6 is transferred to the Viterbi decoding unit 8. The supplementary information processing unit 7 obtains supplementary information based on the information received from the despreading unit 5 and transfers the supplementary information to the Viterbi decoding unit 8. The Viterbi decoding unit 8 reproduces (decodes) the transmission data using the supplementary information.

FIG. 2 is a detailed block diagram of the configuration on the transmitting side shown in FIG. In the following embodiment, serial transmission data is converted into 3-bit parallel data and transmitted.

The serial / parallel conversion circuit 11 converts serial transmission data into a 3-bit parallel format. The convolutional encoder 1 includes registers 12-0, 12-1,
12-2, OR circuits 13 and 14, and latch circuit 1
5 by adding redundancy to the transmission data (x1, x2, x3) in the 3-bit parallel format received from the serial / parallel conversion circuit 11, thereby producing the data (y0, y1, y2, y3) is output. Convolutional encoder 1
As an encoding method, for example, Ungabek (Unge
rboeck).

FIG. 3 is a trellis diagram showing the input / output relationship of the convolutional encoder 1 and the transition states of the registers 12-0, 12-1, and 12-2. The three numbers shown in the vertical direction on the left and right sides of the trellis diagram are the states of the registers 12-0, 12-1, and 12-2 (the values held by the registers).
“0, 0, 0” is the value of the registers 12-0, 12-1, 12
-2 indicates a state in which all “0” are held.
Also, in this trellis diagram, four types of lines are used, but the solid line, the wavy line, the two-dot chain line, and the one-dot chain line are input (x1 is a left bit and x2 is a right bit), respectively. 0
The transition in the case of "0""01""10""11" is shown.

3 on the right side of FIG. 3 show the input / output relationship of the convolutional encoder 1 for each state of the registers 12-0, 12-1, and 12-2. Here, the two digits to the left of the slash are the input value (the left bit is x1, the right is x2), and the three digits to the right of the slash are the output value (the left bit is y0, The middle bit is
y1 and the right bit is y2).

Therefore, for example, the registers 12-0,
The states of 12-1 and 12-2 are "1" and "0", respectively.
If the input data to the convolutional encoder 1 is (x1, x2, x3) = (1, 0, 1) when it is “0”, the registers 12-0, 12-1, 12- 2 transitions to “1”, “0”, and “1”, respectively, and (y
0, y1, y2, y3) = (1, 1, 0, 1).

The data (y0, y1, y2, y3) subjected to the error correction coding (convolutional coding) processing is once held in the latch circuit 15 and then output to the mapping unit 2 at a predetermined timing.

FIG. 4A is a diagram showing a signal point arrangement in the mapping unit 2. As shown in FIG. FIG. 1 shows a phase space, and here, signal points are arranged in a 16QAM system. Each signal point is defined by an amplitude and a phase on the phase space. In addition, each signal point is called signal points AP for convenience. Also, in the figure, the leftmost bit of the 4-bit data assigned to each signal point is y3, and the rightmost bit is y0. That is, for example, at the signal point A, (y3, y2, y1, y0) = (1, 0,
(0, 0) is assigned.

The correspondence between the convolutionally encoded transmission data (y3, y2, y1, y0) and the signal point is, for example, as follows:
It is stored in a memory such as a ROM. FIG. 4B is a configuration example of a memory that stores information on signal point arrangement. As shown in the figure, the amplitude and phase of each of the signal points A to P in the phase space can be searched using the key addresses 0000 to 1111. Therefore, for example, the mapping unit 2 outputs (y3, y
When (2, y1, y0) = (0, 0, 0, 0) is received, the amplitude information and the phase information of the signal point K are extracted from the memory, and the data is arranged at the signal point K. That is, the data (y3, y2, y1, y0) = (0, 0, 0, 0) is transmitted with the amplitude and phase defined by the signal point K.

FIG. 5 is a diagram for explaining a procedure for allocating 4-bit data (y3, y2, y1, y0) to each signal point of 16QAM. The least significant bit (equivalent to y0) of the 4-bit data is used to identify two signal points having the smallest distance between signal points. That is, when “0” is assigned as the least significant bit of signal point A, “1” is assigned to the least significant bit of the signal point having the shortest distance from signal point A. That is, signal points B and E
Is assigned "1" to the least significant bit of. When “1” is assigned to the least significant bits of the signal points B and E, “0” is assigned to the least significant bit of the signal point having the shortest distance from those signal points. That is, “0” is assigned to the least significant bit of the signal points C, F, and I.

Similarly, when the least significant bit of the 4-bit data is set at the 16 signal points, the signal points A, C, F, H, I, K, “0” is assigned as the least significant bit of N and P, and “1” is assigned as each least significant bit of signal points B, D, E, G, J, L, M, and O.

