A joint coding modulation method suitable for short distance high speed optical fiber communication system

Abstract

The application provides an encoding modulation method suitable for a short-distance high-speed optical fiber communication system, which comprises the following steps: step 1, grouping input information bits into a first information sequence u a with a length of k a bits without any operation and a second information sequence u s with a length of k s bits to be shaped; step 2, converting uniformly distributed binary information into a symbol sequence with a target distribution by using a distribution matcher based on an enumeration sphere mapping method, and the symbol sequence after binary labeling is called the shaped information u c ; step 3, taking u c as input information of a forward error correction coding module, and performing twice encoding, first encoding based on a hard decision ladder code, and then encoding based on a soft decision Hamming code. The scheme combines the probability shaping technology based on the enumeration sphere mapping and the forward error correction coding based on the high-speed short-distance optical interconnection, maintains the probability distribution, and avoids the damage of the forward error correction coding to the probability distribution.

Description

Technical Field

[0001] This invention relates to the field of optical fiber communication technology, and in particular to a joint coding and modulation method suitable for short-distance high-speed optical fiber communication systems. Background Technology

[0002] With the continuous development of communication technology, people's demand for high-speed, high-capacity communication networks is becoming increasingly urgent. In the field of short-range optical communication, lower transmission power and higher information rate are the current goals, but the two are often mutually exclusive.

[0003] Forward error correction coding (FEC) is a commonly used error correction method in communication systems. FEC codes help the system detect and reverse the amount of corrupted data in received data without requiring data retransmission. The FEC coding technique first adds redundant bits (parity bits) to the information data through encoding to form a codeword at the transmitter. Then, the codeword is transmitted, and the receiver decodes the received codeword to detect errors and correct corrupted bits in the information data. Robust and efficient error correction coding is an indispensable part of communication systems. In high-speed optical data center interconnects, to meet the ever-increasing data rate demands, the OIF proposed the 400ZR implementation protocol. In the FEC part, hard-decision ladder codes (255, 239) are used as the outer code, and soft-decision Hamming codes (128, 119) are used as the inner code, forming a concatenated coding (CFEC).

[0004] Probabilistic shaping, by adjusting the probability of symbol occurrence to reduce the average power of a signal, is an effective means of improving the spectral efficiency of fiber optic communication systems. Furthermore, some probabilistic shaping algorithms with peak power constraints also possess a certain ability to combat nonlinear effects. Existing high-speed modulation techniques do not incorporate probabilistic shaping, which makes transmit power a persistent critical issue. This proposed solution combines probabilistic shaping with FEC (Flexible Encoding and Coding) to reduce the overall power loss of the communication system while ensuring information accuracy and reliability, showing great promise for future high-speed optical communication networks. Summary of the Invention

[0005] To address the aforementioned technical problems, this invention provides a coding and modulation method suitable for high-speed, short-range optical communication networks. It combines an enumerated ball mapping algorithm with ladder codes, using a bit-to-bit matching method of probabilistic shaping and forward error correction coding. Redundant bits are extracted and combined with a portion of the original source information to be mapped to the symbols of constellation points. This method maintains the probability distribution unchanged and avoids the disruption of the probability distribution caused by forward error correction coding.

[0006] The technical scheme of the present application is: a coding modulation method suitable for short-distance high-speed optical fiber communication system, which combines the probability shaping technology based on enumeration sphere mapping and the forward error correction coding (cascade coding of ladder code and Hamming code) based on high-speed short-distance optical interconnection, and specifically comprises the following steps:

[0007] Step 1, grouping the input information bits into a first information sequence u of k a bit length without any operation, and a second information sequence u a of k s bit length to be shaped; s

[0008] Step 2, using a distribution matcher based on the enumeration sphere mapping method to convert the uniformly distributed binary information into a symbol sequence with a target distribution, and the above symbol sequence after binary labeling is called the shaped information u c .

[0009] Step 3, taking u c as the input information of the forward error correction coding module, which is encoded twice: ladder code based on hard decision and Hamming code based on soft decision.

[0010] Further, before step 1, the input source information is further divided into I and Q two paths, and the two paths are processed in the same way.

[0011] Further, step 4: information interleaving, extracting check bits, and arranging the information in the order of u c , u a , and u OH is further included.

