SNR-MCS dynamic mapping table design method

Abstract

The invention discloses an SNR-MCS dynamic mapping table design method, which belongs to the technical field of wireless communication, and comprises the following steps: continuously updating an SNR-MCS dynamic mapping table: outputting a transmission success probability based on a probability prediction model according to a signal-to-noise ratio and a modulation coding mode; screening an MCS candidate set according to the transmission success probability to construct an MCS prediction model; collecting an actual signal-to-noise ratio, inputting the actual signal-to-noise ratio into the MCS prediction model, outputting a modulation coding mode, and collecting a transmission feedback update data pool; dividing the signal-to-noise ratio into a plurality of segments, and generating an SNR-MCS dynamic mapping table based on the coverage range of the signal-to-noise ratio of each segment and the MCS prediction model; using the SNR-MCS dynamic mapping table for interaction and updating a data pool; and retraining the probability prediction model through the updated data pool. According to the invention, the problems of low transmission efficiency and reliability caused by incapability of adapting to continuously changing channel conditions in the prior art are solved.

Description

Technical Field

[0001] The invention relates to a method for designing an SNR-MCS dynamic mapping table, belonging to the technical field of wireless communications. Background Art

[0002] Adaptive Modulation and Coding (AMC), a typical link adaptation technology, has been widely used in wireless communications since its first proposal in 1968. Research has shown that dynamically adjusting the modulation and coding scheme can significantly improve system throughput in time-varying and multipath fading channels. With the continuous advancement of communications technology, AMC has been incorporated into Long Term Evolution (LTE) systems to adapt to channel characteristics and maximize the effective communication rate. In 5G communication systems, AMC has been further applied to dynamically adapt transmission parameters to changing wireless channel conditions, thereby improving bandwidth utilization and throughput performance. Currently, research on AMC technology focuses on AMC based on the classic Outer Loop Link Adaptation (OLLA) algorithm, table lookup methods based on SNR-MCS mapping, and AMC based on intelligent methods.

[0003] However, existing technologies still have many limitations. First, while the OLLA-based AMC method optimizes MCS selection by adjusting the SNR threshold, its performance is highly dependent on hyperparameter settings, and its convergence speed is limited by the feedback frequency of the Hybrid Automatic Repeat Request (HARQ) process, making it difficult to adapt to highly dynamic channel conditions. Second, while the table lookup method based on SNR-MCS mapping performs well in scenarios with relatively stable channel characteristics, the diversity and time-varying nature of channel conditions make it difficult to construct a universal mapping table. Furthermore, MCS selection is highly sensitive to SNR, limiting its applicability. Finally, while AMC techniques based on intelligent methods, such as Q-learning or deep learning, offer good adaptability, their high implementation complexity, long training time, and reliance on large amounts of training data limit their application in practical systems.

[0004] With the development of 5G and future 6G communication technologies, the communication environment will become more complex, and the time-variability and uncertainty of channels will further increase. Existing AMC technology still has shortcomings in dynamic adaptability, generalization capabilities, and computational efficiency, making it difficult to meet the high efficiency and reliability requirements of future communication systems. Summary of the Invention

[0005] The purpose of the present invention is to provide a method for designing a dynamic SNR-MCS mapping table. By utilizing the advantages of neural networks in processing nonlinear problems, the mapping relationship between SNR and MCS is designed to solve the problem of low transmission efficiency and reliability caused by the inability of existing technologies to adapt to changing channel conditions.

[0006] In order to solve the above technical problems, the present invention is implemented by adopting the following technical solutions:

[0007] The present invention provides a method for designing an SNR-MCS dynamic mapping table, comprising:

[0008] Acquire communication data between users and base stations and build a data pool, wherein the communication data includes signal-to-noise ratio, modulation and coding mode, and transmission feedback;

[0009] Use the data pool to train a probabilistic prediction model;

[0010] Repeat the following steps to continuously update the SNR-MCS dynamic mapping table:

[0011] According to the signal-to-noise ratio and modulation and coding mode, a real-time prediction is performed based on a trained probability prediction model to output the probability of successful transmission of the communication data;

[0012] Establish an MCS candidate set based on the modulation and coding scheme in the data pool;

