Turbo code variable interleaving depth coding method for atmospheric laser communication

Abstract

The application belongs to the technical field of atmospheric laser communication, and particularly relates to a Turbo code variable interleaving depth coding method for atmospheric laser communication, which comprises an encoder, channel estimation, a decoder and a variable depth interleaver, and comprises the following steps: S1, inputting a to-be-encoded sequence into a Turbo encoder, and encoding the sequence through a component encoder, an interleaver and a punctured multiplexer; S2, adding a pilot sequence to the encoded sequence, and sending the sequence into an atmospheric channel through an optical antenna; and S3, performing channel estimation through the known pilot sequence in the received sequence, and feeding back the channel estimation value to the variable depth interleaver. The application adds a channel estimation technology to estimate the channel state in real time according to the time-varying characteristics of the atmospheric channel, provides data reference for the interleaving depth switching of the interleaver in Turbo code encoding, and guarantees the reliability of the communication system under poor channel conditions. The decoder adopts a pipeline parallel decoding structure, effectively reduces the hardware resource consumption, and effectively balances the error correction performance and decoding delay.

Description

Technical Field

[0001] This invention relates to a Turbo code variable interleaving deep encoding and decoding method for atmospheric laser communication, belonging to the field of atmospheric laser communication technology. Background Technology

[0002] The Turbo codes, proposed by Claude Belou et al. in 1993, have had a profound impact on the field of channel coding research due to their decoding performance approaching the Shannon theoretical limit. Turbo code encoding incorporates convolutional coding and interleavers, employing a parallel concatenation method to achieve random channel coding designs that construct long codes from short codes. Turbo code decoding uses a soft-input soft-output (SISO) decoding method, fully utilizing the soft information provided by the decoding of preceding and following symbols. After multiple iterations of SISO, the decoding performance approaches that of maximum likelihood decoding.

[0003] For atmospheric laser communication, the random nature of atmospheric channels causes channel states to change over time. Atmospheric turbulence, in particular, leads to fluctuations in received light intensity, beam drift, and beam spread at the receiver, severely impacting the stability and reliability of laser transmission. Turbo codes, as concatenated codes, are a mature and high-performance error-correcting code, exhibiting excellent coding gain, especially at low signal-to-noise ratios. The component codes employ recursive system convolutional codes, meeting the coding requirements of atmospheric laser communication channels. Considering the characteristics of atmospheric laser communication channels, and minimizing the bit error rate under limited laser emission power and hardware complexity, Turbo code encoding can effectively improve the performance of atmospheric laser communication systems.

[0004] The decoding performance of Turbo codes, especially in low signal-to-noise ratio (SNR) channels, is limited by the interleaving depth of the interleaver. Achieving random coding with a limited interleaving depth is not ideal; the smaller the interleaving depth, the worse the randomness. When the interleaving depth is set appropriately, an interleaver that meets certain distance requirements can achieve good performance. However, traditional Turbo code encoders and decoders use a fixed interleaving depth, which cannot adaptively adjust to changes in channel conditions, resulting in suboptimal error rate performance. Furthermore, increasing the interleaving depth leads to increased decoding delay, making it difficult to balance error rate performance with decoding delay. Therefore, we propose a variable interleaving depth Turbo code encoding and decoding method for atmospheric laser communication to address these issues. Summary of the Invention

[0005] (a) Technical problems to be solved

[0006] To address the shortcomings of existing technologies, this invention provides a Turbo code variable interleaving deep encoding and decoding method for atmospheric laser communication, which solves the problems mentioned in the background section.

[0007] (II) Technical Solution

[0008] To achieve the above objectives, the present invention specifically adopts the following technical solution:

[0009] A Turbo code variable-depth interleaving encoding and decoding method for atmospheric laser communication includes an encoder, channel estimation, a decoder, and a variable-depth interleaving unit, comprising the following steps:

[0010] S1, the sequence to be encoded is input into the Turbo encoder, and the sequence is encoded by the component encoder, interleaver and pruning multiplexing;

[0011] S2, the encoded sequence is added to the pilot sequence and sent into the atmospheric channel through the optical antenna;

[0012] S3, perform channel estimation using the known pilot sequence in the received sequence, and feed the channel estimation value back to the variable depth interleaver;

[0013] S4: Receive sequence input decoder. After the iterative decoding terminates, perform a hard decision on the likelihood ratio output by the component decoder to obtain the final decoding result. Calculate the bit error rate by counting the number of erroneous symbols.