[0012] Further, step 5: planet constellation mapping, extracting u a and u OH as the symbols of the constellation points, and u c as the amplitude of the constellation points, and the specific mapping rule of the I path is the same as that of the Q path.

[0013] Beneficial effects

[0014] The present application combines an enumeration sphere mapping shaping algorithm based on spatial coding with a low-complexity forward error correction coding, thereby achieving the effects of efficient utilization of the energy space of the sequence and high-speed coding at the same time. The joint coding modulation method proposed by the present application makes the average power of the shaped symbols lower and guarantees the accuracy and reliability of the information, and has the advantages of low energy consumption, low latency, high accuracy, etc. BRIEF DESCRIPTION OF DRAWINGS

[0015] Figure 1 : flow chart of the method of the present application; ​

[0016] Figure 2 : Amplitude trellis diagram

[0017] Figure 3 : Ladder-like concatenation of each code block

[0018] Figure 4 : Shaped constellation

[0019] Figure 5 : Ladder code allocation in one embodiment DETAILED DESCRIPTION

[0020] The technical solutions in the embodiments of the present application will be described clearly and completely below in conjunction with the drawings of the embodiments of the present application. Obviously, the described embodiments are only some of the embodiments of the present application, rather than all the embodiments of the present application. Based on the embodiments of the present application, all other embodiments obtained by those of ordinary skill in the art without creative effort are within the protection scope of the present application.

[0021] According to the embodiments of the present application, an encoding and modulation method suitable for a short-distance high-speed optical fiber communication system is provided, which combines a probability shaping technique based on enumeration sphere mapping and forward error correction coding (concatenation coding of ladder code and Hamming code) based on high-speed short-distance optical interconnection, as shown in Figure 1 , and specifically includes the following steps:

[0022] Step 1, grouping the input information bits, into a first information sequence u a of length k a bits without any operation, and a second information sequence u s of length k s bits to be shaped.

[0023] Step 2, using a distribution matcher based on the enumeration sphere mapping method to convert the uniformly distributed binary information into a symbol sequence with a target distribution, which is called the shaped information u c after binary labeling.

[0024] Step 3, taking u c as the input information of the forward error correction coding module, which is encoded twice, first by ladder code coding based on hard decision, and then by Hamming code coding based on soft decision.

[0025] Step 4, information interleaving and constellation mapping.

[0026] Specifically, the main design parameters in this method include: shaping sequence length N, maximum energy E max of shaping sequence, amplitude sequence A = {a1, a2, …}, shaping code rate Rs , the rate of the staircase code R HD , the rate of the Hamming code R SD .

[0027] The encoding and modulation flow chart of the transmitting end is shown in Figure 1 . In the present application, the source information will be divided into two parts, one is the information sequence u a of k a bits without any operation, and the other is the information sequence u s of k s bits to be shaped. The distribution matcher converts the uniformly distributed binary information into a symbol sequence with a target distribution, and the above sequence after binary labeling is called the shaped information u c . The shaped information u c is forward error correction coded as information bits, and additional check information u OH is generated. u c , u a and u OH are formed into constellation points according to the mapping rule of 16QAM and then transmitted in the optical fiber.

[0028] u s is mapped from binary information into an amplitude sequence with a certain length by the distribution matcher based on enumeration sphere shaping. At the same shaping rate, ESS has lower rate loss than constant composition distribution matcher (CCDM), so the probability shaping part in the present application selects the algorithm based on ESS to meet the requirements of high-speed optical communication. s The mapping relationship with the amplitude sequence is determined by the bounded energy amplitude trellis of ESS.

[0029] Take the amplitude trellis structure shown in Figure 2 as an example. In the figure, n represents the number of columns of the amplitude trellis structure; l represents the number of rows of the amplitude trellis structure; and e represents the cumulative energy of the amplitude sequence on the previous n states of the current state. In this structure, each state represents an energy level, each branch represents an amplitude in the amplitude set A = {a1, a2, …}, and each path represents an amplitude sequence from the zero-energy state to the last state. represents the total number of paths from the current energy state to the last energy state, and therefore represents the total number of sequences of the structure. In the joint coding scheme, only the sequence set with the number satisfying the integer power of 2 can establish one-to-one mapping with the source information, so part of the sequences need to be discarded. The shaped sequence set S0 determined by the energy amplitude trellis can establish mapping with the maximum number of bits of

[0030]

[0031] In summary, it can be calculated that the k s bit information sequence u s is mapped to an amplitude sequence with a length of .

