JP2015109610A - Code error correction capability analyzer - Google Patents

Abstract

PROBLEM TO BE SOLVED: To execute analysis of an error correction capability by outputting a coding gain while using error correction coding data to which return light noise is added.SOLUTION: A code error correction capability analyzer includes: first code error rate calculation means including means for generating an error correction coding block, generating first output signal light on which return light noise is superposed, and calculating a first code error rate property; second code error rate calculation means including means for generating second output signal light in which return light noise is superposed on an information block, and calculating a second code error rate property; and means for calculating an output signal light intensity for obtaining a predetermined error rate in the first code error rate property, calculating an output signal light intensity for obtaining a predetermined error rate in the second code error rate property, and calculating a differential between the output signal light intensities.

Description

  The present invention relates to a code error correction capability analyzer mainly used in the communication field, and more specifically to a code error correction capability analyzer for signal light including return light.

  The code error detection and correction technique is a technique for detecting and correcting a code error when it occurs in data, and is widely used in the fields of storage devices and digital communication. As a method of code error correction, a redundant bit is added to an information bit for transmission, and when an error occurs in some bits during transmission, the redundant bit is used to correct an erroneous bit on the receiving side. There is forward error correction (FEC). As types of FEC, there are block coding in which information bits are divided into blocks for each block, and convolutional coding in which the code of each block depends on the previous block. Typical examples of codes used for block coding include Reed-Solomon codes and BHC codes, while typical examples of codes used for convolutional coding include Viterbi codes.

  In particular, RS codes are complicated in code generation, but have high correction capability. Therefore, error correction of terrestrial digital broadcasting, satellite communications, ADSL, and CD, DVD, QR code (registered trademark), etc. are familiar. Has been applied.

  On the other hand, semiconductor lasers used as light sources for communication devices and storage devices may generate output fluctuations called “return light noise”. The return light noise is an output fluctuation that occurs when a part of the laser light emitted from the laser is reflected and returns to the laser again, and the laser transmission becomes unstable. This reflected light is different from the light with the same phase inside the laser, the stimulated emission is confused and the oscillation becomes unstable because the light is out of phase.

  In an optical modulator using a semiconductor laser, in order to avoid return light noise, the return light is blocked by an optical isolator, or a laser having high resistance to return light is used. An optical isolator is an optical element that passes light only in one direction and cuts light in the opposite direction. Optical isolators are expensive and require a certain amount of space when mounted in an optical modulator. Therefore, optical isolators are generally desired to be eliminated in order to reduce the size and price of optical modulators. In the long-distance communication field, it is difficult to remove the optical isolator.

J. Endo, K. Asaka, A. Kanda, T. Ito, M. Yoneyama, N. Ikeda, K. Kenji and M. Urano, "Isolator-free EA-DFB module with Forward Error Correction", OECC2012

  At present, a method has been proposed in which FEC is mounted on a semiconductor laser, and the output fluctuation due to return optical noise is reduced by borrowing FEC correction capability, and the optical isolator is excluded from the optical modulator. FIG. 9 is a flowchart showing a method for calculating error correction characteristics (BER) in a system in which FEC is implemented.

  In step 901, the information block data is subjected to error correction coding to generate a coded block.

  In step 902, the encoded block is converted to a binary number to create digital modulation data.

  In step 903, the laser light is directly modulated using the digital modulation data as a modulation signal to generate signal light, which is output from the optical transmitter.

  In step 904, the output signal light from the optical transmitter is received, 0 or 1 is discriminated and a reproduced digital signal is generated.

  In step 905, the error of the reproduced digital signal is corrected and decoded.

  In step 906, a code error rate characteristic (BER) is calculated based on the decoded data.

  Here, the code error rate is calculated without error correction (uncoded), and compared with the code error rate in the encoding calculated in 6 above, the coding gain is output, and the code error correction is performed. Capabilities can also be analyzed.

  On the other hand, FIG. 10 is a flowchart showing a method of generating output signal light in an isolator-less semiconductor laser. When the output signal light of 903 is generated, direct modulation is performed, but the output light of the semiconductor laser in the presence of the return light can be analyzed by an analysis algorithm to generate the output signal light from the modulation data. In particular, when a rate equation is used for the analysis algorithm, an accurate light intensity can be calculated from a temporal change in light, and an accurate output signal light can be generated.

