JPH0697748B2 - Error correction coding method - Google Patents

Description

Description: TECHNICAL FIELD The present invention relates to an error correction coding method for performing data communication on an extremely bad transmission line such as a fading propagation line in mobile communication.

(Prior Art) As a typical communication path that generates a burst error,
There is a fading channel. As a method for correcting the burst error in this communication channel, Nikkei Electronics (1975,1
2-15, pp. 48-52) There is a method to correct code errors by interleaving random error correction codes and randomizing them as described in detail in "Error correction codes that have been actively used in a wide range of fields". . FIG. 2 is a block diagram showing a configuration example of error correction coding / decoding using conventional interleaving and deinterleaving, in which random error correction is performed. In the figure, 11 is an encoder, 12 is an interleave circuit, 13 is a deinterleave circuit, 14 is a decoder, and 15 is a communication path. The transmission data is encoded by the encoder 11 and interleaved by the interleave 12. The mixed transmission data is a communication path.
It is sent to the deinterleave circuit 13 via 15. At this time,
The transmitted data causes a bursty error on the communication path 15, but after passing through the deinterleave circuit 13, it becomes a random error and is error-corrected by the decoder 14. Therefore, a random error correction code is applied to the error correction code applied to the encoding / decoding.

FIG. 3 shows the configuration of a conventional encoder and decoder that have performed burst error correction. In the figure, 21 is a coder with burst error correction, 22 is a modulator, 23 is a decoder with burst error correction code, 24 is a demodulator, and 25 is a communication path. Since the communication path 25 is a fading communication path, as an application of a code to mobile radio, if a reception power level is high and a fading pitch is 10 to 40 Hz, a 4-bit length burst error correction code is applied to eliminate an error in the communication path 25. Can be corrected.

FIG. 4 is a flow chart of the error correction method when the above-mentioned conventional error correction coding method and the data retransmission method using the error detection code are used.

First, input data is divided into blocks of a predetermined bit length, and a block number (K = 1, 2, ..., KM) is designated for each block.
The data K1 of block number 1 is input and encoded with the error correction code A (step 301). The encoded data is modulated by the modulator 22 (step 302), and the fading channel 2
After passing through 5 (step 303), demodulation is performed by the demodulator 24 on the receiving side (step 304). The demodulated signal is an error correction decoder 23
And is decoded (step 305). Fading channel 25
The erroneous result is corrected by the decoder 23,
If the error cannot be corrected sufficiently, a NAK signal requesting data retransmission is transmitted (step 307), the NAK signal is encoded with the error correction code B (step 308), modulated (step 309), and transmitted. Send back to. On the transmitting side, the NAK signal is demodulated (step 310), decoded (step 311), and recognized as a NAK signal (step 31).
4) Retransmit the data K1 according to the same procedure.

On the other hand, if the error correction code has been corrected without error, the data is accumulated on the receiving side (step 312) and the ACK signal for requesting the data next to the data K1 is transmitted (step 31).
3) encode the ACK signal with the error correction code B (step
308), modulate (step 309) and send back to the sender. The transmitting side demodulates the modulated ACK signal (step 31
0), the data is decoded (step 311), and when it is recognized as an ACK signal (step 314), 1 is added to the block number to update the data by setting K = K + 1 (step 31).
Five). The updated data K2 is transmitted according to the above procedure.
The block signal K of the transmission data becomes the final number KM and the transmission is terminated (step 315).

(Problems to be solved by the invention) Even when any of the above-mentioned methods is used, the mobile radio causes a large change from the situation in which the mobile body is stopped to a speed of several hundred km / h and the reception level is about -100 dBm. − As the burst length changes greatly from several bits to several hundreds bits as it changes to several dBm, interleaving of several hundreds of bits is required for randomization by interleaving, and a delay time occurs. Therefore, the capacity of the random access memory (RAM) becomes large and the throughput efficiency deteriorates. In addition, the throughput efficiency is extremely degraded in order to communicate without errors using only a burst length of several bits and an error correction code.

