KR100400056B1 - Soft decision method of multilevel modulation signal for soft decision Viterbi decoding - Google Patents

Abstract

According to the present invention, soft decision Viterbi decoding is performed through soft decision on each bit constituting a corresponding symbol from I and Q components of a received symbol in an 8PSK / 16QAM / 64QAM high-order modulation communication system using a convolutional code as a channel error correction code. By the following, the present invention relates to a soft decision method of a higher order modulated signal for soft decision Viterbi decoding that can improve coding gain.

Description

Soft decision method of multilevel modulation signal for soft decision Viterbi decoding for soft decision Viterbi decoding

According to the present invention, soft decision Viterbi decoding is performed through soft decision on each bit constituting a corresponding symbol from I and Q components of a received symbol in an 8PSK / 16QAM / 64QAM high-order modulation communication system using a convolutional code as a channel error correction code. By the following, the present invention relates to a soft decision method of a higher order modulated signal for soft decision Viterbi decoding that can improve coding gain.

In recent years, digital communication systems have tended to apply higher-order modulation schemes for high-speed data transmission within a limited frequency band. For example, in European terrestrial digital TV systems, high frequency utilization efficiency is obtained by compressing a large amount of video information and audio information and transmitting signals using various high-order modulations of QPSK, 16QAM, and 64QAM to transmit them using a limited bandwidth. have. In order to transmit a large amount of multimedia information in a limited frequency band in the future, high-order modulation as well as excellent compression on the original information can be considered inevitable. In a digital communication system, an error correction code is used to prevent distortion of information transmitted by multipath fading or impulse noise on a transmission channel. As the error correction code, convolutional codes are used a lot, and convolutional codes show excellent performance against random occurrences of errors. A Viterbi decoder is used to decode the convolutional code, and a value obtained by hard decision or soft decision of each bit constituting a symbol may be input to the Viterbi decoder. The soft decision Viterbi decoder can obtain better coding gain than the hard decision, and the decoding process is performed through branch metric and path metric calculation. Euclidean distance calculation is mainly used as a method of calculating the metric representing the similarity between the received symbol and the codeword occurring at each state transition. In order to apply the Euclidean distance calculation, it is necessary to find the soft decision value for each information bit constituting the received symbol. However, until now, only the soft decision method for BPSK or QPSK modulation has been proposed as the soft decision method for most digital communication systems.

Therefore, in the 8PSK / 16QAM / 64QAM high-order modulation communication system using convolutional code as a channel error correction code, the present invention is performed through soft decision on each information bit constituting the corresponding symbol from the I and Q signal components of the received symbol. It is an object of the present invention to provide a soft decision method for higher order modulated signals for soft decision Viterbi decoding that can improve coding gain by performing decision Viterbi decoding.

The present invention for achieving the above object is an association of each information bit constituting the corresponding symbol from the I, Q components of the received symbol in the 8PSK / 16QAM / 64QAM high-order modulation communication system using a convolutional code as a channel error correction code The soft decision Viterbi decoding is performed through the determination.

The present invention relates to a soft decision method of a received signal for soft decision Viterbi decoding when a convolutional code is used as a channel error correction code in a digital higher order modulation system. Decoding of convolutional code is performed by hard decision or soft decision of the signal. When the signal is soft decision and Viterbi decoding is performed, higher coding gain can be obtained than hard decision. Until now, most communication systems have used the BPSK or QPSK series of modulation schemes, but recently proposed systems for digital broadcasting and wireless multimedia services use higher order modulation schemes. The soft decision method of the received signal for the higher-order modulation method is essential for improving the performance of the system. The branch metric calculation for soft decision Viterbi decoding is performed by obtaining the state transition probability between the codeword value of the branch connected to each state and the received bit string. The state transition probability can be obtained only through soft decision on each bit constituting the symbol, and soft decision on each bit constituting the symbol is performed using the received I and Q signals. In this case, the probability of each state transition can be obtained from the probability distributions of bits '1' and '0' according to the transmission channel. In the case of 8 PSK, the soft decision of the first two bits constituting the symbol is made using only the Q and I signals, respectively, and the soft decision of the other bit uses the phase information consisting of the I and Q signals. In the higher order modulation scheme of 16 QAM or more, odd bits constituting a symbol may be softly determined using an I signal, and even bits may be softly determined using a Q signal. The present invention thus relates to a soft decision method for a received signal necessary for soft decision Viterbi decoding of a higher order modulation system.

