Joint degree distribution design and optimization method for fountain codes

Abstract

The invention provides a joint degree distribution design and optimization method for fountain codes. The method comprises the steps of obtaining to-be-transmitted test data in a current scene; using to-be-transmitted test data to form an optimized joint degree distribution model by taking the minimum average decoding overhead as an optimization target under the joint degree distribution model; the joint degree distribution model is obtained by joint construction of probability distribution of modified Poisson distribution, probability distribution of ideal soliton distribution and probability distribution of sliding robust soliton distribution; based on the optimized joint degree distribution model, performing global optimal solution search processing on a to-be-optimized weight coefficient in the optimized joint degree distribution model by adopting a Monte Carlo simulation method and a double-layer nested search method to obtain a global optimal weight; and substituting the global optimal weight into the joint degree distribution model to obtain final fountain code joint degree distribution in the current scene. Therefore, the rationality of the degree distribution structure, the consistency of the optimization target and the system performance and the improvement of the global optimality of the design method are realized.

Description

Technical Field

[0001] This invention relates to the field of digital fountain code design, and more specifically to a method for designing and optimizing the joint degree distribution of fountain codes. Background Technology

[0002] Fountain codes, also known as digital fountain codes, are a high-efficiency, rate-free coding scheme for binary deletion channels. The core idea is that the transmitter can encode a finite number of original input symbols into a theoretically infinite number of encoded symbols for transmission; the receiver, however, does not need to receive specific symbols, but only needs to successfully collect slightly more encoded symbols than the original symbols to reconstruct the original data with a high probability. This characteristic makes it significantly advantageous in scenarios such as large-scale data distribution, satellite communication, and the Internet of Things (IoT), improving transmission reliability while reducing feedback overhead. The core of fountain code performance lies in its degree distribution design. The degree distribution determines how many original symbols are XORed to generate each encoded symbol, directly affecting the initiation and propagation efficiency of the decoding process, as well as the final decoding success rate and overhead. Therefore, designing a degree distribution that can simultaneously guarantee high decoding success rate and low decoding overhead under complex channel conditions is a key technical issue for improving the overall efficiency of fountain code systems.

[0003] To address the aforementioned issues, existing technologies propose combining multiple single-degree distributions to leverage their respective strengths and compensate for their weaknesses. One example is a dual-proportional-coefficient joint degree distribution design method. This method selects three single-degree distributions with different characteristics: Modified Binary Exponential Distribution (Modified-BED), used to generate a large number of low-degree value (mainly degree 1) encoded packets to ensure smooth startup of the decoding process; Ideal Soliton Distribution (ISD), utilizing its theoretically high iterative decoding efficiency; and Moved Robust Soliton Distribution (Moved-RSD), used to provide sufficient high-degree value encoded packets to ensure coverage of the original data. By introducing two proportional coefficients to balance the weights of these three distributions in the joint distribution, a new joint degree distribution is constructed. The optimization process of this scheme typically relies on Monte Carlo simulations, using decoding success rate as the primary observation indicator, and determining the final distribution parameters by searching for the proportional coefficient combination that maximizes the success rate.

[0004] However, this existing technical solution still has several shortcomings. First, the selected combination of degree distributions may not be optimal in terms of composition. For example, the probability quality of the modified binary exponential distribution is overly concentrated in degrees 1 and 2. Although this is beneficial for decoding initiation, it may lead to an overabundance of low-degree encoded symbols and a relative shortage of medium-degree symbols, thus disrupting the smoothness and diversity of the degree distribution. Second, there is a disconnect between its optimization objective and the ultimate performance indicators of the system. This scheme is mainly optimization-oriented towards maximizing the decoding success rate, which may result in a distribution that, while having acceptable reliability, does not have optimal transmission efficiency. Finally, the design process of existing methods still relies heavily on empirical adjustments and local searches, making it difficult to guarantee globally optimal performance. Summary of the Invention

[0005] To address the aforementioned problems in the prior art, this invention provides a method for designing and optimizing the joint degree distribution of fountain codes.