The second lowest bit (corresponding to y1) of the 4-bit data is the two signal points having the smallest distance between signal points in each group divided into eight signal points by the least significant bit. Are used to identify each other. That is, the second lowest bit is used to identify two signal points having the second smallest distance between signal points when all signal points are targeted.

For example, in the second stage from the top in FIG. 5, if “0” is assigned as the second bit from the lower end of the signal point A, then each group (larger than 8) is divided into eight signal points by the above-described procedure. (Only the black circles are targeted), the lower two points of the signal point having the smallest distance from signal point A
The "1" is assigned to the th bit. That is, “1” is assigned to the second lowest bit of the signal point F. When “1” is assigned to the second lowest bit of the signal point F, “0” is assigned to the second lowest bit of the signal point having the shortest distance from the signal point F in the group. Can be That is, “0” is assigned to the second lowest bit of signal points C, I, and K.

When the second least significant bit of the 4-bit data is set to 16 signal points in the same procedure, FIG.
Signal points A, C, I, K,
"0" as the second lowest bit of B, D, J, L
Are assigned, and signal points F, H, N, P, E, G, M,
“1” is assigned as the second lower bit of O.

By deciding the values of the third and fourth bits of each signal point by the same assignment method, the arrangement shown in FIG. 4A is obtained. In FIG. 4A, for example, the signal points having the shortest distance from the signal point A are the signal points B and E. When the least significant bits of those signal points are compared, the least significant bit of the signal point A is obtained. Is "0", while the least significant bits of signal points B and E are "1".
In other words, the two least significant bits (y0) of the 4-bit data can identify two signal points having the shortest inter-signal distance.

The signal point whose distance from the signal point A is the second smallest is the signal point F. When the second lowest bit of these signal points is compared, the signal point A is "0".
On the other hand, at the signal point F, it is “1”. That is, two signal points having the second closest signal distance can be distinguished from each other by the second bit (y1) from the lower order of the 4-bit data.

Similarly, the lower three bits of the 4-bit data are
The second and fourth bits enable the two signal points having the third closest signal distance to be distinguished from each other.

Next, the PN sequence generator 3 will be described. PN
The sequence generation unit 3 includes a PN selection unit 16 and a PN storage unit 17
Consists of As shown in FIG. 6, the PN storage unit 17 stores four different PN sequences associated with four types of data “00” to “11” represented by two bits. The PN selector 16 outputs the lower 2 bits (y0, y) in the 4-bit data output from the convolutional encoder 1.
When 1) is received, the PN corresponding to the 2-bit data is
The sequence is extracted from the PN storage unit 17. Then, the PN selecting unit 16 transfers the extracted PN sequence to the spreading unit 4.

The spreading unit 4 outputs a PN to the output of the mapping unit 2.
Multiply by the PN sequence output from the sequence generation unit 3. In addition,
In FIG. 2, two output lines from the mapping unit 2 are:
The spreading unit 4 multiplies each of the I-phase signal and the Q-phase signal by a PN sequence. The signal multiplied and spread by the PN sequence is transmitted on a carrier wave.

FIG. 7 is a diagram showing signals in the device on the transmitting side. Transmission data input in a serial format is converted into a 3-bit parallel format by a serial / parallel conversion circuit 11 and input to the convolutional encoder 1. Since the convolutional encoder 1 adds redundancy to the transmission data, its output is 4-bit parallel data. These data are binary data (digital data).

The output of the convolutional code unit 1 is
Are arranged at predetermined signal points. That is, the output of the mapping unit 2 is information representing the signal point, and is multi-value data that can represent an analog amount. Here, in 16QAM shown in FIG. 4 (a), each signal point is represented by a combination of four real values and four imaginary values on the phase plane, so that the I-phase component and the Q-phase component of the output of the mapping unit 2 are represented. Is quaternary data or multivalued data of four or more values.

The spreading unit 4 outputs P to each output of the mapping unit 2.
The PN sequence selected by the PN selection unit 16 of the N sequence generation unit 3 is multiplied. Therefore, the output of the spreading unit 4 is spread multi-valued data. The output of the spreading section 4 is converted into a radio band, that is, put on a carrier wave and transmitted to the receiving side.

As described above, on the transmitting side of the system of the present embodiment, the transmission data is spread and received using the PN sequence corresponding to the value of a part (lower 2 bits) of the data transmitted in the parallel format. To the side.