  Conventionally, when analyzing error characteristics, a method for generating output signal light by solving an analysis algorithm for output light of a semiconductor laser in the presence of return light has not been applied. Furthermore, an apparatus for decoding the signal light obtained by solving the analysis algorithm for the output light including the return light and analyzing the error correction capability has not been established. Has not been strictly verified.

  In order to achieve such an object, a first aspect of the present invention is a code error correction capability analyzing apparatus for analyzing an error correction capability of a transmission signal including return optical noise of a semiconductor laser. Error rate characteristic calculating means for encoding an information block and generating an encoded block; and using the generated encoded block as modulation data and returning the output light of the semiconductor laser to which optical noise is superimposed. Means for generating one output signal light, means for discriminating and reproducing the first output signal light and generating a first digital reproduction signal, decoding the first digital reproduction signal, and decoding block Means for generating, and means for calculating a first code error rate characteristic by comparing the encoded block and the decoded block; and a second code error rate characteristic calculating means, The information block Means for generating a second output signal light in which optical noise is superimposed on the output light of the semiconductor laser, and the second output signal light is discriminated and reproduced to produce a second digital reproduction signal. And means for comparing the information block with the second digital reproduction signal to calculate a second code error rate characteristic. In the calculated first code error rate characteristic, an output signal light intensity for obtaining a predetermined error rate is calculated, and in the second code error rate characteristic calculated by the second code error rate characteristic calculation unit, a predetermined error is obtained. Means for calculating an output signal light intensity for obtaining a rate, and calculating a difference between the output signal light intensities.

According to a second aspect of the present invention, the means for encoding the information block of the error correction capability analyzer of the first aspect is a means for defining an expanded Galois field composed of 2 ^m elements, where m is an integer. And means for converting the information block into a polynomial and performing a block coding operation.

  Further, according to a third aspect of the present invention, the means for generating the decoding block of the error correction capability analysis apparatus according to the first aspect detects an error value, corrects the error, and decodes a decoded symbol string or bit string. It is characterized by providing the means to output.

  According to a fourth aspect of the present invention, the means for encoding the information block of the error correction capability analyzer of the first aspect uses a Reed-Solomon code having the number of information blocks k, and from any data set, An arbitrary number of encoded blocks are extracted so as not to overlap, and the extracted blocks are connected to each other.

  According to a fifth aspect of the present invention, in the semiconductor laser in the presence of return light, the means for generating the output signal light on which the return light noise of the error correction capability analysis apparatus of the first aspect is superimposed is the return signal. An output signal light on which noise caused by light is superimposed is generated.

  By outputting the coding gain using the error correction coded data to which the return optical noise is added, and outputting the coding gain, it is possible to analyze the error correction capability.

It is a flowchart which shows the error correction capability analysis in the error correction capability analyzer which concerns on one Embodiment of this invention. It is a schematic diagram which shows the procedure of the encoding (steps 102-107) in FIG. It is a block diagram which shows the structure of an encoding block. It is a figure which shows the digital modulation signal output in step 108 of FIG. The output signal light of the digital modulation signal obtained from the rate equation is shown. The reproduced digital signal reproduced by discriminating and determining the output signal light is shown. It is a schematic diagram which shows the procedure of the decoding (steps 110-112) in FIG. It is a graph showing the code error rate BER depending on the optical output average intensity Pin. 5 is a flowchart illustrating a method for calculating error correction characteristics (BER) in a system in which FEC is implemented. 3 is a flowchart showing a method for generating output signal light in an isolator-less semiconductor laser.

  FIG. 1 is a flowchart showing error correction capability analysis in an error correction capability analysis apparatus according to an embodiment of the present invention. FIG. 2 is a schematic diagram showing an encoding procedure in the flowchart of FIG.

  In the present embodiment, an RS code having an encoding block number n = 255 bytes, a redundant block number 2t = 32 bytes, and an information block number k = n−2t = 223 bytes is used as an error correction code (this code is RS (n, k)). ). Maximum tbyte error correction is possible.