[0006] The technical problem to be solved by this invention is achieved through the following technical solution: In a first aspect, the present invention provides a method for designing and optimizing the joint degree distribution of fountain codes, including: Obtain the test data to be transmitted in the current scenario; Using the test data to be transmitted, an optimized joint degree distribution model is formed with minimizing the average decoding overhead as the optimization objective under the joint degree distribution model. The joint degree distribution model is jointly constructed by the probability distributions of the modified Poisson distribution, the ideal soliton distribution, and the sliding robust soliton distribution. Based on the optimized joint degree distribution model, the Monte Carlo simulation method and the double-layer nested search method are used to perform global optimal solution search on the weight coefficients to be optimized in the optimized joint degree distribution model, so as to obtain the global optimal weight. Substituting the globally optimal weights into the joint degree distribution model yields the final fountain code joint degree distribution for the current scenario.

[0007] Alternatively, the joint degree distribution model can be expressed as: ; in, Indicates the fountain code encoding packet value The corresponding joint degree distribution model, This represents the first weight coefficient to be optimized. This represents the second weight coefficient to be optimized. Indicates the fountain code encoding packet value The corresponding probability distribution of the modified Poisson distribution, Indicates the fountain code encoding packet value The probability distribution of the corresponding ideal soliton distribution, Indicates the fountain code encoding packet value The probability distribution of the corresponding sliding robust soliton distribution.

[0008] Alternatively, the probability distribution of the modified Poisson distribution can be expressed as: ; ; in, This indicates the corrected Poisson distribution parameters. Indicates the number of packets containing test data to be transmitted; The probability distribution of the ideal soliton distribution is expressed as: ; The probability distribution of the sliding robust soliton distribution is expressed as: ; ; ; ; in, The first parameter represents the probability distribution corresponding to the sliding robust soliton distribution. The second parameter represents the probability distribution corresponding to the sliding robust soliton distribution. The third parameter represents the probability distribution corresponding to the sliding robust soliton distribution. The fourth parameter represents the probability distribution corresponding to the sliding robust soliton distribution.

[0009] Optionally, based on the optimized joint degree distribution model, a Monte Carlo simulation method and a two-level nested search method are used to perform a global optimal solution search on the weight coefficients to be optimized in the optimized joint degree distribution model, to obtain the globally optimal weights, including: S201, The sending end obtains initialization information; S202. The sending end determines the current second weight coefficient to be optimized based on the initialization information, fixes the second weight coefficient to be optimized, and iterates through the first weight coefficient to be optimized to obtain multiple sets of weight coefficient pairs. S203. The transmitting end performs Monte Carlo simulation processing on each weight coefficient pair in the multiple weight coefficient pairs a preset number of times, and generates multiple intermediate data packets based on the optimized joint degree distribution model during the Monte Carlo simulation processing, and sends the multiple intermediate data packets to the receiving end. S204. The receiving end receives multiple intermediate data packets and uses the BP algorithm to perform decoding processing to obtain the average decoding overhead corresponding to all weight coefficients. S205. The weight coefficient pair that minimizes the average decoding cost is taken as the local optimal weight coefficient. S206. Update the current second weight coefficient to be optimized using a preset search step size, and use the new second weight coefficient to be optimized as the current second weight coefficient to be optimized in S202. S207. Repeat S202-S206 to obtain all local optimal weight coefficients and corresponding local optimal average decoding costs within the search range; S207. The local optimal weight coefficient corresponding to the minimum local optimal average decoding cost is taken as the global optimal weight.

[0010] Optionally, the initialization information includes: the number of packets transmitting test data, the parameters corresponding to the probability distribution of the sliding robust soliton distribution, the parameters of the corrected Poisson distribution, the number of Monte Carlo simulations, the search range of the second weight coefficient to be optimized, and the preset search step size.

[0011] Optionally, after substituting the optimal weights into the joint degree distribution model to obtain the final fountain code joint degree distribution in the current scenario, the model further includes: The sending end obtains the actual transmission data in the current scenario; divides the actual transmission data into multiple transmission data packets; uses the final fountain code joint degree distribution to determine the transmission degree value corresponding to multiple transmission data packets; and uses the transmission degree value to select the corresponding multiple transmission data packets for XOR operation to generate the transmission encoded packet. The receiving end decodes the data using the transmitted encoded packets to obtain the decoded information.