Next, the configuration of the receiving side will be described. FIG.
FIG. 4 is a block diagram of a despreading unit 5. Upon receiving the spread signal transmitted on the carrier, the despreading unit 5 first multiplies the received signal by a wave having the same frequency as the carrier to convert the received signal into a baseband. At this time, the I-phase component and the Q-phase component are extracted by multiplying the waves whose phases are shifted by π / 2 from each other.

The signal converted to the baseband is input to matched filters 21-1 to 21-4. The matched filters 21-1 to 21-4 hold PN # 1 to PN # 4 as PN sequences, respectively, and
Multiply by N series. That is, the matched filter 21-1
21-4 despread the received signal using PN # 1-PN # 4, respectively. These PN sequences (PN # 1 to P
N # 4) is the same as the four PN sequences stored in the PN storage unit 17 on the transmission side. Each of the matched filters 21-1 to 21-4 independently performs despreading processing on the I-phase component and the Q-phase component. Each matched filter 21-1 to 21-4 has a filter number (= 1 to 1) as a number for identifying them.
4) is assigned.

The selecting section 22 includes a matched filter 21-1.
21-4 are received, and a matched filter having the largest correlation value is selected from the received outputs. Then, the selection unit 22
Transfers the output signal of the selected matched filter to the remapping 6 and notifies the supplementary information processing unit 7 of a number (filter number) for identifying the selected matched filter. Instead of the filter number, the selection unit 22 may notify the supplementary information processing unit 7 of a PN sequence used by the matched filter having the largest correlation value. In this case, the selection unit 22 determines whether each matched filter 21
-1 to 21-4 recognize the PN sequence used.

According to the above configuration, the PN used on the transmitting side
The sequence can be recognized on the receiving side. For example, assuming that PN # 1 is used as the PN sequence on the transmitting side, the correlation value of the matched filter 21-1 should be maximum on the receiving side. Matched filter 21-1
Is detected to be the maximum, it can be determined that the signal has been spread on the transmitting side using PN # 1.

FIG. 9 is a detailed block diagram of the despreading unit 5. As shown in the figure, the received signal converted to the baseband is input to four A / D converters in the I phase and the Q phase, respectively, where it is converted to digital signals. The output of each A / D converter is input to a digital matched filter DMF. The digital matched filters DMF provided four each in the I phase and the Q phase are provided with a PN sequence (P) held at the output of each A / D converter.
N # 1 to PN # 4).

The output of each digital matched filter DMF is squared. Accordingly, the correlation value (magnitude) in each digital matched filter DMF
Is found.

The sum calculation unit calculates the sum of the I-phase correlation value data and the Q-phase correlation value data for each filter number. For example, in the calculation for the filter number = 1, the sum of the square value of the output of the digital matched filter DMF1 provided for the I phase and the square value of the output of the digital matched filter DMF1 provided for the Q phase is calculated. Ask.

The maximum value detecting section detects the maximum value from the correlation value data for each filter number calculated by the sum calculating section, and notifies the supplementary information processing section 7 of the detected filter number. The sum calculation unit outputs the output of the digital matched filter DMF having the largest correlation value (I sum before calculating the sum).
Phase data and Q phase data) are sent to the remapping unit 6 as despread data.

The remapping unit 6 basically performs the reverse of the process in the mapping unit 2 on the transmitting side. That is, remapping section 6 receives the output signal of despreading section 5 and obtains a corresponding signal point from the I-phase component and the Q-phase component. Also, the remapping unit 6
It has a correspondence table equivalent to the table shown in FIG. 4B, and outputs 4-bit data corresponding to the obtained signal points. For example, if the signal point represented by the I-phase component and the Q-phase component of the received signal is signal point K, remapping section 6 calculates (y3, y2, y1, y0) =
(0,0,0,0) is output.

For example, the remapping unit 6 compares the point on the phase plane represented by the received data with the point 1 shown in FIG.
The distance between each of the six signal points is calculated, and the signal point (the 4-bit data) corresponding to the received data is obtained by detecting the signal point having the shortest calculated distance among them. However, in a poor communication environment such as an environment in which multipath fading occurs, a position of a signal point on a phase plane represented by received data fluctuates, and a correct signal point may not be obtained. In this case, a digital error occurs. The function of correcting such a digital error is a Viterbi decoding process. The Viterbi decoding unit 8 will be described later.

Upon receiving the filter numbers for identifying the matched filters 21-1 to 21-4 from the despreading unit 5, the supplementary information processing unit 7 outputs 2-bit information corresponding to the filter numbers to the Viterbi decoding unit 8. . The details are as follows.

As shown in FIG. 10A, the supplementary information processing section 7 stores output 2-bit information in association with the filter number, reads out the 2-hit information corresponding to the received filter number, and outputs it. I do. For example, assuming that the correlation value of matched filter 21-1 is maximum in despreading section 5, supplementary information processing section 7 receives the filter number "1", and outputs "00" as the 2-bit information. .

