JPS61135234A - System for controlling soft deciding threshold value - Google Patents

Abstract

PURPOSE:To always ensure the actuation under the approximately optimum conditions for correction of errors by estimating a ratio between the demodulated signal power and the noise power and controlling the level, etc. of the demodulated signal based on the estimated ratio and the prescribed relation so that the soft deciding threshold value is set at an optimum level. CONSTITUTION:A demodulator 1 demodulates the reception signal obtained by adding the noise to the transmission signal modulated with a symbol code train received the convolutional coding. the demodulated signal is supplied to a bitabi decoder 3 via a variable attenuator 10 and an A/D converter 2 and quantized by plural soft threshold values to be decoded. A detector 7 detects 6 an error factor from the signal obtained by recoding the digital signal of the decoder 3 with its error corrected by a convolutional coder 4 equal to that at the transmission side and the signal that delayed 5 the soft deciding value of the converter 2. Then the estimated value signal for the ratio between the demodulated signal power and the noise power is applied to a ROM8 of a threshold value control circuit 11. The ROM8 stores the prescribed control amount corresponding to the designated value, and the corresponding control amount is read out by the signal sent from the detector 7. The attenuator 10 is controlled the A/D-converted 9 signal to set the soft threshold value at an optimum level.

Description

[Detailed Description of the Invention] [Field of Industrial Application] The present invention relates to a soft-decision threshold control method, and in particular to a soft-decision Viterbi method that uses soft-decision technology to decode digital information that has been convolutionally encoded and transmitted. The present invention relates to a soft-decision threshold control method for optimally controlling soft-decision threshold intervals in a decoding device.

[Conventional technology]

Soft-decision technology is a transmission system that transmits digital information using a noisy transmission path, and when estimating transmitted digital information from a noisy received signal, extracts information on the probability of the estimated value, and uses this information. Piterbi decoding for convolutional codes is a typical decoding method that uses soft decisions to improve the error rate. FIG. 2 is a block diagram showing an example of the configuration of a conventional Viterbi decoding device using soft decisions. A/D converter 2
．． It consists of a Viterbi decoder 3. A received signal 100 with added noise is demodulated by a demodulator 1, and a demodulated signal 101 is generated.
is quantized by the A/D converter 2 into a soft-decision value of qpi) ((l is usually 2 or 3), and the Viterbi decoder 3 uses this quantized soft-decision value to calculate the path metric in the decoding process. Figure 3 is an explanatory diagram showing the relationship between the demodulated signal when q = 3, the soft decision threshold level of the A/D converter 2, and the quantized output.
is quantized into 3-bit codes as shown in the figure corresponding to the eight regions divided by each soft decision threshold indicated by the thick line,
The corresponding symbol metric is used to calculate the bus metric. For each soft-decision threshold interval, assume that the demodulated signal E8hao (Ea is the information signal power per symbol and No is the one-sided noise power tight button) is determined in advance from the conditions of the transmission line to be used. )3+s/No, which is the value that gives the best error rate.

[Problems to be solved by the invention]

However, since the actual Es/No varies over time depending on the environmental conditions of the transmission line, it cannot be said that the fixed soft-decision threshold as described above is always operating in an optimal state, and error correction The problem is that their abilities are not being fully utilized. An object of the present invention is to solve the above-mentioned problems of the nine conventional methods, and to adjust the interval of soft decision thresholds to K s/
An object of the present invention is to provide a soft-decision threshold control method that controls according to No.

[Means for solving problems]

The soft-decision threshold control method of the present invention demodulates a received signal in which noise is added to a transmitted signal that has been convolutionally encoded or modulated with a Pol code sequence, and performs multiple soft decisions on the demodulated signal. In a soft-decision Viterbi decoding device that quantizes by a threshold value and decodes with a Viterbi decoder, the ratio of signal power to noise power (Es/ No or Es/No
: where Eb is the signal power per bit of digital information), and from this estimated value, the magnitude of the demodulated signal or the voltage of the soft decision threshold is determined by the soft decision threshold according to a predetermined relationship. It is controlled and configured to achieve the optimum value.

〔Example〕

Next, the present invention will be described in detail with reference to the drawings.

FIG. 1 is a block diagram of an embodiment of the present invention, in which, in addition to the conventional demodulator 1, A/D converter 2, and Viterbi decoder 3 shown in the second factor, a convolutional encoder 4, a delay The detection circuit 7 includes a memory 5 and an error rate detector 6, and a threshold control circuit 11 includes a ROM 8, a D/A converter 9, and a variable attenuator 10. In FIG. 1, the detection circuit 7
The error-corrected digital information of the Viterbi decoder 3 is re-encoded by the convolutional encoder 4, which is the same as that on the transmitting side, and the first soft decision value of the A/D converter 2 before Viterbi decoding is re-encoded.
The bit error rate (BER) of the received symbol code without correction is adjusted by matching the phase with the bit error rate (BER) of the received symbol code (MSB) through the delay memory 5, and comparing and checking with the error rate detector 5.
) is detected. 1 for error-corrected received digital information.
If there is a bit error, the symbol code obtained by re-encoding it with a convolutional encoder has errors concentrated over the KV symbol interval (K is the constraint length and 1/v is the coding rate) compared to the transmission symbol code. There is a difference. Therefore, the BER obtained by the detection circuit 7 is slightly worse than the BIR of the actual received symbol code, but using this BKR, the K of the demodulated signal is
It is possible to obtain a/N o. Figure 4 shows the coding rate 1
/2, Es/No vs. BER of optimal quantified code with constraint length 7
FIG. 3 is a characteristic diagram showing the characteristics, where the solid line A shows the BER of two-phase PSK synchronous detection without error correction, the solid line shows the BER of 13-tia bit soft-decision Viterbi decoding, and the broken line shows the BER of Ca 2-bit soft-decision Viterbi decoding. As is clear from FIG. 4, for the same Es/No, the improvement in BER due to entertainment correction is very large except for the portion where Es/No is less than O dB. Therefore, the B'ER of the detection circuit 7 detected by comparing the symbol code obtained by re-encoding the output of the Viterbi decoder 3 with the output of the A/D converter 2 is the same as the BER of the received symbol code and tl. , From this, E of the demodulated signal is determined from Fig. 4.
#1 s/No can be estimated almost without error.

