A combined degree distribution design and optimization method for online fountain codes

Abstract

This invention discloses a method for designing and optimizing a combined degree distribution for online fountain codes. The method comprises designing a combined degree distribution based on four basic degree distributions: ideal arc distribution, robust arc distribution, ripple-reduced degree distribution, and scale-free LT code. The method then optimizes the coefficients of the combined degree distribution using the ocean predator algorithm under a specific decoding success rate or equivalent symbol error rate, subject to a minimum decoding overhead constraint. This optimization method improves the performance of fountain codes by increasing the number of member degree distributions in the combined degree distribution and optimizing the member scaling coefficients using the minimum decoding overhead under a specific symbol error rate as the objective function. Simulation results demonstrate that the combined degree distribution designed and optimized using the proposed method can significantly improve the performance of fountain codes.

Description

Technical Field

[0001] The present invention relates to the field of unmanned aerial vehicle (UAV) communications, and in particular to, but not limited to, a combination degree distribution design and optimization method for an online fountain code. Background Art

[0002] With the development of 5G / 6G communication technologies, the demand for unmanned aerial vehicle (UAV) communications is increasing. UAVs typically operate in a variety of scenarios during flight missions. The mission environment is influenced by factors such as location, weather, altitude, speed, and external interference, resulting in significant randomness in the type and characteristics of UAV communication channels. Therefore, UAV communication systems must adapt to a variety of channels with varying characteristics during missions. Traditional channel coding, which typically has a fixed bit rate and employs a packet-by-packet feedback mechanism, struggles to achieve good performance under these conditions.

[0003] Fountain codes, on the other hand, feature adaptive bit rates and rarely require feedback or retransmissions. They can be applied at the transport or application layer to improve the performance of drone communications. Therefore, they can effectively complement channel coding, enhancing the reliability and stability of data communications. Furthermore, in 6G multipath transmission scenarios, fountain codes can ensure stable end-to-end transmission performance, showing broad potential for application in 6G ultra-reliable low-latency communication (URLLC) and secure communications.

[0004] Although fountain codes with a single degree distribution can adapt to changing channel environments to a certain extent, they cannot fully utilize the rate-free characteristics of fountain codes. Therefore, it is necessary to study the application of fountain codes with combined degree distributions in UAV communication scenarios. Summary of the Invention

[0005] In view of this, an embodiment of the present invention provides a combined degree distribution design and optimization method for online fountain codes, which at least solves the optimization bottleneck problem of single degree distribution fountain codes in adapting to changing channel conditions.

[0006] The technical solutions of the embodiments of the present invention are as follows:

[0007] An embodiment of the present invention provides a method for designing and optimizing the combined degree distribution of an online fountain code, the method comprising:

[0008] Based on four basic degree distributions, namely ideal arc distribution, robust arc distribution, ripple-reduced degree distribution and scale-free LT code, a combined degree distribution is designed. The coefficients of the combined degree distribution are optimized using the ocean predator algorithm under a specific decoding success rate or equivalent symbol error rate and with the minimum decoding overhead as the constraint.

[0009] Furthermore, the combination degree distribution includes a three-member combination degree distribution and a four-member combination degree distribution;

[0010] The combined degree distribution model of the three-member combined degree distribution is as follows:

[0011]

[0012] Where Ω1(x) represents the probability model of the three-member combination degree distribution, a1, a2, a3 are three proportional coefficients, μ(d) represents the robust arc distribution, θ(d) represents the ripple reduction degree distribution, and λ(d) represents the scale-free LT code;

[0013] The combined degree distribution model of the four-member combined degree distribution is as follows:

[0014]

[0015] Where Ω2(x) represents the probability model of the four-member combination degree distribution, a1, a2, a3, a4 represent four proportional coefficients, and ρ(d) represents the ideal arc distribution.

