JPH077942B2 - Code reception system - Google Patents

Description

TECHNICAL FIELD The present invention relates to code reception by diversity reception such as spatial diversity and transmission diversity in which a transmitting side performs block coding of a code and transmits the code, and a receiving side receives a plurality of them. More specifically, the present invention relates to a code reception method for decoding a code in a transmission method for improving reliability by diversity reception when code transmission is performed in an environment where the reception level varies, such as land mobile radio.

“Prior Art” In a conventional mobile communication system, for example, improvement in reliability due to diversity and improvement in reliability due to error correction are often handled independently. For this reason, the information indicating the reliability of each bit unit obtained when performing diversity reception has not been utilized at the stage of decoding the error correction code. That is, as shown in FIG. 1, the conventional code receiving method by spatial diversity reception is provided with a plurality of receiving antennas ₁₁ to 1 _{M which} are spaced apart from each other, and each code is transmitted from the transmitting side as an error correction block code. The transmitted signal is received by these receiving antennas 1 ₁ to 1 _M , and the received signal is demodulated by the receivers 2 ₁ to 2 _M ,
Each of its demodulated output is input to the selection circuit 11, the maximum receiving level that is selected according to the selection result, the maximum output is switched by the switching circuit 12 from the demodulated output of the receiver 2 ₁ to 2 _M It is taken out and supplied to the error correction code decoder 13, where the error correction code is performed, and the code output is output to the output terminal 14. Each signal antenna 1 ₁ to 1 _M and corresponding receiver 2 ₁ to 2 _M
The M reception systems of and are called a reception branch.

According to this space diversity reception method, the demodulation result in the reception branch with the highest reception level is always selected from among the plurality of reception branches, and error correction decoding is performed on the selected reception sequence, so reception of fading, etc. The deterioration of the transmission reliability due to the level fluctuation is improved.

However, in this system, the bit-by-bit decoding performed under the condition that the reception level is maximum and the error correction code decoding are performed completely independently. Therefore, the information on the probability of selection at the time of diversity reception includes error correction. It has not been used to decode the code. Therefore, when an error occurs in the receiving branch with the maximum receiving level, the error of that bit affects the decoding of the error correcting code even if the other receiving branch is able to receive correctly, and the error of the error correcting code When an error that exceeds the correction capability occurs, the frame is not received or erroneously received.

On the other hand, in the decoding of block codes, a soft-decision decoding method using level information at the time of bit-by-bit decoding is described in [D.
Chase, “A Class OF Algorithms for Decoding Block C
odos With Channel Measurement Information ", IEEE Tr
ans.IT-18, No. 1, Jan, 1972], and it is shown that the error correction capability of the block code can be expanded by this method.

in this way, The code word Xj that becomes is the decoded output. here, Respectively.

The expression (1) can be qualitatively explained as follows: That is, in the decoding method of the block code, weights (bits different in the code word and the received word) depending on the reliability of the judgment are It is considered that the codeword that minimizes the distance between the received word and the codeword weighted by the level value li) is transmitted.

With this decoding method, the error correction capability of the error correction code can be expanded. In this decoding method, the reliability information (level value) obtained at the time of bit-by-bit decoding is brought into the decoding process of the error correction code, and the reliability information is used to calculate the inter-code distance between the code word and the received word. Is characterized by defining. As a result, the decoding on a bit-by-bit basis and the decoding of the error correction code are performed integrally, and the error correction capability can be expanded.

"Means for Solving Problems" An object of the present invention is to expand a range in which a block code can be correctly decoded by integrally performing a bit-by-bit composite and a block code decoding when diversity reception is performed. It is to provide the code receiving method.

According to the present invention, bit-by-bit decoding is performed by diversity reception, at that time, information indicating the certainty of decoding of each bit, that is, reliability information is created, and the reliability information is used for the bit-by-bit decoding result. This is introduced in the block code decoding process. In this way, the decodable range is expanded as compared with the conventional code receiving system in which the diversity reception and the decoding of the block code for the result are performed completely independently.

The code receiving system of the present invention will be described in more detail below. In the present invention, first, bit-by-bit decoding is performed by diversity reception. Bit-wise decoding is By. In the equation (2), when M receive branch diversity is performed, the decoding result of each bit in each receive branch is used as the receive word Zi, and the transmit side independently provides each receive branch. Is a decoding method in which the transmission word having a smaller distance between the reception word weighted by the reception level of each reception branch and the transmission word is used as the decoding result.

That is, when the received word Zi is different from the transmitted word Dji due to a transmission error, ZiDji becomes 1. Therefore, in the equation (2), the Dj having the smaller sum of the reception levels ri of the branches in which the transmission error occurs is the transmitted word. Make a decision. This is the reception level
When ri is large, it is generally difficult for bit errors to occur. Therefore, it is unnatural that bit errors occur even if the reception level ri is large, and such a case is rejected.