[0032] The amplitude sequence needs to be converted into a binary sequence u c before forward error correction coding can be performed. Let k" represent the number of bits required to represent an amplitude in binary, and u c contains a number of bits

[0033] In order to reduce the complexity of system calculation, the present application further proposes a new constraint relationship for encoding information and shaping information, which will be described later.

[0034] The present application only takes u c as the input information of the forward error correction coding module, and u c will be encoded twice. The first encoding is a hard-decision-based step code, aiming to reduce the coding complexity, and the second encoding is a soft-decision-based Hamming code, aiming to reduce the error rate. The step code used in the present application combines the ideas of convolutional coding and block coding, and the FEC block code selects a BCH code with a code length of 2m as a component code, and the parity check bit length is r, so the information code word length is 2m-r. A m x m matrix B i (i = 0, 1, 2, …, n) is selected, the left m-r columns constitute the information bits, and the right r columns constitute the check bits. The initial matrix B0 is set as an all-zero matrix, as shown in Figure 3 Each code block is cascaded in a step-by-step manner to form a step code, and the gray part represents the check bits. Each row of the matrix [BT i-1, B i ] is a valid code word of BCH, that is, each row and each column of the step block is a valid code word of BCH. The soft-decision coding mode is Hamming code, which further ensures the accuracy of the information.

[0035] The overhead introduced by the forward error correction coding obeys a uniform distribution, in order to ensure that the shaping distribution does not change, the present scheme requires that the check information u OH and the information u a together represent the symbol of the constellation point in the constellation diagram. This requires that k c , k a and k OH satisfy the following relationship:

[0036]

[0037] where M represents the modulation order, such as M = 4 for 16QAM modulation. From this, the source part ua and u s The length relationship satisfies:

[0038]

[0039] Embodiment 1

[0040] In order to more clearly illustrate the technical solutions in the embodiments of the present application or the prior art, the following will be described with reference to an 800 Gbit / s short-distance optical communication system with DP-16QAM modulation as an example. The embodiment is implemented on the premise of the design method, but is not limited to this transmission system. All other embodiments obtained by those of ordinary skill in the art without creative labor are within the scope of the present application.

[0041] In order to reduce the system operation pressure, the number of repeated operations is set to 6868. The parameter design of the present example is as follows: the shaping sequence length N = 16, the maximum energy E max of the shaping sequence = 72, the amplitude sequence A = {1, 3}, the shaping code rate Step code rate Hamming code rate The scheme flow is as follows:

[0042] (1) The source information is divided into I and Q two paths.

[0043] (2) The same processing is performed on the I and Q paths. Taking the I path as an example, the data to be transmitted is divided into u s (25,483,136 bits) and u a (24,823,808 bits) parts, wherein u s enters the ESS distribution matcher for probability shaping, and u a is left for standby without any processing.

[0044] (3) The distribution matching is performed on u s . The probability amplitude shape is designed according to the following parameters: let the shaping sequence length N = 16, the maximum sequence energy E max = 72, when the shaping amplitude set A = {1, 3}, a total of T0 0 = 26332 groups of sequences can be generated by enumeration operation, wherein the 2 14 groups of sequences with the lowest energy are applied to shaping, and thus the shaping code rate R s = 14 / 16 = 0.8750. The constellation diagram after shaping is shown in Figure 4 After shaping, the probability of the amplitude "1" appearing is 0.6691, the probability of the amplitude "3" appearing is 0.3309, and the average energy of the constellation diagram is 3.6472.

[0045] (4) Represent the amplitude sequence in binary to obtain the input information u of the forward error correction coding module. c (29,123,584 bits).