[0016] Furthermore, the ocean predator algorithm is used to optimize the coefficient of the combination degree distribution under a specific decoding success rate or an equivalent symbol error rate, with the minimum decoding overhead as a constraint, including: initializing a search agent population and setting a maximum number of iterations; wherein the position of each prey in the search agent population represents a set of combination degree distribution coefficients; for each prey in the search agent population, calculating the symbol error rate of the fountain code during LT decoding; if the symbol error rate of a prey is less than a set value, calculating the fitness of the prey; if the symbol error rate of a prey is greater than or equal to the set value, applying a penalty value to the prey; selecting the prey with the smallest fitness as the top predator, and obtaining the corresponding position; if the current number of iterations is less than the maximum number of iterations, updating the positions of all prey in the search agent population, and recalculating the symbol error rate of each prey for iteration; the optimization is completed when the current number of iterations is greater than or equal to the maximum number of iterations, and outputting the parameters of the target prey with the smallest current fitness; wherein the parameters of the target prey represent the optimized combination degree distribution coefficient.

[0017] Furthermore, in the step of calculating the fitness of the prey, the fitness is specifically calculated by the following formula:

[0018]

[0019] Where n is the number of samples required to calculate the fitness, T r represents the number of coded symbols received when the rth sample is successfully decoded;

[0020] The step of calculating the symbol error rate of all prey in the LT decoding process is specifically calculated by the following formula:

[0021]

[0022] Where p LT is the symbol error rate of the prey during LT decoding, k is the code length, l is the decoding step variable, l max is the total number of decoding steps, is the probability that decoding terminates when l=1.

[0023] Furthermore, the method further comprises: using Monte Carlo simulation to evaluate the performance of the combined degree distribution, including performance improvement evaluation obtained by increasing the number of members and performance evaluation of coefficient optimization.

[0024] The beneficial effects of the technical solutions provided in the embodiments of the present application include at least:

[0025] In an embodiment of the present invention, the performance of fountain codes is improved by increasing the number of member degree distributions in the combined degree distribution and optimizing the member scaling coefficients with the minimum decoding overhead under a specific symbol error rate as the objective function. BRIEF DESCRIPTION OF THE DRAWINGS

[0026] In order to more clearly illustrate the technical solutions in the embodiments of the present invention, the following briefly introduces the drawings required for describing the embodiments. Obviously, the drawings described below are only some embodiments of the present invention. Those skilled in the art can also derive other drawings based on these drawings without inventive work, among which:

[0027] Figure 1 A schematic diagram of the flow of the combined degree distribution design and optimization method of online fountain codes provided in an embodiment of the present invention;

[0028] Figure 2 Schematic diagram of ripple evolution of five degree distributions provided in an embodiment of the present invention;

[0029] Figure 3 A schematic diagram showing the relationship between decoding success rate and redundancy provided by an embodiment of the present invention;

[0030] Figure 4 A schematic diagram showing the probability distribution of the number of coded packets required for successful decoding of different code lengths provided by an embodiment of the present invention;

[0031] Figure 5 This is a schematic diagram showing the comparison of MPA algorithm optimization effects under different code length conditions provided by an embodiment of the present invention. DETAILED DESCRIPTION

[0032] In order to make the purpose, technical solutions and advantages of the embodiments of the present invention clearer, the technical solutions in the embodiments of the present invention will be clearly and completely described below in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are part of the embodiments of the present invention, rather than all the embodiments. The following embodiments are used to illustrate the present invention, but are not used to limit the scope of the present invention. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without making creative work are within the scope of protection of the present invention.

[0033] In the following description, reference is made to “some embodiments”, which describes a subset of all possible embodiments, but it will be understood that “some embodiments” may be the same subset or different subsets of all possible embodiments and may be combined with each other without conflict.

[0034] It should be pointed out that the terms "first\second\third" involved in the embodiments of the present invention are only used to distinguish similar objects and do not represent a specific ordering of the objects. It can be understood that "first\second\third" can be interchanged with a specific order or sequence where permitted, so that the embodiments of the present invention described here can be implemented in an order other than that illustrated or described here.