This operation corresponds to the well-known post-two-branch detection selection diversity when M = 2.

Bit-by-bit decoding is performed as described above, and block code decoding is performed on the obtained decoded bit example.
Decoding this block code By. The formula (3) does not represent the soft decision decoding method of the block code, and has a form in which | fi (0) −fi (1) | is substituted in place of the level value li of each diet. ing. This expression (3) is a code that minimizes the sum of | fi (0) -fi (1) | when YiXji = 1, that is, a transmission error (corresponding bits in the received word and the transmitted word differ). Word X
ij is the decoding result. | fi (0) −fi (1)
│ is the sum fi (0) of the reception levels rk in the receiving branch that is received as “1” when “0” is transmitted and the receiving branch that is received as “0” when transmitting “1” Reception level at
It is the absolute value of the difference from the sum rk of fi (1), which becomes small when the reception level becomes small and bit errors occur randomly. From the above, Expression (3) means that the case where a bit error occurs despite the high reception level is rejected as unnatural. Also, | fi (0) -fi (1) | is the i-th
It can also be expressed as information indicating the certainty of diversity reception in digit decoding, and in that sense, equation (3) represents this | f which represents the certainty of diversity reception.
Value of i (0) -fi (1) | (hereinafter referred to as reliability index)
3 shows a decoding algorithm in which a code word that minimizes the inter-code distance between the received word and the code word, which is weighted with, is the decoding result. As described above, the code receiving system of the present invention is characterized by integrating both decodings by using the reliability index obtained by decoding in bit units by diversity reception in the decoding process of the error correction block code. Is.

"Embodiment" FIG. 2 shows an embodiment of the present invention, in which parts corresponding to those in FIG. In this embodiment, the demodulated outputs of the receivers 2 ₁ to 2 _M are bit-by-bit decoders 3 ₁
To 3 _M , and each bit is decoded into “0” or “1”, and these decoding results (received words) Z _{1 to} Z _M are supplied to the diversity bit decoder 15. The reception level r ₁ ~r _M respectively detected by the receiver 2 ₁ to 2 _M is also supplied from the terminal 4 ₁ to 4 _M to diversity bit decoder 15. The diversity bit combiner 15 performs bit-by-bit decoding from the input received words Z _{1 to} Z _M , the reception levels r _{1 to} r _M, and the transmitted words D ₀ and D ₁ by the operation of the equation (2). . The decoding result (reception digit) Yi is supplied to the block code decoder 16. The diversity bit decoder 15 also calculates the reliability index | fi (0) -fi (1) | for each decoding digit Yi, and the reliability index is supplied to the block code decoder 16. The block code decoder 16 inputs these inputs Yi, | fi
The block code is decoded by calculating the equation (2) from (0) -fi (1) | and the code word Xj.

Next, the operation of this embodiment will be described. For simplicity, the block code is a (7,4) Hamming code, which is a kind of error correction block code, and the number of diversity branches is M = 3, and the transmission word T = (1000101) is transmitted from the transmission side. It is assumed that the receivers 1 ₁ , 1 ₂ , 1 ₃ on the receiving side receive at reception levels r ₁ , r ₂ , r ₃ as shown by curves 5 ₁ , 5 ₂ , 5 ₃ in FIG. ₃ , respectively. The reception level at the time of decoding bit by bit for the first to seventh digits in each reception branch is a curve.
From 5 ₁ , 5 ₂ , 5 ₃ Becomes In addition, the bit-wise decoding result of each reception branch (output of the bit-unit decoders 3 ₁ , 3 ₂ , 3 ₃ ) is assumed to be reception branch 1: Z ₁ = (1001001) reception branch 2: Z ₂ = (1101100) Receiving branch 3: Z ₃ = (0000100). First, it is shown that the conventional decoding method shown in FIG. 1 results in erroneous reception. Since the demodulation output of the receiving branch having the maximum receiving level among the receiving branches 1 to 3 in the switching circuit 12 is the decoded bit, the received word becomes Z = (1001100), and the exclusive OR of this and the transmitted word T Is ZT = (0001001), and a 2-bit error has occurred. Humming (7,
4) The code distance between codes is 3, and the error correction capability is 1 bit. Therefore, when this Z is decoded, it is erroneously received.