Here, it will be described that 2-bit information is transmitted from the transmission side to the reception side by the above configuration. First,
It is assumed that the lower two bits (y0, y1) in the 4-bit data output from the convolutional encoder 1 are "0, 0" on the transmission side. In this case, as shown in FIG. 6, PN # 1 is selected as the PN sequence, and the signal is spread using the selected PN # 1. When the signal is spread by PN # 1 on the transmitting side, the correlation value becomes maximum on the receiving side when the signal is despread by PN # 1. That is, the correlation value of the matched filter 21-1 is maximized. As a result, the supplementary information processing section 7 receives “1” as the filter number and outputs “00” as 2-bit information. This 2-bit information corresponds to the lower 2 bits of the transmitted 4-bit data. Thus, according to the above configuration, PN
Two-bit data “00” is transmitted from the transmission side to the reception side using a sequence.

In the above example, the despreading unit 5 notifies the supplementary information processing unit 7 of the filter number.
When notifying the PN sequence instead of the filter number,
The supplementary information processing section 7 stores a correspondence relationship as shown in FIG.

FIG. 11 is a block diagram of the Viterbi decoder 8. The Viterbi decoding unit 8 is a circuit that detects the maximum likelihood data from the error-correction-coded reception data and reproduces the transmission data.

The branch metric calculating unit 31 receives 4-bit data from the remapping unit 6 and receives supplementary information (2-bit data) from the supplementary information processing unit 7. Then, the distances between the points on the phase plane represented by the received 4-bit data and the 16 signal points shown in FIG. 4A are calculated. However, the branch metric calculation unit 31
Does not calculate distances to all signal points, but only distances to signal points specified by supplementary information (2-bit data).

For example, assume that the supplementary information (2-bit data) is “00”. Here, the 2-bit data received as supplementary information is the lower 2 bits (y0, y1) of the 4-bit data output from the convolutional encoder 1. Therefore, if the supplementary information is "00", the 4-bit data is the lower 2 bits out of the 16 points shown in FIG.
This means that the signal has been transmitted using a signal point whose bit is “00”. That is, the 4-bit data has been arranged and transmitted at any of the signal points A, C, I, and K. Therefore, the branch metric calculation unit 31 calculates
When the supplementary information is “00”, only the distances between the signal points represented by the 4-bit data received from the remapping unit 6 and the signal points A, C, I, and K are calculated.

The path metric memory 32 stores the ACS unit 3
3 holds the path calculated and selected and the path metric data (length data of the path). AC
The S (Add & Compare & Select) unit 33 outputs the path metric data extracted from the path metric memory 32 (that is, the path metric data obtained by the previous calculation).
And the distance between each signal point calculated by the branch metric calculation unit 31 are obtained. Also, the ACS unit 33
Holds path metric data obtained in advance for a path that can occur in a certain state. Then, the ACS unit 33 obtains a Euclidean distance or a Hamming distance by comparing each sum data with path metric data obtained in advance. The path with the minimum Euclidean distance or Hamming distance is defined as the maximum likelihood path (the most likely path). This comparison processing is shown in FIG.
This will be described with reference to FIG.

FIG. 12 (a) is a state diagram according to the conventional method, and FIG. 12 (b) is a state diagram according to the present embodiment. Here, how to read in FIG. 12 will be described. FIG. 12 (a) and
In (b), eight small circles arranged in the vertical direction indicate
Each represents one state. Specifically, it corresponds to the states of the registers 12-0 to 12-2 shown in FIG. 2, and sequentially from the top, "000", "001", "010",. . . , "1
11 ". That is, for example, the uppermost state is a state in which the registers 12-0, 12-1, and 12-2 all hold “0”. The state transition is shown in FIG.
Follow the trellis diagram described in.

In FIG. 12 (a) or (b), three digits of four numbers for each state correspond to the lower three bits (y0, y1, y2) in the transmitted 4-bit data. I do.
In the three-digit number, the left bit corresponds to y0, the middle bit corresponds to y1, and the right bit corresponds to y2, as in the description of FIG. Each three-digit number designates a path (path) from one state to another state. Four 3-digit numbers are arranged horizontally in each state, and the leftmost 3-digit number specifies the top path among the four output paths. . Similarly, from the left 2, 3, 4
The third three-digit number designates the second, third, and fourth paths from the top of the four output paths, respectively.