[0013] Screening an MCS candidate set according to the transmission success probability of the communication data to build an MCS prediction model;

[0014] Collect the signal-to-noise ratio in the actual operating environment and input it into the MCS prediction model, and output the corresponding modulation and coding scheme;

[0015] Use the corresponding modulation and coding method to interact between users and base stations and collect transmission feedback;

[0016] Update the data pool according to the signal-to-noise ratio, the corresponding modulation and coding scheme, and the transmission feedback in the data pool;

[0017] The SNR in the data pool is divided into several segments according to its coverage and resolution. Based on the coverage of the SNR in each segment and the MCS prediction model, an SNR-MCS dynamic mapping table is generated.

[0018] The SNR-MCS dynamic mapping table is used for interaction between users and base stations, and the signal-to-noise ratio, modulation and coding scheme, and transmission feedback are collected to update the data pool;

[0019] The probability prediction model is retrained using the updated data pool.

[0020] Furthermore, the probability prediction model includes a binary classification neural network with two hidden layers and a Hard-Sigmoig function. After the binary classification neural network receives the signal-to-noise ratio and modulation coding mode for processing, it inputs the Hard-Sigmoig function for processing and outputs the probability of successful transmission.

[0021] Furthermore, the transmission success probability is expressed as:

[0022] ;

[0023] Where, Indicates the signal-to-noise ratio based on the interaction between the user and the base station and modulation and coding schemes , the transmission success probability is obtained, where Indicates that the transmission is successful. represents the Hard-Sigmoig function, Represents the signal-to-noise ratio of the binary classification neural network receiving the interaction between the user and the base station and modulation and coding schemes Output.

[0024] Furthermore, the loss function for training the probability prediction model is expressed as:

[0025] ;

[0026] Where, Represents the loss value of the binary classification neural network weight, Indicates the number of samples of training data in the data pool, Represents the first training data in the data pool samples, Indicates the Samples were successfully transferred. represents the logarithmic function, Indicates the first The signal-to-noise ratio of samples and modulation and coding schemes , and obtain the transmission success probability.

[0027] Furthermore, an MCS candidate set is established according to the modulation and coding scheme in the data pool, including:

[0028] Determine the modulation order setting range of the modulation coding scheme according to the minimum modulation order and the maximum modulation order of the MCS in the data pool;

[0029] Determine the code rate setting range of the modulation and coding mode based on the minimum code rate and maximum code rate of the MCS in the data pool;

[0030] Selecting multiple modulation order values within the modulation order setting range, selecting multiple code rate values within the code rate setting range, cross-combining the selected modulation order values with the selected code rate values to generate candidate MCS entries;

[0031] Traverse the candidate MCS entries:

[0032] If the spectrum efficiency of one candidate MCS entry is higher than the spectrum efficiency of another candidate MCS entry, and the modulation order is lower than the modulation order of the other candidate MCS entry, then the other candidate MCS entry is eliminated;

[0033] After the traversal is completed, the MCS candidate set is obtained.

[0034] Furthermore, the MCS candidate set includes several MCS entries, and the MCS entries include an index identifier, a modulation mode / order, a code rate, and a spectrum efficiency, wherein the index identifier and the spectrum efficiency have a monotonically increasing relationship.

[0035] Furthermore, screening an MCS candidate set according to the transmission success probability of the communication data and constructing an MCS prediction model includes:

[0036] Traversing the MCS candidate set;

[0037] For each MCS entry, calculate the product of spectrum efficiency and transmission success probability;

[0038] The MCS entry with the maximum product of spectrum efficiency and transmission success probability is selected as the MCS prediction model.

[0039] Furthermore, the MCS entry with the maximum product of spectrum efficiency and transmission success probability is expressed as:

[0040] ;

[0041] Where, The MCS entry that represents the maximum value of the product of spectrum efficiency and transmission success probability, where represents the signal-to-noise ratio, Indicates Get the MCS entry with the maximum value, Indicates the signal-to-noise ratio based on the interaction between the user and the base station and modulation and coding schemes , the transmission success probability is obtained, Modulation and coding scheme spectral efficiency.