The present invention has been made in view of such drawbacks, and even if the reception level changes from several hundred dB / h to a few hundred dB per hour even when the moving body is stopped, the throughput efficiency is improved even if the reception level changes from several low dB to several dB. It is intended to provide an excellent error correction coding method capable of performing data communication with little deterioration.

(Means for Solving Problems) According to the present invention, a transmitting side divides transmission data into blocks having a certain number of bits, encodes each block or a plurality of blocks using an error correction code, transmits the blocks, and receives the blocks. On the side, decoding is performed using an error correction code, error detection is performed, and if error correction is possible and there is no error in the decoded data, an ACK signal indicating the transmission request of the next block is returned to the transmission side, and error correction is performed. If it is impossible and there is an error in the decoded data, a NAK signal indicating a resend request is returned to the transmitting side, and if the transmitting side receives an ACK signal, the next untransmitted block or blocks are transmitted, The present invention relates to an error correction coding method for resending one or a plurality of blocks at present when a NAK signal is received.

Then, in the present invention, a plurality of types of error correction codes are prepared on the transmitting side, the number of corrected bit errors at the time of decoding is counted on the receiving side, and the state of the communication path is determined according to the counting result. One error correction code that conforms to the above is selected, and the selection information indicating the selected error correction code is returned to the transmission side together with the NAK signal or the ACK signal.

In addition, the transmitting side encodes each block or plural blocks, which are currently or not yet transmitted, by using one error correction code according to the selection information and transmits them.

(Operation) The transmission data is input by the transmission terminal device, encoded by the encoder, and transmitted to the communication path through the modulator. At this time, the coded transmission signal is subjected to an error correction code suitable for the line state of the communication path and sent to the communication path. Coded data for performing this error correction coding is stored in advance in the error correction code storage device, and is selected from among them by using return information from the receiving side.

On the receiving side, the received signal is demodulated and decoded, the decoded data is input to the error bit number counting circuit, and the error bit is calculated. This calculation result is input to the error correction code selector,
Convert the number that selects the error correction code into data. The data is encoded and modulated, returned to the transmitting side, and used as data for selecting an error correction code to be applied to the next transmission signal.

(Embodiment) FIG. 1 is a block diagram showing an embodiment of the present invention. The input data from the transmission terminal device 401 is error correction coded by the encoder 402. As the error correction code to be encoded, error correction codes E1 to E4 that meet each condition of the communication path are stored in advance in the error correction code storage device 403. For example, the error correction code used first is E2. The data encoded by the error correction code E2 becomes a modulated signal by the modulator 404 and is sent to the communication path 405. The communication path 405 is a fading communication path, and a bursty error occurs in the modulated signal as the mobile body travels.
The modulated signal passing through the communication path 405 is demodulated by the demodulator 406 and decoded by the decoder 407. The error bit number counting circuit 408 calculates the error bit number from the result of this decoding.
The error detection here uses an error detection method using a cyclic code or the like.

Error correction code selector 409 based on the calculation result of the number of error bits
The correction code is selected according to. The error correction code at this time is a corresponding code from the error correction codes E1 to E4 stored in the error correction code storage device 403, for example, E3.
Is selected, the number (added to the ACK signal or the NAK signal) corresponding to the error correction code E3 is encoded by the encoder 410 as data, and is modulated by the modulator 411. The modulated signal is a communication path 412.
Then, the signal is demodulated by the demodulator 413 on the transmitting side and sent to the decoder 414. The decoder 414 detects the number of the error correction code selected on the receiving side, and the error correction code selector 415 selects the error correction code E3 from the error correction code storage device 403 according to the number of the error correction code and codes the code. Send to the chemical converter 402. Therefore, the next data from the transmission terminal device 401 is encoded with the error correction code E3 selected by the error correction code selector 415 and transmitted.