[0035] Those skilled in the art will understand that, unless otherwise defined, all terms used herein (including technical and scientific terms) have the same meaning as commonly understood by those skilled in the art in the art to which the embodiments of the present invention pertain. It should also be understood that terms such as those defined in common dictionaries should be understood to have meanings consistent with those in the context of the prior art, and will not be interpreted in an idealized or overly formal sense unless specifically defined as herein.

[0036] Figure 1 A schematic diagram of a flow chart of a combined degree distribution design and optimization method for online fountain codes provided by an embodiment of the present invention is shown in FIG. Figure 1 As shown, the method comprises at least the following steps:

[0037] Step S110 , designing a combined degree distribution based on four basic degree distributions: ideal arc distribution, robust arc distribution, ripple-reduced degree distribution, and scale-free LT code.

[0038] Here, the ideal soliton distribution (ISD) is the degree distribution with the lowest expected overhead. The present invention uses the ideal soliton distribution to reduce the risk of decoding failure due to ripple reduction to zero. The probability expression of the ideal soliton distribution is as follows:

[0039]

[0040] Where ρ(d) is the probability of the ideal arc distribution, k is the number of source data packets, i.e., the code length, and d is the encoding packet value.

[0041] Although the ideal arc distribution has the lowest expected overhead in theory, in actual communication system environments, the performance of the ideal arc distribution is not as expected. Robust Soliton Distribution (RSD) can improve the performance of the ideal arc distribution. Therefore, the present invention also adopts the robust soliton distribution to make up for the shortcomings of the ideal arc distribution. The probability expression of the robust soliton distribution is as follows:

[0042]

[0043] Where μ(d) is the probability of robust arc distribution, let is the expected ripple size of the decoding process, then the robustness factor τ(d) can be defined as:

[0044]

[0045] Where k is the number of source packets, i.e., the code length, c is a non-zero positive constant, δ is the allowed probability of decoding failure, and R represents the expected ripple size of the decoding process.

[0046] The Damped Ripple Soliton Distribution (DRSD) is designed to target the ripple size during decoding, and its performance is better than the robust arc distribution and the ideal arc distribution. Using the ripple-reduced distribution in the combined degree distribution can reduce the redundancy required for successful decoding by maintaining an appropriate ripple size during decoding. The expression of the ripple-reduced distribution is as follows:

[0047]

[0048] Where θ(d) is the probability of the ripple reduction distribution, n is the normalization factor, Q is the ripple size of degree 1, and d represents the coding packet value.

[0049] Scale-free LT codes are a type of degree distribution designed based on the concept of scale-free networks. The scale-free property of scale-free LT codes can be exploited in combined degree distributions to achieve lower coding complexity and mitigate the adverse effects associated with random packet loss. Its expression is as follows:

[0050]

[0051] Where λ(d) is the probability of the scale-free LT code, P1 represents the proportion of coded symbols with degree 1, A is the normalization coefficient, and γ is the characteristic coefficient.

[0052] In the embodiment of the present invention, the number of member degree distributions is expanded. The combined degree distribution model consisting of n member degree distributions is defined as follows:

[0053]

[0054] where a1, a2, a3…a n are the proportionality coefficients, λ1(d),λ2(d),λ3(d)…λ n (d) is the membership degree distribution used in the combined degree distribution.

[0055] It's important to note that the performance of fountain codes depends heavily on the design of their degree distribution. This determines the probability of generating encoded symbols, which in turn affects encoding and decoding performance. A well-designed degree distribution can reduce the number of required encoding symbols while maintaining a low error rate, thereby lowering decoding overhead.

[0056] By selecting appropriate degree distributions and combining them in appropriate quantities, embodiments of the present invention can design new combined degree distributions. This combination leverages the strengths of various degree distributions, achieving optimal performance under varying channel conditions. Different degree distributions may have varying impacts on coding coverage, decoding complexity, and decoding overhead.