1 is a block diagram of a transmitter / receiver of a communication system using an OFDM transmission method in a general higher-order modulated communication system requiring soft decision Viterbi decoding proposed in the present invention.

2 shows 8 PSK symbol mapping for soft decision Viterbi decoding.

3 is a diagram showing the correlation of each bit constituting 8 PSK modulation symbols.

4 shows 16 QAM symbol mapping for soft decision Viterbi decoding.

5 shows 64 QAM symbol mapping for soft decision Viterbi decoding.

6 is a performance comparison characteristic chart for hard decision / soft decision Viterbi decoding of 8 PSK.

7 is a performance comparison characteristic chart for hard decision / soft decision Viterbi decoding of 16 QAM.

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

1 shows a transmitter / receiver of a general high-order modulation communication system using an OFDM transmission scheme. The constraint length and code rate of the convolutional encoder 101 are determined according to the service to be applied and the channel environment, and a series of codewords coded through the convolutional encoder 101 are generated in the transport channel. Pass the bit interleaver 102 to turn the concatenation error into a random error. After the symbol interleaving 103 of the subcarriers constituting one OFDM symbol, the symbol mapper 104 is used to determine I and Q signal values corresponding to each symbol. The pilot signal is inserted 105 by an appropriate method to compensate for signal distortion caused by multipath fading, Doppler effect, impulse noise, etc. occurring in the transmission channel. Finally, the transmitter generates an OFDM signal 106 and 107 by inserting an IFFT 106 and a guard interval 107 and transmits the signal through the antenna via the RF unit. The receiver extracts I and Q signals by removing the guard interval 115 from the OFDM signal and taking the FFT 114. In addition, the pilot signal is used to compensate for distortion in the channel (113) and to remove the pilot signal (112). The description so far is a general signal processing process performed until soft decision in a high-order modulation communication system using an OFDM transmission method. The soft decision 111 for the higher-order modulated signal is made as follows according to each modulation scheme.

As can be seen from the 8PSK symbol mapping 201 of FIG. 2, one symbol is composed of three bits.

3 shows a correlation 303 of each bit constituting a symbol. The bits indicated by the rectangle 301 represent '0', and the bits indicated by the circle 302 mean '1'. Before the soft decision is performed, the probability distribution of the BPSK modulated signal is obtained for the corresponding channel. The soft decision of the first information bit of the symbol is made by obtaining a probability value from which a bit '0' and a bit '1' appear for a value obtained by normalizing the Q signal to -sin ([pi] / 8). In the second information bit, soft decision is made by obtaining a probability value from which a bit '0' and a bit '1' appear from a table with respect to a value obtained by normalizing an I signal with -sin ([pi] / 8). The soft decision of the third information bit is a table of probability values of bit '0' and bit '1' for the normalized values according to the quadrant where the phase value (atan (Q / I)) obtained from the I and Q signals exist. By obtaining from

Iii) 1, 3 quadrants, {(atan (Q / I)-π / 4) / (π / 8)}

ii) quadrants 2 and 4, {(atan (-Q / I)-π / 4) / (π / 8)}

4 shows a symbol mapping 401 of 16QAM, in which the soft decision for each information bit constituting a symbol is obtained by bit '0' and bit '1' for normalized I and Q signal values satisfying the following conditions. This is done by looking at the table for the probability of appearance. The soft decision on the first and second bits is obtained using the normalized values of the I and Q signals as follows.

Iii) first bit, (-I)

ii) second bit, (-Q)

iii) third bit,

I ≥0, {(-1) * (I-2)}

I <0, (I + 2)

iv) fourth bit,

Q ≥0, {(-1) * (Q-2)}

Q <0, (Q + 2)

FIG. 5 shows a symbol mapping 501 of 64QAM, in which the soft decision for each information bit constituting a symbol has bit '0' and bit '1' for normalized I and Q signal values satisfying the following conditions. This is done by looking at the table for the probability of appearance. The soft decision on the first and second bits is obtained using the normalized values of the I and Q signals as follows.