For example, at time T2, the register 12
It is assumed that the state of −0 to 12-2 is “000”. In this state, if the lower three bits of the 4-bit data are (y0, y1, y2) = (0, 1, 0), the state of the registers 12-0 to 12-2 at time T3 is "010".
Transitions to. The path indicated by a bold line in FIG. 12 (a) or (b) corresponds to the path from time T1 to time T5.
Lower 3 bits of bit data (y0, y1, y2)
Are generated in the order of (111, 010, 001, 010).

The operation of the ACS unit 33 at the time T3 will be described. At time T3, the path metric memory 3
2 stores the path metric data of the path leading to the state at time T3. Here, it is assumed that the path indicated by the thick line is selected as the maximum likelihood path in the processing at the time T2. The ACS unit 33 reads the path metric data of this path from the path metric memory 32. Also, A
The CS unit 33 receives, from the branch metric calculation unit 31, distance data between the signal point represented by the received 4-bit data and each signal point of 16QAM. As described above, since the branch metric calculation unit 31 calculates only the distance to the signal point specified by the supplementary information (2-bit data), the ACS unit 33 receives four pieces of distance data between signal points. Then, the ACS unit 33 adds the distance data between each signal point to the path metric data read from the path metric memory 32. After this, ACS
The unit 33 compares the path metric data of one or more paths that can occur in the state transition from the time T3 to the time T4 with the data obtained by the above-described addition operation, and selects the path having the smallest Euclidean distance.

By the way, as shown in FIG. 12, there are four paths that transition from one state to another state.
For example, in the state “0, 0, 1” at time T3,
(Y0, y1, y2) = (001), (101), (01)
1) Four paths specified by (111) can occur.

However, the ACS unit 33 does not use all of these four paths for comparison, but only the path specified by the supplementary information notified from the supplementary information processing unit 7. For example, at time T3, the register 12-
When “00” is received as supplementary information when the states of 0 to 12-2 are “0, 0, 1”, the value of the lower 2 bits of the 4-bit data is (y0, y1) = (0 , 0), and a path designated by a numeral whose left two bits are both “0” among the three-digit numbers described in four for each state on the left side of FIG. Only as a control. That is, in this case, (y0, y1, y2) =
The path specified by (0,0,1), that is, the state “0,0,1” at time T3 to the state “1,0” at time T4
The path to "0,0" is selected for comparison.

Thereafter, the ACS unit 33 compares the path metric data of the path to be compared with the above-mentioned four addition operation data, and obtains the Euclidean distance. Then, the path having the minimum Euclidean distance is set as the maximum likelihood path. The maximum likelihood path thus obtained is written to the path memory 34, and the path metric data of the path is written to the path metric memory 32.

The output of the path memory 34 is the result of decoding. That is, in the convolutional encoding process, the relation between one data and one pass is determined by the encoding process method, so that if the path is obtained, decoded data determined by the pass is obtained. . This procedure is the same as a general Viterbi decoding process.

In the conventional Viterbi decoding process, as shown in FIG. 12 (a), a maximum likelihood path is detected using all possible paths as comparisons. On the other hand, in the present embodiment, only the path specified by the supplementary information is targeted. That is, the path indicated by the broken line in FIG. Therefore, the processing time is greatly reduced.

FIG. 13 is a flowchart showing the operation of the Viterbi decoder 8. In step S1, 4-bit data is received from the remapping unit 6, and in step S2, supplementary information is received from the supplementary information processing unit 7. In step S3, one signal point is extracted from the six signal points of 16QAM. In step S4, it is determined whether or not the signal point extracted in step S3 is a signal point to be subjected to distance calculation with reference to the supplementary information.

If it is a signal point to be subjected to distance calculation, in step S5, the distance between the signal point represented by the 4-bit data received in step S1 and the signal point extracted in step S3 is calculated. And step S
In step 6, it is checked whether or not the above processing has been performed for all signal points. If any signal points have not been processed, the process returns to step S3. On the other hand, if it is determined in step S4 that the signal point is not a signal point at which distance calculation is to be performed, step S4 is performed.
The process proceeds to step S6 without executing step 5.

By the steps S3 to S6, one or more (four in the embodiment) signal point distance data is obtained. In step S7, the path metric data of the path up to the present time is read from the path metric memory, and the distance data calculated in step S5 is added to the path metric data. As a result, one or more (four in the embodiment) path metric data can be obtained.

In step S8, one of the paths that can occur when moving from one state to the next state is extracted. In the embodiment, four paths can be generated in one state, and one path is selected from the four paths. In step S9,
Referring to the supplementary information received in step S2, step S2
It is checked whether the path extracted in step 8 is a comparison path. In the example shown in FIG. 12 (b), the path for comparison is indicated by a solid line, and the path not to be compared is indicated by a dashed line.