[0042] Furthermore, the method for determining the coverage and resolution of the signal-to-noise ratio in the data pool includes:

[0043] By statistically analyzing the signal-to-noise ratio in the data pool, the mean of the signal-to-noise ratio is calculated and standard deviation ;

[0044] Set the lower limit of the signal-to-noise ratio coverage to , the upper limit is set to ;

[0045] Set the SNR resolution based on the SNR coverage.

[0046] Furthermore, the signal-to-noise ratio is divided into several segments, and based on the signal-to-noise ratio coverage of each segment and the MCS prediction model, an SNR-MCS dynamic mapping table is generated, including:

[0047] The signal-to-noise ratio is divided into several segments according to its coverage and resolution;

[0048] Input the lower bound of the signal-to-noise ratio coverage range of each segment into the MCS prediction model to obtain the corresponding MCS prediction value;

[0049] By constructing a mapping between the coverage range of the signal-to-noise ratio and the corresponding MCS prediction value, an SNR-MCS dynamic mapping table is generated.

[0050] Compared with the prior art, the present invention has the following beneficial effects:

[0051] 1. The present invention can capture changes in channel quality in real time and intelligently screen the optimal MCS scheme based on a trained probability prediction model, so that the target communication system can dynamically select the modulation and coding method with the highest matching degree with the current SNR while ensuring transmission reliability. Compared with the traditional static mapping table scheme, the present invention can improve the spectrum efficiency in scenarios with rapidly changing channel conditions. The present invention also ensures that the target communication system always operates at the optimal balance point between spectrum efficiency and transmission reliability through a continuously updated SNR-MCS dynamic mapping table. The present invention solves the problem of low transmission efficiency and reliability caused by the inability of existing technologies to adapt to constantly changing channel conditions.

[0052] 2. The present invention continuously updates the SNR-MCS mapping relationship through a closed-loop feedback mechanism, enabling the target communication system to automatically select the optimal modulation and coding scheme based on real-time channel conditions, maximizing spectrum utilization efficiency while ensuring transmission reliability. The present invention breaks through the efficiency bottleneck of traditional MCS selection schemes through a screening strategy that maximizes the product of spectrum efficiency and transmission success rate, achieving better spectrum resource allocation under the same bit error rate indicator. BRIEF DESCRIPTION OF THE DRAWINGS

[0053] Figure 11 is a flow chart of a method for designing an SNR-MCS dynamic mapping table provided by an embodiment of the present invention;

[0054] Figure 2 Schematic diagram of a data pool update mechanism provided by an embodiment of the present invention;

[0055] Figure 3 is a schematic diagram of a process for applying an MCS prediction model provided by an embodiment of the present invention;

[0056] Figure 4 1 is a schematic diagram of a simulation comparing the spectrum efficiency performance of the present invention and the outer loop link adaptation algorithm under different signal-to-noise ratios provided by an embodiment of the present invention;

[0057] Figure 5 This is a simulation diagram comparing the bit error rate performance of the present invention and the outer loop link adaptation algorithm under different signal-to-noise ratios provided by an embodiment of the present invention. DETAILED DESCRIPTION

[0058] The technical solution of the present invention is described in detail below through the accompanying drawings and specific embodiments. It should be understood that the embodiments of the present invention and the specific features in the embodiments are detailed descriptions of the technical solution of the present invention, rather than limitations on the technical solution of the present invention. In the absence of conflict, the embodiments of the present invention and the technical features in the embodiments can be combined with each other.

[0059] Example 1

[0060] like Figure 1 As shown, this embodiment introduces a method for designing an SNR-MCS dynamic mapping table, including:

[0061] Step 1: Acquire communication data between the user and the base station and build a data pool. The communication data includes signal-to-noise ratio, modulation and coding mode, and transmission feedback.

[0062] The present invention constructs a data pool through communication data of multiple traditional fixed modulation and coding methods. The continuous updating of the data pool enables the target communication system to perceive the signal-to-noise ratio and short-term dynamic changes.

[0063] Step 2: Use the data pool to train the probabilistic prediction model.

[0064] The present invention maps the signal-to-noise ratio and modulation and coding mode combination in the data pool into a measurable success probability, providing a quantitative basis for MCS screening. The trained probability prediction model can quickly predict the transmission performance based on any signal-to-noise ratio and modulation and coding mode, adapting to real-time decision-making needs.