[0012] Optionally, the receiving end uses the transmitted encoded packets to decode the data, obtaining decoded information including: The receiving end receives the transmitted encoded packet and uses the BP algorithm to decode the data in the transmitted encoded packet to obtain the decoded information.

[0013] Secondly, the present invention provides a joint degree distribution design and optimization device for fountain codes, which includes: an acquisition unit, a model construction unit, an optimization processing unit, and an output unit. The acquisition unit is used to acquire the test data to be transmitted in the current scenario. The model building unit is used to form an optimized joint degree distribution model by using the test data to be transmitted and minimizing the average decoding overhead under the joint degree distribution model. The joint degree distribution model is jointly constructed by the probability distributions of the modified Poisson distribution, the ideal soliton distribution, and the sliding robust soliton distribution. The optimization processing unit is used to perform global optimal solution search processing on the weight coefficients to be optimized in the optimized joint degree distribution model based on the Monte Carlo simulation method and the double-layer nested search method to obtain the global optimal weight. The output unit is used to substitute the globally optimal weights into the joint degree distribution model to obtain the final fountain code joint degree distribution in the current scenario.

[0014] Thirdly, the present invention provides a fountain code degree distribution design and optimization device, comprising: a processor, a storage medium and a bus, wherein the storage medium stores machine-readable instructions executable by the processor, and when the fountain code degree distribution design and optimization device is running, the processor communicates with the storage medium via the bus, and the processor executes the machine-readable instructions to perform the steps of the fountain code degree distribution design and optimization method of the first aspect described above.

[0015] This invention provides a method for designing and optimizing the joint degree distribution of fountain codes, comprising: acquiring test data to be transmitted in the current scenario; using the test data to be transmitted, forming an optimized joint degree distribution model with minimizing the average decoding overhead as the optimization objective under the joint degree distribution model; the joint degree distribution model is jointly constructed using the probability distributions of the modified Poisson distribution, the ideal soliton distribution, and the sliding robust soliton distribution; based on the optimized joint degree distribution model, using Monte Carlo simulation and a double-layer nested search method to perform a global optimal solution search on the weight coefficients to be optimized in the optimized joint degree distribution model to obtain the globally optimal weights; substituting the globally optimal weights into the joint degree distribution model to obtain the final joint degree distribution of the fountain codes in the current scenario. In this invention, firstly, a model is jointly constructed using the modified Poisson distribution, the ideal soliton distribution, and the sliding robust soliton distribution, because the three types of distributions have complementary advantages in low-degree smoothness, medium-degree diversity, and high-degree robustness, respectively. By dynamically adjusting the proportions of the three factors (degree distribution, degree distribution, and system optimization) using weighting coefficients, the problem of excessive concentration of probability quality at low degree values ​​is avoided, thereby enhancing the smoothness and diversity of the degree distribution and improving the balance of the encoded symbol composition. Then, minimizing the average decoding overhead is used as the optimization objective, directly linking transmission efficiency and decoding reliability. Compared to simply pursuing decoding success rate, this objective simultaneously balances the probability of successful decoding with the required number of encoded symbols, prompting the degree distribution design to improve transmission efficiency while ensuring reliability, overcoming the problem of the original objective being disconnected from the ultimate performance indicator. Finally, Monte Carlo simulation of actual transmission scenarios is used, combined with a two-layer nested search to systematically optimize the weighting coefficients, replacing the original local search method that relied on experience-based adjustments. This mechanism allows for extensive exploration within the parameter space, significantly increasing the probability of finding the global optimum, thus providing a theoretically global optimization guarantee for the degree distribution performance. Based on this, improvements are achieved in the rationality of the degree distribution structure, the consistency between the optimization objective and system performance, and the global optimality of the design method, resulting in systematic improvements in decoding reliability, transmission efficiency, and engineering applicability.