[0014] S5, the variable depth interleaver updates the coding parameters based on the feedback channel estimate, and changes the interleaver's interleaving depth to re-encode the next frame of the sequence to be sent.

[0015] Furthermore, in S1, the Turbo code encoder interleaver adopts a random interleaving method with variable interleaving depth. The initial state of the interleaving depth of the interleaver is set to a fixed value of 512, so as to perform the initial channel estimation for decoding the received sequence with the minimum decoding delay.

[0016] Furthermore, in S3, the channel estimation adopts the least squares (LS) algorithm based on training symbols. The channel estimation is performed using comb pilot signals. Based on this, linear interpolation and quadratic polynomial interpolation are performed on the pilot positions to estimate the channel carrying data symbols. The estimated value of the signal-to-noise ratio is calculated from the channel estimation value.

[0017] Furthermore, the Turbo code decoder in S4 adopts a pipelined iterative decoding structure. During the iterative decoding process, the forward and backward metrics are optimized at certain recursion intervals. The optimization target value is obtained by pre-simulating the change in the transmitted data frame length in laser communication, and the statistical values ​​of the forward and backward metrics are empirical reference values.

[0018] Furthermore, the Turbo code decoder in S4 adopts the Log-MAP decoding algorithm and calculates the log-likelihood ratio (LLR) in parallel. The LLR is calculated in reverse recursion and is performed in parallel and synchronously with the backward metric calculation unit, which effectively reduces the hardware resource consumption of the decoding algorithm.

[0019] Furthermore, in S5, the interleaving depth of the variable-depth interleaver is variable. Under different atmospheric turbulence intensities, the atmospheric channel turbulence state is determined based on the channel estimation feedback value, and the encoder interleaving depth is reset according to the signal-to-noise ratio estimation threshold range.

[0020] (III) Beneficial Effects

[0021] Compared with existing technologies, this invention provides a Turbo code variable interleaving deep encoding and decoding method for atmospheric laser communication, which has the following advantages:

[0022] This invention incorporates channel estimation technology to estimate atmospheric channel conditions in real time, providing data reference for selecting the interleaving depth in the encoder. By adjusting the interleaving depth based on the time-varying characteristics of the atmospheric channel environment, the reliability of the communication system under adverse channel conditions can be guaranteed, and the system's anti-interference capability can be improved. While ensuring the system's target bit error rate, the interleaving depth can be reduced in scenarios with good channel conditions to effectively control decoding delay. Simultaneously, the decoder employs a pipelined parallel decoding structure, which effectively reduces hardware resource consumption caused by decoding complexity, achieving an effective balance between Turbo code error correction performance and decoding delay. Attached Figure Description

[0023] Figure 1 This is a schematic diagram summarizing the Turbo code encoding and decoding structure of this invention;

[0024] Figure 2 This is a schematic diagram of a pipelined iterative decoding structure in an embodiment of the present invention;

[0025] Figure 3 for Figure 2 Schematic diagram of the Log-MAP algorithm structure for a medium component decoder;

[0026] Figure 4 This is a flowchart of the variable depth interleaver switching process of the present invention;

[0027] Figure 5 A diagram showing the threshold and interleaving depth settings for dividing the signal-to-noise ratio estimation range. Detailed Implementation

[0028] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0029] Example

[0030] like Figures 1-5 As shown in the figure, an embodiment of the present invention proposes a Turbo code variable-depth interleaving encoding and decoding method for atmospheric laser communication, including an encoder, a channel estimate, a decoder, and a variable-depth interleaver; wherein,

[0031] S1, the sequence to be encoded is input into the Turbo encoder, and the sequence is encoded by the component encoder, interleaver and pruning multiplexing;

[0032] The Turbo code encoder consists of a component encoder, an interleaver, and a pruning multiplexer. The component encoder uses recursive systematic convolutional codes and employs a parallel concatenation method to effectively combine short codes to construct long codes. The interleaver is introduced to achieve random encoding of the input information sequence. During the initial encoding, the interleaver is set with a fixed interleaving depth to maximize compatibility with channel conditions and ensure normal communication.