  The error correction capability analysis of this embodiment starts at step 101, and includes frame data that does not include errors used for actual measurement (the number of frames is 100, the number of data per frame is 1062 bytes. Each data is in units of 8 bits. 8 bit unit symbols are sequentially read and converted into decimal numbers for each symbol data dec_pre-enc. It is output as dat (step 102). The output dec_pre-enc. dat includes data of the number of symbols totalframe = 106200 bytes (1062 bytes × 100 frames).

  Next, a method for calculating the encoded code error rate (BER-ec) will be described.

  First, in step 103, the start line number $ start = 1 + k * (i−1), the end line number $ end = $ start + k−1 (i is an integer), and dec_pre-enc. From dat, a symbol (information block) used for an RS code between $ start and $ end is extracted. Here, k = 223 decimal data is extracted symbol data extdec_pre-enc. Output as dat.

Next, in step 104, a coded block of RS code is generated and output. Here, first, an expanded Galois field GF (2 ^m ) composed of 2 ^m elements is defined. In this embodiment, since m = 8, the expanded Galois field is
GF (2 ⁸ ) = {0, α ⁰ ,..., Α ²⁵⁴ }
It is. Since the number of redundant blocks is 2t, the generator polynomial is
G (x) = (x−α ⁰ ) (x−α ¹ ) (x−α ^2t−1 )
It is.

The extdec_pre-enc. The dat data set {d ₀ , d ₁ ,..., d _k-1 } is used as an information block data set, and an information block polynomial Data (x) is generated. Then
Data (x) = d _k−1 x ^k−1 +... + D ₁ x ¹ + d ₀
It can be expressed. If the polynomial of the redundant block is R (x), the encoded polynomial C (x) is
C (x) = x ^nk Data (x) + R (x)
It is. C (x) must be divisible by G (x)
R (x) = x ^nk Data (x) mod G (x)
It is. A data set {r ₀ , r1,..., R _nk } of redundant blocks is generated from R (x).

The data set of the data set and the information block of the generated redundancy blocks linked, coded blocks _{{r 0, r 1, ···} , r k-1, d 0, d 1, ···, d k-1 } _I is generated and encoded block data dec_post-enc. Output as dat. FIG. 3 is a configuration diagram showing the configuration of the encoding block, and shows one (201-i) of the encoding blocks 201-1 to 201-k in FIG.

  i is set to i = i + 1, and the process returns to step 103 again. Symbols (information blocks) between $ start and $ end are extracted, and an encoded block is generated (step 104). This loop is repeated until i = imax1 = int (totalframe / k) (step 105).

Concatenating the respective encoding blocks 201-1 to 201-k output in step 104,
{R ₀ , r ₁ ,..., R _nk−1 , d ₀ , d ₁ ,..., D _k−1 } _{i = 1} ,
{R ₀ , r ₁ ,..., R _nk−1 , d ₀ , d ₁ ,..., D _k−1 } _{i = 2} ,
..., {r ₀ , r ₁ , ..., r _nk-1 , d ₀ , d ₁ , ..., d _k-1 } _{i = imax}
Are encoded block data dec_all_encode. It is output as dat (step 107). In FIG. 2, a coding block 202 is obtained by connecting coding blocks 201-1 to 201-k.

  Returning to FIG. 1 again, in step 108, the concatenated encoded block data dec_all_enc. dat is converted into a binary number and the digital modulation signal bin_all_encode. Output as dat. Here, FIG. 4 is a diagram showing the digital modulation signal output in step 108.

  In step 109, the digital modulation signal in step 108 is used as modulation data, and analysis by an algorithm is performed to generate output signal light on which noise is superimposed. Specifically, in the presence of the return light, the distributed feedback laser is connected to the digital modulation signal bin_all-encode. Modulation is directly performed by dat, and as an algorithm analysis, an output signal light is generated from a digital modulation signal by numerically solving a rate equation. The solvated light output is output from the output signal light solve. Output as dat. FIG. 5 shows the output signal light obtained from the rate equation.