[0016] The present invention will be further described in detail below with reference to the accompanying drawings and embodiments. Attached Figure Description

[0017] Figure 1 A flowchart illustrating a method for designing and optimizing the joint degree distribution of fountain codes, provided in an embodiment of the present invention; Figure 2 An exemplary flowchart of the process for solving the globally optimal weights based on the method of the present invention is shown; Figure 3 An illustrative probability distribution diagram of the joint degree distribution model provided by the present invention is shown; Figure 4 An exemplary comparison of the decoding overhead of the method of the present invention with other comparative algorithms is shown; Figure 5 A schematic diagram of a joint degree distribution design and optimization device for fountain codes provided in an embodiment of the present invention; Figure 6 This is a schematic diagram of a joint degree distribution design and optimization device for fountain codes provided in an embodiment of the present invention. Detailed Implementation

[0018] The present invention will be further described in detail below with reference to specific embodiments, but the implementation of the present invention is not limited thereto.

[0019] To achieve systematic improvements in the decoding reliability, transmission efficiency, and engineering applicability of fountain codes, this invention provides a method for designing and optimizing the joint degree distribution of fountain codes. Figure 1 This is a flowchart illustrating a method for designing and optimizing the joint degree distribution of fountain codes according to an embodiment of the present invention, as shown below. Figure 1 As shown, it includes: S101. Obtain the test data to be transmitted in the current scenario.

[0020] S102. Using the test data to be transmitted, an optimized joint degree distribution model is formed with minimizing the average decoding overhead as the optimization objective under the joint degree distribution model.

[0021] The joint degree distribution model is constructed by jointly using the probability distributions of the modified Poisson distribution, the ideal soliton distribution, and the sliding robust soliton distribution.

[0022] Alternatively, the joint degree distribution model MPMRID (Dual-Proportion Combined Degree Distribution Based on MPD, MRSD and ISD) can be expressed as: ; in, Indicates the fountain code encoding packet value The corresponding joint degree distribution model, This represents the first weight coefficient to be optimized. This represents the second weight coefficient to be optimized. Indicates the fountain code encoding packet value The corresponding probability distribution of the modified Poisson distribution, Indicates the fountain code encoding packet value The probability distribution of the corresponding ideal soliton distribution, Indicates the fountain code encoding packet value The probability distribution of the corresponding sliding robust soliton distribution.

[0023] In this invention, it is necessary to satisfy the following: , and .

[0024] Alternatively, the probability distribution of the modified Poisson distribution can be expressed as: ; ; in, This indicates the corrected Poisson distribution parameters. Indicates the number of packets containing test data to be transmitted; The probability distribution of the ideal soliton distribution is expressed as: ; The probability distribution of the sliding robust soliton distribution is expressed as: ; ; ; ; in, The first parameter represents the probability distribution corresponding to the sliding robust soliton distribution. The second parameter represents the probability distribution corresponding to the sliding robust soliton distribution. The third parameter represents the probability distribution corresponding to the sliding robust soliton distribution. The fourth parameter represents the probability distribution corresponding to the sliding robust soliton distribution.

[0025] The joint degree distribution model employed in this invention has the core advantage of systematically addressing the performance limitations of existing single or combined distributions through a design of complementary structures and adjustable weights. Specifically, the modified Poisson distribution ensures rapid decoding startup, the ideal soliton distribution maintains efficient and stable progress, and the sliding robust soliton distribution guarantees reliable final convergence. The three are organically combined to form a complete chain covering the entire decoding lifecycle. This joint structure not only enhances the smoothness and diversity of degree values ​​and avoids excessive concentration of probability quality, but also, through optimizing the low-dimensional interface of weight coefficients, enables the distribution form to flexibly and automatically adapt to different scenarios. It directly targets the ultimate performance goal of "minimizing decoding overhead" for global optimization, thereby achieving synergistic improvements in decoding reliability, transmission efficiency, and engineering adaptability.

[0026] S103. Based on the optimized joint degree distribution model, the Monte Carlo simulation method and the double-layer nested search method are used to perform global optimal solution search on the weight coefficients to be optimized in the optimized joint degree distribution model to obtain the global optimal weight.