[0033] The encoder employs a random interleaving method, assigning a randomly generated mapping address to each bit in the sequence. Then, a lookup table is used to implement symbol interleaving, reducing information correlation and increasing code weight. The interleaving depth of the interleaver is initially set to a fixed value of 512 to maximize compatibility with complex and variable atmospheric channel fading conditions. This ensures normal communication first, followed by initial channel estimation using the first received frame sequence, providing real-time channel state information for the encoding and decoding process and adjusting the interleaving depth.

[0034] S2, the encoded sequence is added to the pilot sequence and sent into the atmospheric channel through the optical antenna;

[0035] The encoded sequence is added to the pilot sequence as the sequence to be transmitted, and the pilot positions and intervals are determined. The sequence to be transmitted is fed into the electro-optic modulator to output the laser signal, which is then sent into the atmospheric channel through a collimating lens.

[0036] S3, perform channel estimation using the known pilot sequence in the received sequence, and feed the channel estimation value back to the variable depth interleaver;

[0037] Channel estimation employs a least squares (LS) algorithm based on training symbols, using comb-shaped pilot signals to determine pilot positions and spacing. After passing through the communication channel, the received sequence is acquired, and channel estimation of pilot positions is performed based on the received pilot signals. Furthermore, linear interpolation and quadratic polynomial interpolation are applied to the pilot positions to estimate the channel carrying data symbols. The solution for channel estimation obtained from the transmitted training symbols and the received training signals is as follows:

[0038]

[0039] In the above formula, X represents the transmitted training symbol, and Y represents the received training symbol. This represents an estimate of the atmospheric channel. Then, the signal power and noise power are estimated separately to calculate the estimated signal-to-noise ratio (SNR). The estimated SNR can be calculated using the following formula:

[0040]

[0041] In the above formula, noise power The following formula can be used to obtain:

[0042]

[0043] In the above formula, This represents the channel estimate at the current symbol frequency k. Let W(k) represent the channel estimate at the previous symbol frequency k. W(k) and W′(k) represent the Fourier transforms of the current and previous AWGN noise, respectively, and they are independent and identically distributed. Therefore, the above equation can be transformed into:

[0044]

[0045] Therefore, the noise power can be expressed as:

[0046]

[0047] Bundle Substitute into the formula The signal-to-noise ratio estimate can be expressed as:

[0048]

[0049] S4: Receive sequence input decoder. After the iterative decoding terminates, perform a hard decision on the likelihood ratio output by the component decoder to obtain the final decoding result. Calculate the bit error rate by counting the number of erroneous symbols.

[0050] The decoder consists of two component decoders, an interleaver, and a deinterleaver, using the soft output information of one component decoder as the input to the next decoding unit. However, different interleaving depth settings in the interleaver can cause varying degrees of decoding delay, making it difficult to provide real-time feedback of the prior information from the component decoder outputs in the hardware design. To meet the real-time requirements of the communication system, a solution such as... Figure 2 The pipelined decoding structure is shown. Where y s For system information, and There are two verification messages; the prior information for the initial decoding is set to 0 by default.

[0051] During the iterative decoding process, a constant is subtracted from the forward and backward metric values ​​at regular recursion intervals. This constant is obtained by pre-simulating the changes in the transmitted data frame length in laser communication, and the difference between the statistical values ​​of the forward and backward metrics obtained and the current iterative recursion value is used as an empirical reference value to prevent overflow when recursively calculating the forward and backward metrics due to changes in the interleaving depth during the decoding process.