Iii) first bit, (-I)

ii) second bit, (-Q)

iii) third bit,

I ≥0, {(-1) * (I-4)}

I <0, (I + 4)

iv) fourth bit,

Q ≥0, {(-1) * (Q-4)}

Q <0, (Q + 4)

v) fifth bit,

I ≥0,

I <4, (I-2)

I ≥4, {(-1) * (I-6)}

I <0,

I ≥-4, {(-1) * (I + 2)}

I <-4, (I + 6)

vi) sixth bit

Q ≥0,

Q <4, (Q-2)

Q ≥4, {(-1) * (Q-6)}

Q <0,

Q ≥-4, {(-1) * (Q + 2)}

Q <-4, (Q + 6)

As described above, according to the present invention, by performing soft decision Viterbi decoding on a high-order modulation communication system using convolutional code, at least 2 dB or more coding gain improvement effect is shown compared to hard decision as shown in FIGS. 6 and 7. Can be.

Claims (4)

delete In the soft decision method of higher order modulated signal for soft decision Viterbi decoding in 8PSK high order modulation system using convolutional code as channel error correction code, (a) obtaining a probability distribution of the BPSK modulated signal for the corresponding channel; And (b) determining a probability value at which bit '1' and bit '0' can appear by using the probability distribution for each information bit constituting the symbol from the I and Q signals of the received symbol. step (b) The first information bit is soft-determined by the Q signal, and the probability value of bit '0' and bit '1' for the value obtained by normalizing the Q signal to -sin (π / 8) is calculated using the probability distribution. Determining; The second information bit is soft-determined by the I signal, and the probability value of bit '0' and bit '1' for the value obtained by normalizing the I signal to -sin (π / 8) is calculated using the probability distribution. Determining; And The third information bit is a probability value of bit '0' and bit '1' of the normalized quadrant value according to the quadrant where the phase value (atan (Q / I)) obtained from the I and Q signals exist. Soft decision method for soft decision Viterbi decoding, characterized in that it comprises the step of determining using the probability distribution. Ix) 1, 3 quadrants, {(atan (Q / I)-π / 4) / (π / 8)} ii) 2, 4 quadrants, {(atan (-Q / I)-π / 4) / ( π / 8)} In the soft decision method of higher order modulated signal for soft decision Viterbi decoding in 16QAM high order modulation system using convolutional code as channel error correction code, (a) obtaining a probability distribution of the BPSK modulated signal for the corresponding channel; And (b) determining a probability value at which bit '1' and bit '0' can appear by using the probability distribution for each information bit constituting the symbol from the I and Q signals of the received symbol. step (b) Determining the probability value of bit '0' and bit '1' for the normalized I and Q signal values satisfying the corresponding conditions for each information bit constituting the symbol as shown in Table 2 using the probability distribution The soft decision method of the higher-order modulation signal for soft decision Viterbi decoding, comprising a. Iii) first bit, (-I) ii) second bit, (-Q) iii) third bit, I ≥ 0, {(-1) * (I-2)} I <0, (I + 2) iv) 4th bit, Q ≥0, {(-1) * (Q-2)} Q <0, (Q + 2) In the soft decision method of higher order modulated signal for soft decision Viterbi decoding in 64QAM high order modulation system using convolutional code as channel error correction code, (a) obtaining a probability distribution of the BPSK modulated signal for the corresponding channel; And (b) determining a probability value at which bit '1' and bit '0' can appear by using the probability distribution for each information bit constituting the symbol from the I and Q signals of the received symbol. step (b) Determining the probability value of bit '0' and bit '1' for the normalized I and Q signal values satisfying the corresponding conditions for each information bit constituting the symbol as shown in Table 3 using the probability distribution The soft decision method of the higher-order modulation signal for soft decision Viterbi decoding, comprising a. Iii) first bit, (-I) ii) second bit, (-Q) iii) third bit, I ≥ 0, {(-1) * (I-4)} I <0, (I + 4) iv) 4th bit, Q ≥0, {(-1) * (Q-4)} Q <0, (Q + 4) v) 5th bit, I ≥0, I <4, (I-2) I ≥4 , {(-1) * (I-6)} I <0, I ≥-4, {(-1) * (I + 2)} I <-4, (I + 6) vi) 6th bit Q ≥0, Q <4, (Q-2) Q ≥4, {(-1) * (Q-6)} Q <0, Q ≥-4, {(-1) * (Q + 2)} Q <-4, (Q + 6)