[0057] Step S120 , using the Ocean Predator algorithm to optimize the coefficients of the combined degree distribution under a specific decoding success rate or equivalent symbol error rate and with a minimum decoding overhead as a constraint.

[0058] Here, the member degree distributions of a combined degree distribution indirectly influence the performance of the combined degree distribution by affecting the structure of the overall degree distribution. As the number of member degree distributions increases, optimizing the combined coefficients becomes challenging. This paper proposes a method that uses the Marine Predator Algorithm (MPA) to optimize the combined degree distribution coefficients under a specific decoding success rate, subject to a minimum decoding overhead constraint. The MPA exhibits strong robustness and excellent convergence performance.

[0059] During the optimization process, an objective function must be determined to evaluate the performance of fountain codes under different scaling factors. This objective function is related to the symbol error rate (SER) and decoding overhead. By minimizing this objective function, the optimal scaling factor can be found. To find the optimal scaling factor, various optimization algorithms can be applied, such as differential evolution and sequential quadratic programming (SQP). These algorithms can find the optimal solution within a complex search space.

[0060] In an embodiment of the present invention, the performance of fountain codes is improved by increasing the number of member degree distributions in the combined degree distribution and optimizing the member scaling coefficients with the minimum decoding overhead under a specific symbol error rate as the objective function.

[0061] In some embodiments, the combination degree distribution includes a three-member combination degree distribution and a four-member combination degree distribution; wherein the combination degree distribution model of the three-member combination degree distribution is as follows:

[0062]

[0063] Where Ω1(x) represents the probability model of the three-member combination degree distribution, a1, a2, and a3 are three scaling coefficients, μ(d) represents the probability distribution of the robust arc distribution, θ(d) represents the probability distribution of the ripple reduction degree distribution, and λ(d) represents the probability distribution of the scale-free LT code.

[0064] The combined degree distribution model of the four-member combined degree distribution is as follows:

[0065]

[0066] Where Ω2(x) represents the probability model of the four-member combination degree distribution, a1, a2, a3, a4 represent four proportional coefficients, and ρ(d) represents the probability distribution of the ideal arc distribution.

[0067] To study the performance of combined degree distributions, the present invention designs two combined degree distributions. The first is a three-member combined degree distribution (TMCD), which uses a robust arc distribution, a ripple-reduced degree distribution, and a scale-free LT code, as defined in Formula (7). The other is a four-member combined degree distribution (FMCD), which uses an ideal arc distribution, a robust arc distribution, a ripple-reduced degree distribution, and a scale-free LT code, as defined in Formula (8).

[0068] In some embodiments, the above step S120 of "optimizing the coefficients of the combined degree distribution using the ocean predator algorithm under a specific decoding success rate or equivalent symbol error rate and subject to a minimum decoding overhead constraint" can be implemented by the following steps:

[0069] S121, initializing a search agent population and setting a maximum number of iterations; wherein the position of each prey in the search agent population represents a set of combined degree distribution coefficients;

[0070] S122, calculating the symbol error rate of the fountain code during the LT decoding process for each prey in the search agent population;

[0071] S123, if the symbol error rate of a certain prey is less than a set value, calculating the fitness of the prey;

[0072] S124, if the symbol error rate of a certain prey is greater than or equal to the set value, impose a penalty value on the prey;

[0073] S125, select the prey with the smallest fitness as the top predator and obtain the corresponding position;

[0074] S126, if the current number of iterations is less than the maximum number of iterations, updating the positions of all prey in the search agent population, and recalculating the symbol error rate of each prey to iterate;

[0075] S127, when the current number of iterations is greater than or equal to the maximum number of iterations, the optimization is completed, and the parameters of the target prey with the minimum current fitness are output; wherein the parameters of the target prey represent the optimized combination degree distribution coefficient.