FIG. 5 shows the transition of each stage of the error correction codes E1 to E4. The state of the error correction code E1 is that when the number of bits in one frame is k bits (for example, 1024 bits), nk bits (n is a positive integer) are one block, there is no error correction bit, and there is no error detection parity. It is a correction code to which only check bits are added. The error correction code E2 is a correction code in which when the number of bits in one frame is k bits, k bits are one block, and there are no error correction bits and only parity check bits are added. The error correction code E3 is an error correction code having an information rate of 1/2 by applying an error correction code with k bits as one block as described above. The error correction code E4 has k bits as one block and is subjected to error correction code, and is in a state of reliably transmitting data even if the coding rate is deteriorated to a fraction by a majority logic decoding method or the like.

The error correction codes E1 to E4 are previously stored in the error correction code storage device 40.
It is stored in 3 and selected according to the line status. Specifically, if the error occurring in the communication path 405 was corrected using the error correction code E2, if the state of the communication path 405 and the speed of the vehicle are substantially constant, the error correction code E2 does not change, Transmission is continued using the code.

When the state of the communication path 405 deteriorates, the number of error bits increases, and the state shifts to the error correction code E3 according to the information from the receiving side. On the contrary, when the state of the communication path 405 is improved, the state is changed to the error correction code E1 according to the information from the receiving side.

FIG. 6 is a flow chart of error correction. The procedure of error correction will be described according to the flow of the figure.

K indicates a block number of transmission data. The data of the first block is called by the transmission terminal device 401 (step 501).
Next, one of the error correction codes E1 to E4 is selected (step
502). The error-corrected signal (step 503) is sent to the modulator 40.
Modulated by 4 (step 504) demodulator 40 via communication line
Demodulate with 6 (step 506). The demodulated signal is decoded by the error correction decoder 407 (step 507). The error detection is performed together with the error correction decoding, and it is detected whether the error correction has been completed (the decoded data has no error) or whether only the error has been detected (the decoded data has an error) (step
508). Further, the error bit number counting circuit 408 counts the number of bit errors corrected at the time of decoding. If it is uncorrectable and there is an error in the decoded data, the number of corrected bit errors also exceeds the correction limit, so the error correction code selector 409
Then, the error correction code having the next higher error correction capability is selected (step 509). NAC depends on the information of this error correction code.
A signal is generated (step 512), the NAC signal is encoded by the error correction encoder 410 (step 513), modulated by the modulator 411 (step 514), and then demodulated via the communication line 412.
The data is demodulated by 413 (step 516) and decoded by the decoder 414 (step 517). ACK signal or NAK from the decoding result
The signal is identified (step 518), and if it is an ACK signal, the block number K is updated by 1 and the content of the next block is read (step 519). In the case of the NAK signal, the current block is read and encoded without updating the block number K. Next, as a result of error detection on the receiving side, step 50
8) If there is no error, store the data (step 52)
0). At this time, if the correction bit error number Eb counted by the error bit number counting circuit 408 is equal to or higher than the predetermined rate Kb, the error correction code selector 409 selects the current error correction code, and the predetermined rate Kb.
If it is smaller, the error correction code having the next lower error correction capability is selected. An ACK signal is generated according to the result of this error correction code (step 521), and the ACK signal is encoded by the error correction encoder 410 (step 513). When all the blocked input data is sent, that is, the maximum block
The transmission stops when the KM is reached.

FIG. 7 shows the throughput efficiency with respect to the bit error rate. Car phones and mobile phones are relatively active when they are stopped. For ease of explanation, communication path 412
When the car is stopped and the car is 20 km to 100 k / h
We recommend assuming that the vehicle is running at about m.