[0046] (5) Information u c Ladder coding is performed, where the code rates of the ladder codes are R. HD and R SD Ladder coding will incur overhead u OH The initial matrix B0 is set to a 512×512 matrix of all zeros. Then B1 is 512×510. The left 478 columns of B1 are used as information bits, and the right 32 columns are parity bits. Subsequent ladder blocks are constructed in the same manner as block B1. i Each row of the code satisfies the selected component code BCH(1022,990). However, a problem arises during ladder construction: since the blocks after B1 are not constructed as an m×m matrix, they cannot be directly assembled, resulting in a 2-column error. Therefore, the solution is to add a 512×2 all-zero matrix after encoding B1 and the subsequent matrices, forming a 512×512 square matrix. This allows for ladder combination using the traditional ladder encoding method. Figure 5 As shown, the gray area represents the parity bit, and the diagonally filled area represents the added 512×2 all-zero matrix. Regarding the coding efficiency of the ladder code, the coding efficiency of the ladder code constructed from the component code BCH(1022,990) is...

[0047] (6) Hamming Encoding. The codewords encoded using the echelon code are then encoded using Hamming code to further ensure the accuracy of the information. The Hamming code rate is...

[0048] (7) Information interleaving. Extract the check bit, and according to u c u a u OH The information is arranged in the correct order.

[0049] (8) Constellation mapping. Extract u a and u OH As a symbol for constellation points, u c As for the amplitude of constellation points, the specific mapping rules for the I path are shown in the table below, and the same applies to the Q path.

[0050]

[0051]

[0052] The data with a total number of 345,508,091,392 bits is modulated by joint coding modulation of probability shaping and forward error correction, additional shaping information and check information are introduced, and finally 400,041,549,824 bits of data are modulated into the optical fiber channel according to the 16QAM modulation format. The total code rate of the embodiment is 0.86368.

[0053] Although the foregoing describes the specific embodiments of the present application in detail, so as to make the person skilled in the art understand the present application, it should be clear that the present application is not limited to the scope of the specific embodiments, and for the person skilled in the art, as long as various changes are within the spirit and scope of the present application defined and determined by the appended claims, all the inventions utilizing the concept of the present application are within the scope of protection.

Claims

1. A coding and modulation method suitable for short-distance high-speed optical fiber communication systems, characterized in that: Step 1: Divide the input information bits into groups of length [length to be filled in] without any operation. First information sequence And the length to be reshaped is The second information sequence ; Step 2: A distribution matcher based on the enumerated ball mapping method is used to convert the uniformly distributed binary information into a symbol sequence with a target distribution. This symbol sequence after binary labeling is called the shaped information. ; Step 3, The input information to the forward error correction coding module is encoded twice: first based on hard-decision ladder code, and then based on soft-decision Hamming code. In step 2, a bit-to-bit matching method based on probabilistic integer shaping and forward error correction coding is used. For the second information sequence, let the verification information... and symbol information The symbols that together represent constellation points in a constellation diagram, information bits. , and check bits The following relationship must be satisfied: in, RHD It is the code rate of the ladder coding. RSD It is the code rate of Hamming encoding. M Indicates the modulation order; The design parameters include: the length N of the integer sequence, and the maximum energy of the integer sequence. , amplitude sequence Integer coding rate Step code bit rate Hamming code rate ; n represents the number of columns in the amplitude lattice structure; l represents the number of rows in the amplitude lattice structure; e represents the cumulative energy of the amplitude sequence over the previous n states, where each state represents an energy level and each branch represents an amplitude set. In the amplitude of the path, each path from the zero-energy state to the last state represents an amplitude sequence. This represents the total number of paths from the current energy state to the last energy state. This represents the total number of sequences in this structure; The set of shaped sequences determined by the energy amplitude lattice The maximum number of bits required to establish a mapping is （1） Based on the above calculations, it is found that... Information sequence Mapped to a length of amplitude sequence; use This represents the number of bits required to represent one amplitude in binary. The number of bits contained is ; Obtain the source part and The length relationship satisfies: 。 2. The coding and modulation method for short-distance high-speed optical fiber communication systems according to claim 1, characterized in that: Before step 1, the process also includes dividing the input source information into two paths, I and Q, and performing the same processing on both paths.

3. The coding and modulation method for short-distance high-speed optical fiber communication systems according to claim 1, characterized in that: It also includes step 4: Information intertwined, extract check bits, and follow , , The information is arranged in the correct order.

4. The coding and modulation method for short-distance high-speed optical fiber communication systems according to claim 1, characterized in that, It also includes step 5: Perform constellation mapping and extract and As a symbol of constellation points As for the amplitude of constellation points, the specific mapping rules for the I-path are the same as those for the Q-path.

Images