[0076] Here, the goal of MPA optimization is to minimize the decoding overhead under a specific decoding success rate (or equivalent symbol error rate). The optimization problem can be expressed as min(f(Ω)) ,in f(Ω) is a function for calculating decoding overhead, and its expression is shown in formula (9). The specific process of the coefficient optimization method of the combined degree distribution is shown in Table 1:

[0077] Table 1 Coefficient optimization method of combination degree distribution

[0078]

[0079] In this way, the design and optimization of the combined degree distribution can significantly improve the performance of fountain codes under various channel conditions, especially in environments with high error rates or high delays. This optimization method provides an effective strategy for the application of fountain codes in practical communication systems.

[0080] In some embodiments, the step of calculating the fitness of the prey is specifically calculated by the following formula:

[0081]

[0082] Where n is the number of samples required to calculate the fitness, T r represents the number of coded symbols received when the rth sample is successfully decoded;

[0083] The step of calculating the symbol error rate of all prey in the LT decoding process is specifically calculated by the following formula:

[0084]

[0085] Where p LT is the symbol error rate of the prey during LT decoding, k is the code length, l is the decoding step variable, l max is the total number of decoding steps, is the probability that decoding terminates when l=1.

[0086] Here, when optimizing the coefficients, it is necessary to calculate the symbol error rate (SER) during the LT decoding process. A related art method for calculating the SER, called GERS, uses a classical probability model, but its computational complexity is relatively high. Another SER performance analysis scheme for LT codes, called DDEA, uses the calculation method shown in Equation (10) and has lower complexity than GERS. Therefore, the present invention uses DDEA to calculate the SER for LT encoding and decoding.

[0087] In some embodiments, the method further comprises: using Monte Carlo simulation to evaluate the performance of the combined degree distribution, including performance improvement evaluation obtained by increasing the number of members and performance evaluation of coefficient optimization.

[0088] Here, through simulation and experiments, we can evaluate the performance of optimized fountain codes under different channel conditions. Performance indicators may include bit error rate, decoding overhead, encoding and decoding time, etc.

[0089] During BP decoding, a sufficient number of degree-1 packets should be available early on to facilitate the decoding process. However, if the initial ripple size is too large, this means that there are too many low-degree encoded packets, making it difficult to cover all source packets. If the ripple size during decoding is too small, the decoding process may fail. Therefore, the ripple size should be reduced at an appropriate rate during the decoding process and maintained above a certain level.

[0090] Ripple evolution is used to evaluate the performance of different types of degree distributions. Figure 2 The evolution of ripple size of various degree distributions is shown when k=1000 and redundancy α is 0.05. Figure 2 As shown in FIG, the TMCD and FMCD distributions proposed in the embodiment of the present invention achieve a continuous reduction in the ripple size during the decoding process and keep the fluctuation within a relatively small range. This reduces the probability of decoding failure due to the lack of data packets with a degree of 1 during the decoding process. Figure 2 In the ripple size analysis, DRS outperforms the TMCD and FMCD proposed in this paper. However, DRS's decoding success rate performance is not superior to the proposed combined degree distribution. This is because DRS focuses on designing the desired ripple size during decoding, while neglecting the optimization of the degree distribution itself.

[0091] First, the performance improvement obtained by increasing the number of members is evaluated as follows:

[0092] Assuming ideal channel conditions, the simulations are performed using the degree distributions listed in Table 2 for comparison. These degree distributions represent a single degree distribution and combined degree distributions with two, three, and four members, respectively. The parameter settings for DRSD, SF-LT, and RSF-LT (a combination of RSD and SF-LT) are based on empirical conclusions from existing research. The degree distributions TMCD and FMCD proposed in this paper are given by Equations (7) and (8), respectively.

[0093] Table 2 Detailed parameters of degree distribution

[0094]

[0095]

[0096] Simulations are performed with code lengths of k = 500, k = 1000, and k = 2000 to evaluate the performance of combined degree distributions with different numbers of members.