[0065] Step 3: Repeat the following steps to continuously update the SNR-MCS dynamic mapping table:

[0066] According to the signal-to-noise ratio and modulation coding mode, real-time prediction is performed based on the trained probability prediction model to output the probability of successful transmission of the communication data.

[0067] The present invention inputs the current signal-to-noise ratio and modulation coding mode into a trained probability prediction model, calculates the prediction success probability of each MCS, and evaluates the reliability of different MCSs under current channel conditions in real time, thus avoiding the rigidity problem of a fixed mapping table.

[0068] An MCS candidate set is established based on the modulation and coding schemes in the data pool.

[0069] The present invention combines the physical layer constraints of the target communication system to generate an MCS candidate set that meets the conditions, thereby ensuring that the establishment of the MCS candidate set complies with the system hardware capabilities and protocol specifications.

[0070] An MCS candidate set is screened according to the transmission success probability of the communication data to construct an MCS prediction model.

[0071] The present invention traverses the MCS candidate set, calculates the product of the spectrum efficiency and the prediction success probability of each MCS, and selects the MCS with the largest product as the optimal solution. While ensuring the transmission success rate, it maximizes the spectrum utilization, breaking through the limitation of traditional solutions that are difficult to balance the two.

[0072] The signal-to-noise ratio in the actual operating environment is collected and input into the MCS prediction model, which then outputs the corresponding modulation and coding scheme.

[0073] The present invention monitors the signal-to-noise ratio of the current channel in real time, inputs the signal-to-noise ratio into a constructed MCS prediction model, outputs a recommended optimal MCS, and applies the results of the MCS prediction model to actual communication links, completing a closed loop from theoretical optimization to engineering practice.

[0074] Use the corresponding modulation and coding method to interact between users and base stations and collect transmission feedback.

[0075] The present invention verifies the validity of the MCS prediction model through actual transmission results, and provides real-time samples for subsequent data pool updating and probability prediction model retraining.

[0076] The data pool is updated according to the signal-to-noise ratio, the corresponding modulation and coding scheme, and the transmission feedback in the data pool.

[0077] The present invention appends the newly collected signal-to-noise ratio, the corresponding modulation and coding mode, and the transmission feedback to the data pool, which can continuously learn the dynamic changes of the channel and avoid the failure of the probability prediction model due to environmental drift.

[0078] The SNR is divided into several segments according to the coverage and resolution of the SNR in the data pool. Based on the coverage of the SNR of each segment and the MCS prediction model, an SNR-MCS dynamic mapping table is generated.

[0079] The present invention discretizes the continuous signal-to-noise ratio space into finite intervals, reduces the complexity of the mapping table, automatically matches the actual channel quality distribution through a segmentation strategy, and avoids performance loss caused by fixed segmentation.

[0080] The SNR-MCS dynamic mapping table is used for interaction between users and base stations, and the signal-to-noise ratio, modulation and coding mode, and transmission feedback are collected to update the data pool.

[0081] The present invention converts complex model prediction results into a mapping relationship that can be directly looked up in a table, reducing the complexity of real-time decision-making. At the same time, the SNR-MCS dynamic mapping table is dynamically adjusted as the data pool is updated, always reflecting the latest channel status.

[0082] The probability prediction model is retrained using the updated data pool.

[0083] The present invention uses the updated data pool to retrain the probability prediction model and adjust the neural network weights to adapt to channel changes, thereby avoiding the probability prediction model from becoming invalid due to long-term channel changes or upgrades of the target communication system. Incremental training enables the probability prediction model to gradually approach the actual channel characteristics, reducing prediction errors.

[0084] Example 2

[0085] Based on the same inventive concept as Example 1, this embodiment introduces the implementation steps of a method for designing an SNR-MCS dynamic mapping table, including:

[0086] Step 1: Acquire communication data between the user and the base station and build a data pool. The communication data includes signal-to-noise ratio, modulation and coding mode, and transmission feedback.

[0087] Step 2: Use the data pool to train the probabilistic prediction model.