[0027] Optionally, S103 may specifically include: S201, The sending end obtains initialization information; S202. The sending end determines the current second weight coefficient to be optimized based on the initialization information, fixes the second weight coefficient to be optimized, and iterates through the first weight coefficient to be optimized to obtain multiple sets of weight coefficient pairs. S203. The transmitting end performs Monte Carlo simulation processing on each weight coefficient pair in the multiple weight coefficient pairs a preset number of times, and generates multiple intermediate data packets based on the optimized joint degree distribution model during the Monte Carlo simulation processing, and sends the multiple intermediate data packets to the receiving end. S204. The receiving end receives multiple intermediate data packets and uses the BP algorithm to perform decoding processing to obtain the average decoding overhead corresponding to all weight coefficients. S205. The weight coefficient pair that minimizes the average decoding cost is taken as the local optimal weight coefficient. S206. Update the current second weight coefficient to be optimized using a preset search step size, and use the new second weight coefficient to be optimized as the current second weight coefficient to be optimized in S202. S207. Repeat S202-S206 to obtain all local optimal weight coefficients and corresponding local optimal average decoding costs within the search range; S207. The local optimal weight coefficient corresponding to the minimum local optimal average decoding cost is taken as the global optimal weight.

[0028] The sending end is typically a device or server that possesses complete data and needs to efficiently and reliably distribute it to one or more receivers. Its core characteristic is having sufficient computing power to execute the fountain code encoding algorithm. In some typical scenarios, the sending end can be a video streaming server, such as IPTV or live streaming platforms, using fountain codes to combat network fluctuations and ensure smooth viewing for users.

[0029] The receiving end is typically a terminal device that needs to acquire data but is located in an unstable network environment or is unable to perform complex interactions with the sending end, such as mobile terminals and consumer electronic devices, including smartphones and tablets.

[0030] In the aforementioned execution process, setting "minimizing the average decoding overhead" as the core optimization objective and solving for the joint weight coefficients accordingly offers a fundamental advantage: it achieves a systematic upgrade of the design paradigm from local experience to global optimum. This objective directly integrates the system's ultimate performance requirements (reliability and efficiency), enabling the optimization of degree distribution to precisely align with the ultimate demands of engineering applications for the first time. Simultaneously, it reduces the complex distributed design problem to a clear low-dimensional parameter optimization problem. Through an automated solution framework combining Monte Carlo simulation and double-layer nested search, it ensures that the obtained weight coefficients are globally optimal or near-optimal solutions under the current model. This theoretically guarantees the scientific validity and optimality of the design results, while in practice, it significantly improves the adaptability and performance predictability of the solution.

[0031] Optionally, the initialization information includes: the number of packets transmitting test data, the parameters corresponding to the probability distribution of the sliding robust soliton distribution, the parameters of the corrected Poisson distribution, the number of Monte Carlo simulations, the search range of the second weight coefficient to be optimized, and the preset search step size.

[0032] In this embodiment of the invention, the initialization information can be set as shown in Table 1 below.

[0033] Table 1. Initialization Information Settings

[0034] To illustrate the overall process of solving the globally optimal weights for the joint degree distribution design and optimization method of fountain codes proposed in this invention, Figure 2 An exemplary flowchart illustrating the solution process for the globally optimal weights based on the method of this invention is shown. Figure 2 As shown, first, the process begins by initializing and setting parameters, including... The range of values, step size, and number of Monte Carlo simulations. Then, iterate through... The entire range. For each To determine the value, perform the following steps: Next, initialize the relevant parameters (such as...) (starting value), and enter the inner loop to iterate through Value. For each Value, Execution A second Monte Carlo simulation is performed to calculate the average decoding overhead and success rate. Then, it is determined whether the simulation results meet preset conditions (such as a success rate requirement). If they do, the current result is updated. Minimum cost and corresponding optimal Value. Then, take the next one. Repeat the above simulation and judgment process until all values ​​have been traversed. Value. After the inner loop completes, record the current value. The optimal And minimum overhead.

[0035] Next, take the next one. The value is repeated throughout the entire inner process (including...). (traversal, simulation, and update), until all are traversed. Value. Finally, from all records and the corresponding optimal In the process, compare and find the one corresponding to the global minimum cost. and The output is the optimal solution, and the process ends.