Next, it is shown that the code receiving system according to the present invention can correctly receive. First, on the assumption that "0" or "1" is transmitted at the i-th digit on the transmitting side, the values on the left side of equation (2), fi (0) and fi (1), are calculated, and "0" Assumption: fi (0) = (34 4 0 26 28 0 22) Assuming "1": Received to "0" By the same procedure as for the branch, we obtain fi (1) = (3 42 40 12 5 84 28). With this value, bit-wise compounding is performed according to equation (2). That is, for digits, with fi (0)
The smaller transmission word D ₀ = 0 or D ₁ = 1 with fi (1) is taken to obtain the decoding result Yi for each digit. As a result,
The received word is Y = (1001100), and this Y is the output Z of the switching circuit 12 according to the conventional method.
Therefore, a 2-bit error has occurred. Next, the reliability index is calculated. | fi (0) −fi
The value of (1) | for the first to seventh digits is | fi (0) −fi (1) | = (31 38 40 14 23 84 6). For simplicity, these values can be When quantized into a binary value by the following function, the sequence of F {| fi (0) -fi (1) |} becomes F {| fi (0) -fi (1) |} = (1110110). The above process is also shown in FIG.
The codeword of the (7,4) Hamming code is 1 as shown in FIG.
There are six, and when the left side of equation (3) is calculated for these, the values shown in the right column in FIG. 4 are obtained, and when X ₁ = (1000101), Therefore, it can be seen that this is the minimum and this code word x ₁ is the transmission word T shown above. On the other hand, (7,4)
It is one of the code words of the Hamming code.

Considering X ₉ = (1001110), YX ₉ = (0000010), resulting in a 1-bit error, and a codeword X ₉ that is different from the transmission word T and is erroneous as a codeword that can be decoded by error correction processing. become. However, in this invention, the codeword X ₉
The result of the operation on the left side of equation (3) for Therefore, this is not the minimum, and this X ₉ is not the decoding result.

[Advantages of the Invention] As described above, according to the present invention, it is possible to correctly decode a received word that is not received or is erroneously received by the conventional method. An efficient algorithm for minimizing the equations (2) and (3) is described in detail in the above-mentioned documents and the like, and this method can be easily realized. That is, the code receiving system of the present invention has an advantage that the block code decodable range can be expanded by modifying the software for executing the conventional decoding algorithm.

In the above-mentioned embodiment, the spatial diversity is taken up, and it is assumed that the demodulation is performed by a plurality of receivers.However, for example, on the transmission side, the transmission diversity for transmitting the same code a plurality of times at timings separated in time is used. If you do
Only one receiver is required, and it suffices that the decoding output of each bit of the multiple transmissions corresponds to the combined output Zi of each receiving branch. Also, as the weight of equation (3), | fi (0) −fi
(1) Instead of | Similar effects can be obtained by using values such as.

Further, it will be easily understood that the present invention is effective not only when applied to the error correction block code but also when receiving code for a simple block code which is not error-correctable. In the diversity bit decoder 15, the transmission word D ₀ = {00 ...
0}, D ₁ = {11 ... 1}, the transmission word with the smaller sum of the reception levels of the reception branches that do not match is taken as the bit decoding result Yi, but the transmission word D ₀ = {00 ... 0}, D _The bit decoding result Yi may be the transmission word having the larger sum of the reception levels of the reception branches that match ₁ = {11 ... 1}.

Similarly, in the above description, the code word Xj that minimizes the sum of the reliability indexes at the different bits between the received word Y and the code word Xj is determined to be the decoded word, but the reliability at the matching bit between the received word Y and the code word Xj is determined. The code word Xj that maximizes the sum of the frequency indices may be determined as the decoded word.

[Brief description of drawings]

FIG. 1 is a block diagram showing a conventional diversity reception decoding system, FIG. 2 is a block diagram showing an embodiment of the present invention, and FIG. 3 is a reception level of each receiver and decoding of each digit in bit units. Result, reception level of each digit, fi
FIG. 4 is a diagram showing (0), fi (1) and the like, and FIG. 4 is a diagram showing a (7,4) Hamming code word and a calculation result of the left side of Expression (3). 1 ₁ to 1 _M : Antenna, 2 ₁ to 2 _M : Receiver, 3 _{1 to} 3 _M : Bit-based decoder, 4 ₁ to 4 _M : Detection reception level output terminal, 14: Output terminal, 15: Diversity bit decoding , 16: block code decoder.

Claims (1)

[Claims] 1. A code receiving system in which a code is block-coded and transmitted from a transmitting side and diversity receiving is performed by a plurality of receiving branches at a receiving side, and bit-by-bit decoding means for decoding bit by bit for each branch, and each of them. Level detection means for detecting the reception level in bit-by-bit decoding for each branch, and the sum of the reception branch reception levels matching “1” in the corresponding bit decoding output of each reception branch by the bit-unit decoding means, "1" or "0", which is the larger of the sum of the reception levels of the reception branches that match 0 ", is set as the decoding result Yi, or the reception level of the reception branch that does not match" 1 "in the corresponding bit decoding output of the reception branch. Diversity of the decoding result Yi is the smaller one of the sum of the received branch and the sum of the received branches that do not match "0". Difference between the received word and each possible code word of the block code by using the sum information of the reception levels corresponding to “1” and “0” for each bit unit as a diversity reliability index. Or a block code decoding means for determining that a codeword having a minimum or maximum sum of reliability indices in coincidence bits is transmitted.