[0097] Figure 3 The relationship between decoding success rate and decoding redundancy overhead is demonstrated. The decoding success rate of LT codes under low redundancy conditions improves with increasing the number of member degree distributions. Specifically, compared to the two-member RSF-LT, the three-member TMCD degree distribution improves the decoding success rate by 20% when k = 1000 and redundancy is 0.1. Furthermore, under the same redundancy conditions, the four-member FMCD degree distribution improves the decoding success rate by 12% when k = 1000 and 18% when k = 2000 compared to TMCD. Therefore, increasing the number of member degree distributions in a combined degree distribution can significantly improve the probability of successful decoding.

[0098] The comparison of the average decoding overhead and average degree value required for successful decoding under different degree distributions is shown in Table 3. It can be seen that as the number of member degree distributions increases, the decoding overhead required for successful decoding gradually decreases, but the average degree value increases slightly.

[0099] Table 3 Average overhead required to decode all input symbols and average degree value of each encoded symbol

[0100]

[0101] Figure 4 The distribution of the number of code packets required for successful decoding under four degree distributions is shown. As the number of member degree distributions increases, the number of code packets required for successful decoding decreases, and the distribution becomes more concentrated. This demonstrates that combining degree distributions can enhance the performance of LT codes and reduce the overall decoding overhead.

[0102] Then, the performance evaluation process for coefficient optimization is as follows:

[0103] The above experiments show that increasing the number of member degree distributions is an effective way to improve the performance of the combined degree distribution. In order to verify the effectiveness of the algorithm for increasing the number of members through MPA optimization of the combined degree distribution coefficient, the TMCD performance before and after MPA optimization is compared. The comparison results are shown in Figure 2. Figure 5 As shown, the optimized proportional coefficient is shown in Table 4:

[0104] Table 4 Optimized proportional coefficients under different code lengths

[0105] RSD DRSD SF-LT k＝500 0.5637 0.2495 0.1868 k＝1000 0.5215 0.1969 0.2815 k＝2000 0.4255 0.4793 0.0952

[0106] like Figure 5 As shown, when k = 2000, after MPA optimization, the performance of the combined degree distribution improves compared to the unoptimized version at a redundancy of 0.05. At a redundancy of 0.1, the probability of successful decoding increases by 18% after MPA optimization. At a redundancy of 0.15, the decoding success probability of the MPA-optimized TMCD exceeds 99%, while the unoptimized TMCD requires a redundancy of 0.35 to achieve similar performance. At this point, the MPA-optimized TMCD can reduce decoding overhead by approximately 20%.

[0107] Table 5 Comparison of degree distribution effects before and after MPA optimization under different code lengths

[0108]

[0109] Table 5 shows the characteristics of the MPA-optimized TMCD. Compared to the unoptimized version, the MPA-optimized TMCD reduces decoding overhead by 11.3% to 23.6%. The optimized TMCD slightly increases the average degree value, which reduces the probability of not covering all source packets during the encoding process.

[0110] This paper investigates methods for designing combined degree distributions from various complementary single degree distributions and evaluates their impact on fountain code performance. To determine the optimal coefficients for the combined degree distributions, the Marine Predator Algorithm (MPA) is employed to optimize these coefficients with the goal of minimizing decoding overhead at a specific symbol error rate. Simulation results show that increasing the number of member degree distributions in a combined degree distribution and optimizing their coefficients using MPA can improve decoding success rates by 10%-20%. Furthermore, simulation results demonstrate that combined degree distributions with more members exhibit more stable performance and significantly higher overall decoding success rates.

[0111] It should be understood that "one embodiment" or "an embodiment" mentioned throughout the specification means that the specific features, structures or characteristics related to the embodiment are included in at least one embodiment of the present invention. Therefore, "in one embodiment" or "in an embodiment" appearing throughout the specification does not necessarily refer to the same embodiment. In addition, these specific features, structures or characteristics can be combined in one or more embodiments in any suitable manner. It should be understood that in various embodiments of the present invention, the size of the serial numbers of the above-mentioned processes does not mean the order of execution. The execution order of each process should be determined by its function and internal logic, and should not constitute any limitation on the implementation process of the embodiment of the present invention. The serial numbers of the above-mentioned embodiments of the present invention are for description only and do not represent the advantages and disadvantages of the embodiments.