[0036] In addition, to illustrate the performance of the joint degree distribution design and optimization method for fountain codes provided by this invention, Figure 3 An exemplary probability distribution diagram of the joint degree distribution model provided by the present invention is shown. From Figure 3 As can be seen, the degree distribution designed in this invention exhibits peaks at degrees of 2, 10, and 26, covering small, medium, and large degree value ranges. Furthermore, this embodiment also compares and simulates various methods with the joint degree distribution method of this invention. Figure 4 An illustrative comparison of the decoding overhead performance of the method of the present invention with other comparative algorithms is shown. For example... Figure 4 As shown, the method of the present invention (MPMRID) and other comparative methods have lower decoding overhead at the same bit error rate. This further illustrates the effectiveness of the method of the present invention.

[0037] S104. Substitute the globally optimal weights into the joint degree distribution model to obtain the final fountain code joint degree distribution in the current scenario.

[0038] Optionally, after S104, it also includes: The sending end obtains the actual transmission data in the current scenario; divides the actual transmission data into multiple transmission data packets; uses the final fountain code joint degree distribution to determine the transmission degree value corresponding to multiple transmission data packets; and uses the transmission degree value to select the corresponding multiple transmission data packets for XOR operation to generate the transmission encoded packet. The receiving end decodes the data using the transmitted encoded packets to obtain the decoded information.

[0039] Optionally, the receiving end uses the transmitted encoded packets to decode the data, obtaining decoded information including: The receiving end receives the transmitted encoded packet and uses the BP algorithm to decode the data in the transmitted encoded packet to obtain the decoded information.

[0040] This invention provides a joint degree distribution design and optimization method for fountain codes. First, a model is jointly constructed using a modified Poisson distribution, an ideal soliton distribution, and a sliding robust soliton distribution. These three distributions have complementary advantages in low-degree smoothness, medium-degree diversity, and high-degree robustness, respectively. By dynamically adjusting the proportions of these three distributions through weight coefficients, the problem of excessive concentration of probability quality in low-degree values ​​is avoided, thereby enhancing the smoothness and diversity of the degree distribution and improving the compositional balance of the encoded symbols. Then, minimizing the average decoding overhead is used as the optimization objective, directly linking transmission efficiency and decoding reliability. Compared to simply pursuing decoding success rate, this objective simultaneously balances the decoding success probability with the required number of encoded symbols, enabling the degree distribution design to improve transmission efficiency while ensuring reliability, overcoming the problem of the original objective being disconnected from the ultimate performance indicator. Finally, Monte Carlo simulation is used to simulate actual transmission scenarios, and a two-layer nested search is combined to systematically optimize the weight coefficients, replacing the original local search method that relies on empirical adjustments. This mechanism allows for extensive exploration within the parameter space, significantly increasing the probability of finding the global optimum, thus providing a theoretically global optimization guarantee for the degree distribution performance. Based on this, the rationality of the degree distribution structure, the consistency between the optimization objective and system performance, and the global optimality of the design method are improved, thereby forming a systematic improvement in decoding reliability, transmission efficiency, and engineering applicability.

[0041] The method provided in this embodiment of the invention can be applied to electronic devices. Specifically, the electronic device can be a desktop computer, a portable computer, a smart mobile terminal, a server, etc., and this embodiment of the invention does not limit the application to such devices.

[0042] Based on the same inventive concept, embodiments of the present invention also provide a device for designing and optimizing the joint degree distribution of fountain codes. Figure 5 This is a schematic diagram of a joint degree distribution design and optimization device for fountain codes provided in an embodiment of the present invention, as shown below. Figure 5 As shown, it includes: an acquisition unit 501, a model building unit 502, an optimization processing unit 503, and an output unit 504; Acquisition unit 501 is used to acquire the test data to be transmitted in the current scenario; Model building unit 502 is used to form an optimized joint degree distribution model by using the test data to be transmitted and minimizing the average decoding overhead under the joint degree distribution model. The joint degree distribution model is jointly constructed by the probability distribution of the modified Poisson distribution, the probability distribution of the ideal soliton distribution, and the probability distribution of the sliding robust soliton distribution. The optimization processing unit 503 is used to perform global optimal solution search processing on the weight coefficients to be optimized in the optimized joint degree distribution model based on the Monte Carlo simulation method and the double-layer nested search method to obtain the global optimal weight. Output unit 504 is used to substitute the globally optimal weights into the joint degree distribution model to obtain the final fountain code joint degree distribution in the current scenario.