[0052] The component decoder is implemented using the Log-MAP decoding algorithm. This algorithm represents all likelihood operations in the MAP algorithm using log-likelihood values, transforms complex exponential operations into addition and subtraction operations, and introduces the Jacobi function to simplify logarithmic operations into the sum of finding the maximum / minimum value and the correction function.

[0053] MAX * (x,y)=ln(e x +e y )=max(x,y)+ln(1+e -|x-y| )

[0054] =max(x,y)+f c (xy)

[0055] In the formula f c (·) is the correction function, which can be pre-created using a lookup table based on the variable values ​​to simplify the calculation. For soft-input soft-output component decoding, this includes calculating the forward metric α, the backward metric β, and the branch transition metric γ, and then calculating the log-likelihood ratio (LLR).

[0056] Since the calculation of both the forward metric α and the backward metric β utilizes the calculated value of the branch transition metric γ, the calculation of the branch transition metric γ will be performed first. The formula for calculating the branch transition metric is:

[0057]

[0058] In the formula, D k (s',s) is the logarithm of the branch transition metric γ between states s' and s; These represent the information bits and check bits received by the component decoder at time k, respectively; u k , La(u) represents the information bits and check bits output by the component encoder at time k, respectively; k ) is prior information. Represent u using logarithms k The probability of a decision being 1 or 0 is obtained by interleaving or deinterleaving the external information output by the component decoder cascaded in parallel with it.

[0059] Formula for calculating forward metric:

[0060] A k (s)=ln(α k (s))=MAX * (A k-1 (s')+D k (s',s))

[0061] In the formula, A k (s) is the logarithm of the forward metric α in the MAP algorithm; D k (s',s) is the logarithm of the branch transition metric γ in the MAP algorithm.

[0062] Formula for calculating backward metric:

[0063]

[0064] In the formula, B k-1 (s) is the logarithm of the backward metric β in the MAP algorithm; D k (s',s) is the logarithm of the branch transition metric γ in the MAP algorithm.

[0065] u k The formula for calculating the log-likelihood ratio (LLR) is:

[0066] L(u k ) = MAX * (A k-1 (s')+D k (s',s)+B k (s))-MAX * (A k-1 (s')+D k (s',s)+B k (s))

[0067] In the above formula, A k-1 (s) is the logarithm of the forward metric α in the MAP algorithm, B k (s) is the logarithm of the backward metric β in the MAP algorithm, D k(s',s) is the logarithm of the branch transition metric γ in the MAP algorithm.

[0068] like Figure 3 As shown, in the pipeline iterative decoding process, a method for parallel computation of the log-likelihood ratio (LLR) in the Log-MAP decoding algorithm is proposed. By recursively calculating the LLR in reverse and in parallel with the backward metric computation unit, the hardware resource consumption of the decoding algorithm is effectively reduced. This method includes the following steps:

[0069] (1) The received information sequence is decomposed into a system information sequence and two-way check information sequences through a demultiplexing operation, and the prior information of the information sequence is calculated.

[0070] (2) The three sequences pass through the branch transition metric calculation unit to calculate the branch transition metric γ and save and transfer the calculated value to the forward metric calculation unit.

[0071] (3) The forward metric α calculation unit uses the obtained branch transition metric γ value to calculate the value of forward metric α and saves the calculated value synchronously. In forward recursion, the branch transition metric calculation unit and the forward metric unit can work sequentially or in parallel;

[0072] (4) Until a complete sequence data frame is received, the branch transition metric calculation unit and the forward metric unit stop running and start backward recursive calculation. The backward metric calculation unit and the soft output calculation unit calculate the backward metric β and the soft output LLR value in parallel.

[0073] The decoding process is repeated. When the decoding performance reaches the iteration termination condition after the maximum number of iterations, a hard decision is made on the likelihood ratio output by the component decoder to obtain the final decoding result. The number of erroneous symbols is counted to calculate the bit error rate.