Step S10 is a process performed in normal Viterbi decoding. The added value data calculated in step S7 is compared with the path metric data of the path extracted in step S8, and the Euclidean distance (convolution code In the case of conversion, the Hamming distance is obtained. Then, the one with the shortest distance is selected. Step S
If it is determined in step 9 that it is not a comparison, the processing in step S10 is not performed. In step S11,
It is checked whether or not the above-described processing has been performed for all possible paths. If there is an unprocessed path, the process returns to step S8.

In step S12, the maximum likelihood path is determined. When step S10 is performed a plurality of times in the loop processing of steps S8 to S11, a path having the smallest Euclidean distance is selected among them. Then, the maximum likelihood path is written to the path memory, and the path metric data is written to the path metric memory.

In step S13, decoded data is obtained from the maximum likelihood path. The process of obtaining the decoded data from the path is the reverse procedure of the encoding algorithm, and the original transmission data can be obtained from the relationship shown in FIG.

In the Viterbi decoding process of the present embodiment,
Since the distance between signal points is obtained only for the signal point specified by the supplementary information, step S in the above flowchart is performed.
5 can be reduced. For this reason, the number of times of the addition process in step S7 and the number of comparison processes in step S10 are reduced. Further, since the Euclidean distance is obtained only for the path specified by the supplementary information, the number of executions of step S10 is further reduced.

Next, with reference to FIG. 14, a description will be given of how the error rate is reduced by the method of the present embodiment. Here, (y3, y2, y1, y0) = (1, 0, 0, 0)
A case where bit data is arranged at a signal point A and transmitted will be described as an example.

The data arranged and transmitted at signal point A on the transmitting side is ideally detected on signal point A also on the receiving side. However, if the communication environment is bad and the transmission signal point fluctuates, it is detected at a position shifted from the signal point A. Here, it is assumed that the transmission data is detected at the position indicated by the triangle in FIG.

In this case, in the conventional method, there is a possibility that the transmission point is recognized as having been transmitted with the transmission data arranged at the signal point closest to the mark, that is, the signal point E. Therefore, on the receiving side, the transmission data is (1,
(1, 1, 1).

However, in the method of the present embodiment, the PN sequence used on the transmitting side is detected on the receiving side, so that
The lower two bits (y1, y0) of the bit data are (0,
0), the transmission data is at signal points A,
It is presumed that it has been placed and transmitted in any one of C, I, and K. Therefore, it is determined that the transmission data has been arranged at the signal point closest to the mark, that is, at signal point A among these four signal points.
For this reason, on the receiving side, even if the transmission data is detected at the position indicated by the triangle, the transmission data is (1, 0, 0,
0) can be determined.

In general, if the distance between signal points is large, the error rate becomes low. However, according to the method of the present embodiment, only specific signal points are substantially reduced on the receiving side by supplementary information transmitted using the PN sequence. Signal points, the distance between signal points is virtually increased, and digital errors are reduced. For example, when the distance between signal points is shorter than the distance in a white Gaussian noise environment, 3 to ⁵ dB is obtained as an S / N ratio improvement value for achieving an error rate of 10 ^-5 .

FIG. 15 is a block diagram of another embodiment. FIG. 1 shows an example in which 4-bit parallel data is transmitted by a spread spectrum method. On the transmitting side, the upper two bits of the 4-bit data are input to the mapping unit 41, and are arranged at signal points determined by the two-bit values (00, 01, 10, 11). In order to assign one signal point to each value of 2-bit data, four signal points may be provided, and the mapping method is 4PSK.
And

The lower 2 bits of the 4-bit data are:
It is input to the PN sequence generation unit 42 and its 2-bit value (0
0, 01, 10, 11) are generated. PN sequence generating section 42 is the same as PN sequence generating section 1 shown in FIG.
One of PN # 4 is output.

Spreading section 43 applies the PN generated by PN sequence generating section 42 to the signal mapped with the upper two bits.
Multiply and spread the sequence. On the receiving side, the despreading unit 44 despreads the received signal. The despreading unit 44 is the same as the despreading unit 5 shown in FIG. 1, transfers the output of the matched filter having the highest correlation value to the remapping unit 45, and converts the PN sequence used by the matched filter into the PN inverse conversion unit. Transfer to 46.

The remapping unit 45 performs a decoding process based on 4PSK and reproduces the upper 2 bits of the 4-bit parallel data. The PN inverse conversion section 46 performs processing in the opposite direction to the processing in the PN sequence generation section 42,
The lower 2 bits of 4-bit parallel data are reproduced from the PN sequence obtained in step 4.