Let ts be the time when the vehicle is stopped and communicating and tm be the time that the vehicle is traveling and communicating, and the ratio of each is 7 to 3 with respect to the total communication time. Communication line error rate when the vehicle is stopped and running is Ps
Pm. According to the experimental data, Ps = 0 and Pm = 10 ⁻³ . Under the above conditions, the throughput efficiency can be approximated by the equation (1).

Here, tt is the sum of communication time and is shown by the sum of ts and tm. Ks / Ns
And Km / Nm are coding rates for error correction. Ks / Ns = 1 because no error occurs when the vehicle is stopped, and Km / Nm = 0.5 because the error is corrected when the vehicle is running. Nm is the number of information bits and Km is the number of control bits, and Nm = 100 and Km = 50.
The result obtained according to the equation (1) is shown by the solid line in FIG. 7, and the result when selective retransmission is performed by applying only a single error correction code is shown by the broken line. As is clear from the figure, the throughput efficiency is greatly improved.

(Effect of the Invention) As described above, according to the present invention, in the communication of the mobile body, the wireless line constantly changes due to the influence of fading. This is an extremely effective error correction technique when performing data communication using a line that changes drastically in this way, and it is expected that throughput will be improved because the optimum error correction code is selected according to the state of the communication path. .

BRIEF DESCRIPTION OF THE DRAWINGS FIG. 1 is a block diagram of an error correction coding method according to the present invention, FIG. 2 is a block diagram of a conventional interleave error correction method, and FIG. 3 is a conventional burst error correction method. FIG. 4 is a block diagram, FIG. 4 is a flowchart showing a conventional error correction procedure, and FIG. 5 is a state transition diagram of the error correction code of the present invention.
FIG. 6 is a flowchart showing the error correction procedure of the present invention,
FIG. 7 is an explanatory diagram of throughput efficiency. 11 ... Encoder, 12 ... Interleave circuit, 13 ... Deinterleave circuit, 14 ... Decoder, 15,25 ... Communication path, 21 ...
... Encoder, 22 ... Modulator, 23 ... Decoder, 24 ... Demodulator, 401 ... Transmitting terminal device, 402 ... Encoder, 403 ...
Error correction code storage device, 404 Modulator, 405, 412
Communication channel, 406 ... Demodulator, 407 ... Decoder, 408 ... Error bit number counting circuit, 409 ... Error correction code selector, 410
... Encoder, 411 ... Modulator, 413 ... Demodulator, 414 ...
... Decoder, 415 ... Error correction code selector, 416 ... Receiving terminal device.

Front page continued (72) Inventor Haruhiko Akiyama 1-7-12 Toranomon, Minato-ku, Tokyo Oki Electric Industry Co., Ltd. (72) Inventor Manabu Kawabe 1-7-12 Toranomon, Minato-ku, Tokyo Oki Electric Industry Co., Ltd. In-house (56) References JP 56-169952 (JP, A) JP 58-146012 (JP, A) JP 59-149442 (JP, A)

Claims (1)

[Claims] 1. A transmission side divides transmission data into blocks of a fixed number of bits, each block or a plurality of blocks is encoded using an error correction code and transmitted, and a reception side uses the error correction code. When the error is correctable and there is no error in the decoded data, an ACK signal indicating the transmission request of the next block is returned to the transmitting side, and error correction is impossible and the decoded data is erroneous. If there is, a NAK signal indicating the resend request is sent back to the sender, and the sender sends A
In the error correction coding method, when the CK signal is received, the next untransmitted one or more blocks are transmitted, and when the NAK signal is received, the current one or more blocks are retransmitted. An error correction code is prepared, the receiving side counts the number of corrected bit errors at the time of decoding, and selects one of the error correction codes suitable for the state of the communication path according to the counting result. The information indicating the selected error correction code is returned to the transmission side together with the NAK signal or the ACK signal, and the transmission side transmits the present or untransmitted block, or each of the plurality of blocks, to the selection. An error correction coding method, characterized in that it is coded and transmitted using one of the error correction codes corresponding to information.