[0074] S5, the variable depth interleaver updates the coding parameters based on the feedback channel estimate, and changes the interleaver's interleaving depth to re-encode the next frame of the sequence to be sent;

[0075] Variable-depth interleavers employ random interleaving with variable interleaving depth. They adjust the interleaving depth based on real-time channel conditions obtained from channel estimation feedback, achieving an optimal balance between error rate performance and decoding delay under different channel conditions while ensuring system performance. For example... Figure 4 As shown, the specific workflow for switching interleaving depth includes the following steps:

[0076] (1) During system initialization, the interleaver is started to calculate the interleaving address mapping under each interleaving depth parameter and cache it in advance. Since the interleaver uses random interleaving, the number of interleaving address combinations under the same interleaving depth parameter is not less than the number of frames of the information sequence to be sent;

[0077] (2) The interleaving depth of the first frame information sequence encoder is selected as 512. The channel estimation is performed by low-delay decoding to obtain atmospheric channel parameters. Based on the feedback value of the initial channel estimation, the encoder interleaving depth of the subsequent transmitted information sequence is provided to provide accurate channel state information reference.

[0078] (3) Atmospheric channel state estimation is performed using the received sequence, and the coding interleaving depth parameter for the next transmitted sequence is set according to the threshold range where the signal-to-noise ratio estimate is located. For example... Figure 5 As shown, the encoder interleaving depth is reset according to the signal-to-noise ratio (SNR) estimation threshold range. Taking a weak turbulence channel as an example, the interleaving depth values ​​(512, 1024, 2048, 4096, 6144) are selectable, and the SNR estimation is divided into 5 reference intervals accordingly. The starting value of each interval is based on the minimum bit error rate (BER) set by the system. min Obtained through repeated experiments.

[0079] (4) Update the coding interleaving depth parameter, perform interleaving coding according to the interleaving address mapping corresponding to the currently updated interleaving depth parameter, and send it to the atmospheric channel.

[0080] (5) Update the interleaving parameters of the deinterleaver in the decoder. Deinterleaving in decoding is the inverse operation of interleaving. The interleaver depth in the encoding and the deinterleaver depth in the decoding of the same frame information sequence are the same. The received sequence is temporarily buffered in ascending order of the mapping address during interleaving. Then, the data is read sequentially according to the address order after buffering to complete the deinterleaving process of the information sequence.

[0081] This invention addresses the significant impact of atmospheric turbulence on laser communication performance, incorporating channel estimation technology to estimate atmospheric channel conditions in real time. This provides data reference for selecting the interleaving depth in the encoder, allowing for informed adjustment of the interleaving depth to ensure the reliability of the communication system under adverse channel conditions. Compared to traditional Turbo code encoding and decoding methods with fixed interleaving depths, this invention offers superior error correction capabilities. While maintaining the system's target bit error rate, the interleaving depth can be reduced in favorable channel conditions to effectively control decoding delay. Furthermore, the decoder employs a pipelined parallel decoding structure, effectively reducing hardware resource consumption caused by decoding complexity, achieving a good balance between Turbo code error correction performance and decoding delay.

[0082] Finally, it should be noted that the above descriptions are merely preferred embodiments of the present invention and are not intended to limit the present invention. Although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art can still modify the technical solutions described in the foregoing embodiments or make equivalent substitutions for some of the technical features. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the protection scope of the present invention.