  In step 110, the output signal light solve. dat is discriminated and determined by a predetermined discriminating point, and a digital signal is reproduced, and a digital reproduction signal data_recover. Output as dat. FIG. 6 shows the output reproduced digital signal.

  Thereafter, the reproduced digital signal is decoded. FIG. 7 is a schematic diagram showing a decoding procedure in FIG. Returning to FIG. 3, in step 111, the digital reproduction signal data_recover. dat is divided into 8-bit unit symbols, each symbol is converted into a decimal number, and the digital reproduction signal bin_all-solve. Output as dat.

In step 112, the digital playback signal is decoded to generate a decoding block 701-i. Here, it is assumed that start line number $ start = 1 + n * (i−1), end line number $ end = $ start + n−1 (i is an integer), bin2dec. Data of n = 255 bytes is extracted from the dat data. The syndrome is calculated from the extracted data, the number of errors and the error location are detected, error correction is performed, and decoding is performed. Decoded 255- _byte decoded block {r ₀ , r ₁ ,..., R _nk−1 , d ₀ , d ₁ ,..., D _k−1 } _{i = 1, decode}
Dec_decode. Output as dat.

  i is set to i = i + 1, and the process returns again to step 112, where the digital reproduction signal is decoded to generate a decoded block. This loop is repeated until i = imax2 = int (totalframe / n) (step 113).

In step 114, the decoding blocks 701-1 to 701-n in step 113 are concatenated,
{R ₀ , r ₁ ,..., R _nk−1 , d ₀ , d _1, ..., D _k−1 } _{i = 1, decode} ,
{R ₀ , r ₁ ,..., R _nk−1 , d ₀ , d ₁ ,..., D _k−1 } _{i = 2, decode} ,
..., {r ₀ , r ₁ , ..., r _nk-1 , d ₀ , d ₁ , ..., d _k-1 } _{i = imax, decode}
Dec_all-decode. Output as dat. In FIG. 7, a decoding block 702 is obtained by concatenating the decoding blocks 701-1 to 701-n.

  Returning to FIG. 1 again, in step 115, the concatenated decoded block data dec_all-decode. dat is converted to a binary number and bin_all-decode. dat and output bin_all-solve. The binary sequence of dat is converted into a symbol of 8-bit unit, each symbol is converted into a decimal number, and dec_all-solve. Output as dat. dec_all-solve. The process of outputting as dat is bin_all-solve. As long as dat output is executed, processing may be performed in parallel with other processing.

  In step 116, an encoded code error rate (BER-ec) is calculated. The error-free bit data bin-all-encode. dat and the target bit data bin-all-decode. dat is compared, the number of bits having different values is counted, and divided by the total number of bits of the target bit data. The total number of bits is 255 bytes × 4 × 100 frames × 8 bits = 816000 bits.

  The step size dP is added to the light output average intensity Pin to obtain a new light output average intensity Pin, and the process returns to step 109 to generate output signal light on which noise is superimposed again. The loop of steps 109 to 116 is repeated, and the BER-ec Pin-dependent output is output until the light output average intensity Pin becomes equal to or greater than a predetermined value Pmax (here, dP = 0.2 dB, Pmax = −20 dBm). Repeat (step 117).

Next, a method for calculating an unsigned code error rate (BER-woec) will be described. Again, the extdec_pre-enc. Let dat's data set {d ₀ , d ₁ ,..., d _k−1 } be an information block data set.

  i is set to i = i + 1, and the process returns to step 103 again to extract symbols (information blocks) between $ start and $ end. This loop is repeated until i = imax1 = int (totalframe / k) (step 106).

  In step 118, the information blocks output in step 103 are connected, and information block data dec_all_pre-encode. Output as dat.

  In step 119, the connected information block data dec_all_pre-encode. dat data is converted into a binary number and the digital modulation signal bin_all_pre-encode. Output as dat.

  In step 120, the digital modulation signal bin_all_pre-encode. Using dat as modulation data, output signal light on which noise is superimposed is generated. Specifically, in the presence of the return light, the distributed feedback laser is connected to the digital modulation signal bin_all_pre-encode. The output signal light is generated by directly modulating with dat and solving the rate equation numerically. The solved light output is output as output signal light solve_pre-encode. Output as dat.