As described above, the signal is spread on the transmitting side using the PN sequence corresponding to a part of the transmission data (here, the lower 2 bits), and the PN sequence is detected on the receiving side. The receiving side recognizes a part of the transmission data. This is substantially equivalent to transmitting a part of the transmission data using the PN sequence. For this reason, the amount of data (here, only the upper 2 bits) transmitted at the signal point is reduced, and the number of signal points is reduced (1
(6 points → 4 points), the distance between signal points can be increased, and a reduction in bit errors can be expected.

FIG. 16 is a block diagram of a configuration in which the amount of data transmitted using a PN sequence is increased as compared with the configuration of FIG. FIG. 1 shows an example in which 5-bit parallel data is transmitted by a spread spectrum method.

In the configuration shown in FIG. 16, the lower 3 bits of the 5-bit data are input to the PN sequence generator, where the PN sequence corresponding to the value of the 3-bit data is converted to the PN #.
1 to PN # 8 and output. The despreading unit on the receiving side has eight matched filters, and the PN
The PN sequence used on the transmitting side is recognized by multiplying each of # 1 to PN # 8 and detecting that the correlation value is maximized. In the above configuration, the mapping unit and the remapping unit are the same as those shown in FIG.
This is a 4PSK mapping method.

As described above, when the amount of data to be transmitted increases (4 bits → 5 bits), if the amount of data to be transmitted using the PN sequence is increased accordingly (2 bits → 3 bits), the signal point arrangement The distance between signal points can be increased as in the case of FIG. 14 without changing the system, and a reduction in bit errors can be expected. In other words, by increasing the amount of data to be transmitted using the PN sequence, the data transmission speed can be improved while maintaining the transmission quality. In the above-described embodiment, the direct spread system, which is a typical system among the spread spectrum communication systems, has been described. However, the present invention can be applied to other systems. For example, the present invention can be applied to a frequency hopping scheme.

[0094]

In a spread spectrum communication system,
When arranging data at a predetermined signal point and transmitting the data, the signal is spread using a PN sequence corresponding to a part of the data,
Since the receiving side recognizes a part of the transmission data by detecting the used PN sequence, the error rate at the time of reproducing the data from the signal point of the received signal is reduced.

In the error correction encoding process, since a part of the transmission data to be reproduced is known, the decoding process is simplified.

[Brief description of the drawings]

FIG. 1 is a block diagram illustrating a basic configuration of a spread spectrum communication system according to an embodiment.

FIG. 2 is a detailed block diagram of a configuration on a transmission side shown in FIG. 1;

FIG. 3 is a trellis diagram showing an input / output relationship of a convolutional code unit and a transition state of each register.

FIG. 4 is a diagram showing a signal point arrangement in a mapping unit.

FIG. 5 is a diagram illustrating a procedure for allocating 4-bit data to each signal point of 16QAM.

FIG. 6 is a diagram schematically showing a PN storage unit.

FIG. 7 is a diagram illustrating signals in a device on the transmission side.

FIG. 8 is a block diagram of a despreading unit.

FIG. 9 is a detailed block diagram of a despreading unit.

FIG. 10 is an example of information stored by a supplementary information processing unit.

FIG. 11 is a block diagram of a Viterbi decoding unit.

12A and 12B are diagrams illustrating state transitions and paths, where FIG. 12A is a conventional method and FIG. 12B is a method of the present embodiment.

FIG. 13 is a flowchart illustrating a Viterbi decoding process.

FIG. 14 is a diagram illustrating the effect of reducing the error rate.

FIG. 15 is a block diagram (part 1) of another embodiment.

FIG. 16 is a block diagram (part 2) of another embodiment.

17A is a diagram showing a general model of a spread spectrum communication system, and FIGS. 17B and 17C are block diagrams of a system in which the S / N ratio is improved by a conventional method. is there.

18 (a) to (c) show 4PSK and 16PS, respectively.
It is a figure which shows the signal point arrangement | sequence example of K and 16QAM.

[Explanation of symbols]

 Reference Signs List 1 convolutional code unit 2, 41 mapping unit 3, 42 PN sequence generation unit 4, 43 spreading unit 5, 44 despreading unit 6, 45 remapping unit 7 supplementary information processing unit 8 Viterbi decoding unit 12-0 to 12-2 Register 16 PN selection unit 17 PN storage unit 21-1 to 21-4 Matched filter 22 Selection unit 31 Branch metric calculation unit 32 Path metric memory 33 ACS (Add & Compare & Select) unit 34 Path memory 46 PN despreading unit

Claims (8)