[0043] Figure 6 A schematic diagram of a fountain code federation degree distribution design and optimization device provided in an embodiment of the present invention includes: a processor 710, a storage medium 720, and a bus 730. The storage medium 720 stores machine-readable instructions executable by the processor 710. When the fountain code federation degree distribution design and optimization device is running, the processor 710 and the storage medium 720 communicate via the bus 730. The processor 710 executes the machine-readable instructions to perform the steps of the above-described method embodiment. Specific implementation methods and technical effects are similar and will not be described in detail here.

[0044] The storage medium may include random access memory (RAM) or non-volatile memory (NVM), such as at least one disk storage device. Optionally, the storage medium may also be at least one storage device located remotely from the aforementioned processor.

[0045] The processors mentioned above can be general-purpose processors, including central processing units (CPUs), network processors (NPs), etc.; they can also be digital signal processors (DSPs), application-specific integrated circuits (ASICs), field-programmable gate arrays (FPGAs), or other programmable logic devices, discrete gate or transistor logic devices, or discrete hardware components.

[0046] Although the invention has been described herein in conjunction with various embodiments, those skilled in the art, by reviewing the accompanying drawings and the disclosure, will understand and implement other variations of the disclosed embodiments in carrying out the claimed invention. In this description, the word "comprising" does not exclude other components or steps, "a" or "an" does not exclude a plurality, and "a plurality" means two or more, unless otherwise explicitly specified. Furthermore, while different embodiments may describe certain measures, this does not mean that these measures cannot be combined to produce good results.

[0047] The above description, in conjunction with specific preferred embodiments, provides a further detailed explanation of the present invention. It should not be construed that the specific implementation of the present invention is limited to these descriptions. For those skilled in the art, various simple deductions or substitutions can be made without departing from the inventive concept, and all such modifications and substitutions should be considered within the scope of protection of the present invention.

Claims

1. A method for designing and optimizing the joint degree distribution of fountain codes, characterized in that, include: Obtain the test data to be transmitted in the current scenario; Using the test data to be transmitted, an optimized joint degree distribution model is formed with minimizing the average decoding overhead as the optimization objective under the joint degree distribution model. The joint degree distribution model is jointly constructed by the probability distributions of the modified Poisson distribution, the ideal soliton distribution, and the sliding robust soliton distribution. Based on the optimized joint degree distribution model, the Monte Carlo simulation method and the double-layer nested search method are used to perform a global optimal solution search on the weight coefficients to be optimized in the optimized joint degree distribution model to obtain the global optimal weights. Substituting the globally optimal weights into the joint degree distribution model yields the final fountain code joint degree distribution for the current scenario.

2. The method for joint degree distribution design and optimization of fountain codes according to claim 1, characterized in that, The joint degree distribution model is expressed as follows: ； in, Indicates the fountain code encoding packet value The corresponding joint degree distribution model, This represents the first weight coefficient to be optimized. This represents the second weight coefficient to be optimized. Indicates the fountain code encoding packet value The corresponding probability distribution of the modified Poisson distribution, Indicates the fountain code encoding packet value The probability distribution corresponding to the ideal soliton distribution, Indicates the fountain code encoding packet value The probability distribution corresponding to the sliding robust soliton distribution.

3. The method for joint degree distribution design and optimization of fountain codes according to claim 2, characterized in that, The probability distribution of the modified Poisson distribution is expressed as: ； ； in, This indicates the corrected Poisson distribution parameters. Indicates the number of packets containing test data to be transmitted; The probability distribution of the ideal soliton distribution is expressed as: ； The probability distribution of the sliding robust soliton distribution is expressed as follows: ； ； ； ； in, The first parameter represents the probability distribution corresponding to the sliding robust soliton distribution. The second parameter represents the probability distribution corresponding to the sliding robust soliton distribution. The third parameter represents the probability distribution corresponding to the sliding robust soliton distribution. The fourth parameter represents the probability distribution corresponding to the sliding robust soliton distribution.