[0112] It should be noted that, in the present invention, the terms "comprises," "includes," or any other variations thereof are intended to encompass non-exclusive inclusion, such that a process, method, article, or apparatus comprising a series of elements includes not only those elements but also other elements not explicitly listed, or elements inherent to such process, method, article, or apparatus. In the absence of further limitations, an element defined by the phrase "comprising a ..." does not exclude the presence of other identical elements in the process, method, article, or apparatus comprising the element.

[0113] In the several embodiments provided herein, it should be understood that the disclosed methods can be implemented in other ways. The methods disclosed in the several method embodiments provided herein can be combined arbitrarily, unless they conflict, to produce new method embodiments. The features disclosed in the several method embodiments provided herein can be combined arbitrarily, unless they conflict, to produce new method embodiments.

[0114] The above description is merely an embodiment of the present invention, but the scope of protection of the present invention is not limited thereto. Any modifications or substitutions that can be easily conceived by a person skilled in the art within the technical scope disclosed in the present invention should be included in the scope of protection of the present invention. Therefore, the scope of protection of the present invention should be based on the scope of protection of the claims.

Claims

1. A method for designing and optimizing the combined degree distribution of online fountain codes, characterized in that: include: Based on four basic degree distributions, namely ideal arc distribution, robust arc distribution, ripple-reduced degree distribution and scale-free LT code, a composite degree distribution is designed. With the help of the Ocean Predator algorithm, using a specific decoding success rate or equivalent symbol error probability as a constraint and minimizing the decoding overhead as the optimization goal, the coefficients of each member in the combined degree distribution are optimized through the following process: Initialize the search agent population and set the maximum number of iterations; wherein the position of each prey in the search agent population represents a set of combination degree distribution coefficients; for each prey in the search agent population, calculate the symbol error rate of the fountain code during LT decoding; if the symbol error rate of a prey is less than a set value, calculate the fitness of the prey; if the symbol error rate of a prey is greater than or equal to the set value, impose a penalty value on the prey; select the prey with the smallest fitness as the top predator and obtain the corresponding position; if the current number of iterations is less than the maximum number of iterations, update the positions of all prey in the search agent population, recalculate the symbol error rate of each prey and iterate; when the current number of iterations is greater than or equal to the maximum number of iterations, the optimization is completed, and the parameters of the target prey with the smallest current fitness are output; wherein the parameters of the target prey represent the optimized combination degree distribution coefficient; The step of calculating the fitness of the prey is specifically calculated by the following formula: ； Where n is the calculated fitness The number of samples required, represents the number of coded symbols received when the rth sample is successfully decoded; The step of calculating the symbol error rate of all prey in the LT decoding process is specifically calculated by the following formula: ； Where, is the symbol error rate of the prey during LT decoding, k is the code length, l is the decoding step variable, l max is the total number of decoding steps, is the probability that decoding terminates when l=1.

2. The method according to claim 1, characterized in that The combination degree distribution includes a three-member combination degree distribution and a four-member combination degree distribution; The combined degree distribution model of the three-member combined degree distribution is as follows: ； Where, represents the probability model of the degree distribution of three-member combinations, are three proportional coefficients, represents the probability distribution of the robust arc distribution, represents the probability distribution of the ripple reduction degree distribution, represents the probability distribution of scale-free LT codes; The combined degree distribution model of the four-member combined degree distribution is as follows: ； Where, represents the probability model of the degree distribution of four-member combinations, Represents four proportional coefficients, The probability distribution representing the ideal arc distribution.

3. The method according to claim 1 or 2, characterized in that The method further comprises: Monte Carlo simulation is used to evaluate the performance of the combined degree distribution, including the performance improvement evaluation obtained by increasing the number of members and the performance evaluation of coefficient optimization.

Images