[Claims] 1. A spread spectrum communication system in which data in N bits is arranged at 2 ^N signal points and transmitted, wherein an error correction coding process is performed on transmission data in M (M ≦ N) bits. Error correction coding means for generating transmission data in units of N bits, spreading sequence storage means storing a plurality of spreading sequences, and transmission data in units of N bits error-corrected and coded by the error correction coding means. Extracting means for extracting a spread sequence from the spread sequence storage means based on at least a part thereof; mapping means for arranging the error correction coded N-bit transmission data at a signal point corresponding to the data value; A spreading unit for multiplying the output of the mapping unit by the spreading sequence extracted by the extracting unit; and a spreading unit for detecting the spreading sequence used by the spreading unit. Column detecting means; remapping means for performing despreading processing on the received signal using the spread sequence detected by the spread sequence detecting means, and generating N-bit unit data based on the resulting signal; A supplementary information processing means for obtaining at least a part of the N-bit transmission data based on the spread sequence detected by the spread sequence detection means; an N-bit unit data generated by the remapping means; Error correction decoding means for performing error correction decoding processing based on at least a part of the transmission data in N bits obtained by the supplementary information processing means to reproduce the transmission data in M bits, Spread spectrum communication system provided. 2. The spread sequence detecting means multiplies a received signal by a plurality of spread sequences stored in the spread sequence storage means and examines a spread sequence having the highest correlation among the received signals to use the spread sequence. The spread spectrum communication system according to claim 1, wherein the detected spread sequence is detected. 3. When associating each of the 2 ^N signal points with data in N-bit units, two signal points having the smallest distance between signal points are identified from each other by using a first bit of N bits. And the second, third,. . . , N, the signal point distances are second, third,. . . , The N-th smallest signal point is distinguished from each other, and when selecting bits for extracting a spread sequence from the N-bit transmission data, predetermined bits are sequentially assigned from the first bit. The spread spectrum communication system according to claim 1, wherein the number is selected. 4. When associating each of the 2 ^N signal points with data in units of N bits, two signal points having the smallest distance between signal points are identified from each other by using a first bit of the N bits. 2. The spread spectrum communication system according to claim 1, wherein the first bit is included in at least a part of the transmission data in the unit of N bits. 5. A transmitting apparatus comprising: a spreading sequence generating means for generating a spreading sequence based on at least a part of transmission data; and a spreading means for multiplying the transmission data by the spreading sequence generated by the spreading sequence generating means. A spreading sequence detecting means for detecting a spreading sequence used by the spreading means; a supplementary information processing means for obtaining at least a part of the transmission data based on the spreading sequence detected by the spreading sequence detecting means; Data obtained by despreading processing using the spread sequence detected by the sequence detection means, and reproduction means for reproducing transmission data based on at least a part of the transmission data obtained by the supplementary information processing means. Spread spectrum communication system provided on the receiving side. 6. A spread spectrum communication system for transferring data comprising a first segment and a second segment, comprising: a transmitting side generating a spread sequence for generating a spread sequence based on the first segment. Means, and spreading means for spreading the second segment using the spreading sequence generated by the spreading sequence generating means, and detecting means for detecting the spreading sequence used by the spreading means on the receiving side; Calculating means for obtaining the first segment based on the spreading sequence detected by the detecting means; despreading means for obtaining the second segment by despreading processing using the spreading sequence detected by the detecting means; Spread spectrum communication system provided with. 7. A spread-spectrum transmitter used in a communication system for arranging N-bit data at 2 ^N signal points and transmitting the data, wherein M (M ≦ N) -bit transmission data has an error. Error correction coding means for performing correction coding processing to generate transmission data in N-bit units, spread sequence storage means storing a plurality of spread sequences, and N error-correction coded by the error correction coding means. Extracting means for extracting a spreading sequence from the spreading sequence storage means based on at least a part of the bit-unit transmission data; and transmitting the error-correction-coded N-bit transmission data to a signal point corresponding to the data value. Spread spectrum transmission comprising: mapping means for mapping; and spreading means for multiplying the output of the mapping means by the spreading sequence extracted by the extracting means and outputting the result. Vessel. 8. A transmission device performs error correction encoding on transmission data in M bits to generate transmission data in N (M ≦ N) bits, and divides the transmission data in N bits into 2 ^N transmission data. A spread-spectrum receiver used in a communication system for arranging and transmitting signals at signal points, comprising: a spread-sequence detection unit for detecting a spread sequence used in a transmission device of the communication system; Remapping means for performing a despreading process on a received signal using the spread sequence thus generated, and generating N-bit data based on the resulting signal; and a spread sequence detected by the spread sequence detection means. Supplementary information processing means for obtaining at least a part of the transmission data in N-bit units based on Error correction for performing error correction decoding processing on the basis of at least a part of the N-bit transmission data obtained by the supplemental information processing means and reproducing the M-bit transmission data. A spread spectrum receiver comprising: decoding means.