4. The method for joint degree distribution design and optimization of fountain codes according to claim 1, characterized in that, The optimized joint degree distribution model is used to perform a global optimal solution search on the weight coefficients to be optimized in the optimized joint degree distribution model using Monte Carlo simulation and a double-layer nested search method, to obtain the globally optimal weights, including: S201, The sending end obtains initialization information; S202. The sending end determines the current second weight coefficient to be optimized based on the initialization information, fixes the second weight coefficient to be optimized, and iterates through the first weight coefficient to be optimized to obtain multiple sets of weight coefficient pairs. S203. The transmitting end performs a preset number of Monte Carlo simulations on each of the multiple sets of weight coefficient pairs, and generates multiple intermediate data packets based on the optimized joint degree distribution model during the Monte Carlo simulation process, and sends the multiple intermediate data packets to the receiving end. S204. The receiving end receives the multiple intermediate data packets and performs decoding processing using the BP algorithm to obtain the average decoding overhead corresponding to all the weight coefficient pairs. S205. The weight coefficient pair corresponding to the minimum average decoding cost is taken as the local optimal weight coefficient; S206. Update the current second weight coefficient to be optimized using a preset search step size, and use the new second weight coefficient to be optimized as the current second weight coefficient to be optimized in S202. S207. Repeat S202-S206 to obtain all local optimal weight coefficients and corresponding local optimal average decoding costs within the search range; S207. The local optimal weight coefficient corresponding to the minimum local optimal average decoding cost is taken as the global optimal weight.

5. The method for joint degree distribution design and optimization of fountain codes according to claim 4, characterized in that, The initialization information includes: the number of packets transmitting the test data, the parameters corresponding to the probability distribution of the sliding robust soliton distribution, the parameters of the corrected Poisson distribution, the number of Monte Carlo simulations, the search range of the second weight coefficient to be optimized, and the preset search step size.

6. The method for joint degree distribution design and optimization of fountain codes according to claim 1, characterized in that, After substituting the optimal weights into the joint degree distribution model to obtain the final fountain code joint degree distribution in the current scenario, the process further includes: The sending end acquires the actual transmission data in the current scenario; divides the actual transmission data into multiple transmission data packets; uses the final fountain code joint degree distribution to determine the transmission degree value corresponding to the multiple transmission data packets; and uses the transmission degree value to select the corresponding multiple transmission data packets for XOR operation to generate a transmission encoded packet. The receiving end uses the transmitted encoded packet to decode the data and obtain the decoded information.

7. The method for joint degree distribution design and optimization of fountain codes according to claim 6, characterized in that, The receiving end uses the transmitted encoded packet to decode the data, obtaining the decoded information including: The receiving end receives the transmitted encoded packet and uses the BP algorithm to decode the transmitted encoded packet to obtain the decoded information.

8. A device for designing and optimizing the joint degree distribution of fountain codes, characterized in that, The joint degree distribution design and optimization device for the fountain code includes: an acquisition unit, a model building unit, an optimization processing unit, and an output unit; The acquisition unit is used to acquire the test data to be transmitted in the current scenario; The model building unit is used to form an optimized joint degree distribution model by using the test data to be transmitted and minimizing the average decoding overhead as the optimization objective under the joint degree distribution model. The joint degree distribution model is jointly constructed by the probability distribution of the modified Poisson distribution, the probability distribution of the ideal soliton distribution, and the probability distribution of the sliding robust soliton distribution. The optimization processing unit is used to perform a global optimal solution search on the weight coefficients to be optimized in the optimized joint degree distribution model based on the optimized joint degree distribution model, using Monte Carlo simulation method and double-layer nested search method, to obtain the global optimal weight. The output unit is used to substitute the globally optimal weights into the joint degree distribution model to obtain the final fountain code joint degree distribution in the current scenario.

9. A device for designing and optimizing the joint degree distribution of fountain codes, characterized in that, include: The device includes a processor, a storage medium, and a bus. The storage medium stores machine-readable instructions executable by the processor. When the fountain code federation distribution design and optimization device is running, the processor communicates with the storage medium via the bus, and the processor executes the machine-readable instructions to perform the steps of the fountain code federation distribution design and optimization method as described in any one of claims 1-7.

Images