Claims

1. A Turbo code variable-depth interleaving encoding and decoding method for atmospheric laser communication, comprising an encoder, channel estimation, a decoder, and a variable-depth interleaving unit, characterized in that, Includes the following steps: S1, the sequence to be encoded is input into the Turbo encoder, and the sequence is encoded by the component encoder, interleaver and pruning multiplexing; S2, the encoded sequence is added to the pilot sequence and sent into the atmospheric channel through the optical antenna; S3, perform channel estimation using the known pilot sequence in the received sequence, and feed the channel estimation value back to the variable depth interleaver; S4, receive the sequence input decoder, after the iterative decoding terminates, perform a hard decision on the likelihood ratio output by the component decoder to obtain the final decoding result, and count the number of erroneous symbols to calculate the bit error rate; S5, the variable depth interleaver updates the coding parameters based on the feedback channel estimate, and changes the interleaver's interleaving depth to re-encode the next frame of the sequence to be sent; The variable-depth interleaver employs random interleaving with a variable interleaving depth. It adjusts the interleaving depth based on real-time channel status obtained from channel estimation feedback, achieving an optimal balance between bit error rate performance and decoding delay under different channel conditions while ensuring system performance. The specific workflow for interleaving depth switching includes the following steps: (1) During system initialization, the interleaver is started to calculate the interleaving address mapping under each interleaving depth parameter and cache it in advance; since the interleaver adopts random interleaving, the number of interleaving address combinations under the same interleaving depth parameter is not less than the number of frames of the information sequence to be sent; (2) The interleaving depth of the first frame information sequence encoder is selected as 512. The channel estimation is performed by low-delay decoding to obtain atmospheric channel parameters. Based on the feedback value of the initial channel estimation, the encoder interleaving depth of the subsequent transmitted information sequence is provided to provide accurate channel state information reference. (3) Use the received sequence to estimate the atmospheric channel state, and set the coding interleaving depth parameter of the next transmission sequence according to the threshold range of the signal-to-noise ratio estimate; reset the encoder interleaving depth according to the threshold range of the signal-to-noise ratio estimate. (4) Update the coding interleaving depth parameters, perform interleaving coding according to the interleaving address mapping corresponding to the currently updated interleaving depth parameters, and send it to the atmospheric channel; (5) Update the interleaving parameters of the deinterleaver in the decoder. Deinterleaving in decoding is the inverse operation of interleaving. The interleaver depth in the encoding and the deinterleaver depth in the decoding of the same frame information sequence are the same. The received sequence is temporarily buffered in ascending order of the mapping address during interleaving. Then, the data is read sequentially according to the address order after buffering to complete the deinterleaving process of the information sequence.

2. The Turbo code variable interleaving depth coding and decoding method for atmospheric laser communication according to claim 1, characterized in that: In S1, the Turbo code encoder and interleaver adopt a random interleaving method with variable interleaving depth. The initial state of the interleaving depth of the interleaver is set to a fixed value of 512, and the initial channel estimation is performed on the decoding of the received sequence with the minimum decoding delay.

3. The Turbo code variable interleaving depth coding and decoding method for atmospheric laser communication according to claim 1, characterized in that: In S3, the channel estimation adopts the least squares (LS) algorithm based on training symbols. The channel estimation is performed using comb pilot signals. Based on this, linear interpolation and quadratic polynomial interpolation are performed on the pilot positions to estimate the channel carrying data symbols. The estimated signal-to-noise ratio is then calculated from the channel estimation value.

4. The Turbo code variable interleaving depth coding and decoding method for atmospheric laser communication according to claim 1, characterized in that: The Turbo code decoder in S4 adopts a pipelined iterative decoding structure. During the iterative decoding process, the forward and backward metrics are optimized at certain recursion times. The optimization target value is obtained by pre-simulating the change in the data frame length transmitted in laser communication. The statistical values ​​of the forward and backward metrics are empirical reference values.

5. A Turbo code variable interleaving depth coding and decoding method for atmospheric laser communication according to claim 1, characterized in that: The Turbo code decoder in S4 uses the Log-MAP decoding algorithm to calculate the log-likelihood ratio (LLR) in parallel. The LLR is calculated in reverse recursion and is performed in parallel and synchronously with the backward metric calculation unit, which effectively reduces the hardware resource consumption of the decoding algorithm.

6. The Turbo code variable interleaving depth coding and decoding method for atmospheric laser communication according to claim 1, characterized in that: The variable depth interleaver in S5 has a variable interleaver depth. Under different atmospheric turbulence intensities, the atmospheric channel turbulence state is determined based on the channel estimation feedback value, and the encoder interleaver depth is reset according to the signal-to-noise ratio estimation threshold range.

Images