Furthermore, the optimal value Topt of the soft decision threshold interval T and E
The relationship between s/N o has been theoretically and experimentally investigated and clarified (for example, IEICE technical research report: C380-126, p.31-p.36). Figure 5 shows the relationship between B a/N o and the optimal soft decision threshold interval T opt that optimizes the BER for the same sign as in Figure 4, and the solid line indicates the 3-bit soft decision. case,
A broken line E shows the case of 2-bit soft decision, normalized by 5, which is the transmission code amplitude. If the BER of the received symbol code is known from this FIG. 5 and the above-mentioned FIG. 4, the optimal soft-decision threshold interval T opt can be found.

As shown in FIG. 1, if a variable attenuator 10 is inserted between the demodulator 1 and the A/D converter 2 to control the amount of attenuation, it is equivalent to changing the interval T of the soft decision threshold. Because there is, RO
If the BER of the received symbol code and the control amount of the variable attenuator 10 for setting Topt are stored in M8 in correspondence, even if Es/No changes, the output of the detection circuit 7
The variable attenuator 10 can be controlled via the A converter 9 so that the soft decision threshold interval is always optimal, and the BIR of the decoded digital information can always be the best. By the way, in the case of the above-mentioned coding rate 1/2 and constraint length 7.3-bit soft decision, the optimal soft decision threshold intervals are o, zs, r5- when Es/No is 5 dB from Fig. 5. However, the ratio of the BER when the soft decision threshold interval is fixed to this value and when it is controlled to the optimal value is 'Ets/No.
= OdB and Ks/N.

, , , 73 dB, it is calculated to be approximately L5 times and 3 times, respectively (Electronic Communication Academic Journal: VOL
, J67-B, A8. (See pages 930-931).

In the embodiment described above, the detection circuit 7 detects the Viterbi decoder 3
0 BIR of received symbol code from input information and output information
Estimate B s/N o of the demodulated signal by finding Es/N
We have explained the method of controlling the variable attenuator according to the relationship between E tx/N o and the optimal soft-decision threshold interval.
The method of detecting is not limited to the embodiments, but for example, a method of determining from the magnitude relationship of path metrics in the decoding process of a Viterbi decoder, a method of determining from temporal changes (IEICE technical research report: C382-43, 917 - p. 24) are known, and these methods may also be used. Furthermore, the configuration of the threshold control circuit that controls the magnitude of the demodulated signal is not limited to the embodiment, and may be configured to directly control the soft-decision threshold voltage instead of the magnitude of the demodulated signal. Good too.

〔Effect of the invention〕

As explained in detail above, according to the soft-decision threshold control method of the present invention, even if E a/N o of the received signal changes, the soft-decision threshold interval can always be set to the optimal value. This has the effect of maximizing the error correction function.

[Brief explanation of drawings]

FIG. 1 is a block diagram of an embodiment of the present invention, FIG. 2 is a block diagram showing the configuration of a conventional soft-decision Viterbi decoding device, and FIG.
The figure is an explanatory diagram of the soft decision threshold, and Fig. 4 shows the demodulated signal E s
A characteristic diagram showing the relationship between /N o and BER, Figure 5 is E
FIG. 3 is a characteristic diagram showing the relationship between s/N o and standardized optimal threshold interval. 1... Demodulator, 2... A/D converter,
3... Viterbi decoder, 4... Convolutional encoder, 5... Error rate detector, 6...
Delay memory, 7...Detection circuit, 8...---
ROM, 9...D/A converter, 10...
- Variable attenuator, 11...Threshold control circuit. Figure 1 2I\l Grass 3I! I

Claims (1)

[Claims] A soft method that demodulates the received signal, which is a transmitted signal modulated with a convolutionally encoded symbol code string with noise added to it, quantizes the demodulated signal using multiple soft-decision thresholds, and decodes it with a Viterbi decoder. In the decision Viterbi decoding device, the ratio of the signal power to the noise power of the demodulated signal is estimated from the information of the Viterbi decoder, and the magnitude of the demodulated signal or the soft decision threshold is determined from this estimated value according to a predetermined relationship. A soft decision threshold control method characterized in that a voltage of a value is controlled so that the soft decision threshold becomes an optimal value.