  In step 121, the output signal light solve_pre-encode. dat is identified and reproduced, and the digital reproduction signal data_recover_pre-encode. Output as dat.

  In step 122, the error-free bit data bin_all_pre-encode. dat and the target bit data data_recover_pre-encode. dat is compared, and BER-wofec is calculated in the same manner as the calculation of BER-ec in step 116. The total number of bits is 223 bytes × 4 × 100 frames × 8 bits = 713600 bits.

  The step dP is added to the optical output average intensity Pin to obtain a new optical output average intensity Pin, and the process returns to step 120 to generate output signal light on which noise is superimposed again. The loop of steps 120 to 123 is repeated, and the BER-wooc Pin-dependent output is output until the light output average intensity Pin becomes equal to or greater than a predetermined value Pmax (here, dP = 0.2 dB, Pmax = −20 dBm). Repeat (step 123).

  Finally, in step 124, a code error rate BER characteristic depending on the optical output average intensity Pin is calculated for the encoded code error rate and the unsigned code error rate. FIG. 8 is a chart showing the code error rate BER depending on the optical output average intensity Pin. In FIG. 8, the optical reception intensity in the unsigned code error rate characteristic for obtaining a predetermined code error rate BER = BER_g is Pin0, the optical reception intensity in the encoded code error rate characteristic is Pin1, and the coding gain G = Pin0. -Pin1 is output and the process ends in step 125.

  In addition to the bit error characteristics, it is also possible to analyze a symbol error rate by an arbitrary symbol, and to execute the above method using data subjected to processing such as scrubbing in addition to an error correction code.

  As described above, it is possible to output the coding gain using the error correction encoded data to which the return optical noise is added and to analyze the error correction capability.

201-1-201-k coding block 202 concatenated coding block 701-1-701-n decoding block 702 concatenated decoding block

Claims (5)

A code error correction capability analyzer for analyzing error correction capability of a transmission signal including return optical noise of a semiconductor laser,
First code error rate characteristic calculating means, comprising:
Means for encoding an information block and generating an encoded block;
Means for generating the first output signal light in which the generated encoded block is used as modulation data and the output light of the semiconductor laser is superimposed on the optical noise;
Means for identifying and reproducing the first output signal light to generate a first digital reproduction signal;
Means for decoding the first digital reproduction signal and generating a decoded block;
Means for comparing the encoded block and the decoded block to calculate a first code error rate characteristic; and
Second code error rate characteristic calculating means,
Means for generating the second output signal light in which the information block is modulated data and the output light of the semiconductor laser is superimposed on the optical noise;
Means for identifying and reproducing the second output signal light and generating a second digital reproduction signal;
Means for comparing the information block and the second digital reproduction signal to calculate a second code error rate characteristic;
A first code error rate characteristic calculated by the first code error rate characteristic calculation unit; and calculating an output signal light intensity for obtaining a predetermined error rate, and calculating the second code error rate characteristic Means for calculating an output signal light intensity for obtaining a predetermined error rate in the second code error rate characteristic calculated by the means, and calculating a difference between the output signal light intensities. Capability analysis device. The means for encoding the information block comprises means for defining an expanded Galois field consisting of 2 ^m elements, where m is an integer, and means for converting the information block into a polynomial and performing block encoding operations. The code error correction capability analysis apparatus according to claim 1.   2. The code error correction capability analysis according to claim 1, wherein the means for generating the decoded block comprises means for detecting an error value, correcting the error, and outputting a decoded symbol string or bit string. apparatus.   The means for encoding the information block uses a Reed-Solomon code having k information blocks, extracts an arbitrary number of encoded blocks from an arbitrary data set so as not to overlap, and connects the extracted blocks. The code error correction capability analysis apparatus according to claim 1, further comprising: means.   The means for generating the output signal light on which the return light noise is superimposed generates the output signal light on which the noise caused by the return light is superimposed in a semiconductor laser in the presence of the return light. The code error correction capability analyzer according to 1.

Images