[0088] In this embodiment, the probability prediction model includes a binary classification neural network with two hidden layers and a Hard-Sigmoig function. After the signal-to-noise ratio and modulation coding mode are received and processed by the binary classification neural network, the Hard-Sigmoig function is input for processing to output the probability of successful transmission.

[0089] In this embodiment, the transmission success probability is expressed as:

[0090] ;

[0091] Where, Indicates the signal-to-noise ratio based on the interaction between the user and the base station and modulation and coding schemes , the transmission success probability is obtained, where Indicates that the transmission is successful. represents the Hard-Sigmoig function, Represents the signal-to-noise ratio of the binary classification neural network receiving the interaction between the user and the base station and modulation and coding schemes Output.

[0092] In this embodiment, the loss function for training the probability prediction model is expressed as:

[0093] ;

[0094] Where, Represents the loss value of the binary classification neural network weight, Indicates the number of samples of training data in the data pool, Represents the first training data in the data pool samples, Indicates the Samples were successfully transferred. represents the logarithmic function, Indicates the first The signal-to-noise ratio of samples and modulation and coding schemes , and obtain the transmission success probability.

[0095] Step 3: Repeat the following steps to continuously update the SNR-MCS dynamic mapping table:

[0096] Step 3.1: Based on the signal-to-noise ratio and modulation coding mode, a real-time prediction is performed based on the trained probability prediction model, and the transmission success probability of the communication data is output.

[0097] Step 3.3: Create an MCS candidate set based on the modulation and coding schemes in the data pool;

[0098] In this embodiment, establishing an MCS candidate set according to the modulation and coding scheme in the data pool includes:

[0099] The modulation order setting range of the modulation coding scheme is determined according to the minimum modulation order and the maximum modulation order of the MCS in the data pool.

[0100] The code rate setting range of the modulation and coding mode is determined according to the minimum code rate and maximum code rate of the MCS in the data pool.

[0101] A plurality of modulation order values are selected within the modulation order setting range, a plurality of code rate values are selected within the code rate setting range, and the selected modulation order values and the selected code rate values are cross-combined to generate candidate MCS entries.

[0102] Traverse the candidate MCS entries:

[0103] If the spectrum efficiency of one candidate MCS entry is higher than that of another candidate MCS entry, and the modulation order of one candidate MCS entry is lower than that of another candidate MCS entry, the another candidate MCS entry is eliminated.

[0104] After the traversal is completed, the MCS candidate set is obtained.

[0105] In this embodiment, the MCS candidate set includes several MCS entries, and the MCS entry includes an index identifier, a modulation mode / order, a code rate, and a spectrum efficiency, wherein the index identifier and the spectrum efficiency have a monotonically increasing relationship.

[0106] Step 3.3: Filter the MCS candidate set according to the transmission success probability of the communication data to construct an MCS prediction model.

[0107] In this embodiment, the flow chart of applying the MCS prediction model is as follows: Figure 3 As shown, wherein, screening the MCS candidate set according to the transmission success probability of the communication data and constructing the MCS prediction model includes:

[0108] Step 3.3.1: Traverse the MCS candidate set:

[0109] For each MCS entry, the product of spectral efficiency and transmission success probability is calculated.

[0110] The MCS entry with the maximum product of spectrum efficiency and transmission success probability is selected as the MCS prediction model.

[0111] In this embodiment, the MCS entry with the maximum value of the product of spectrum efficiency and transmission success probability is expressed as:

[0112] ;

[0113] Where, The MCS entry that represents the maximum value of the product of spectrum efficiency and transmission success probability, where represents the signal-to-noise ratio, Indicates Get the MCS entry with the maximum value, Indicates the signal-to-noise ratio based on the interaction between the user and the base station and modulation and coding schemes , the transmission success probability is obtained, Modulation and coding scheme spectral efficiency.

[0114] Step 3.4: Collect the signal-to-noise ratio in the actual operating environment and input it into the MCS prediction model, and output the corresponding modulation and coding scheme.

[0115] Step 3.5: Use the corresponding modulation and coding method to interact with the user and the base station and collect transmission feedback.

[0116] Step 3.6: Update the data pool according to the signal-to-noise ratio in the data pool, the corresponding modulation and coding scheme, and the transmission feedback.

[0117] Step 3.7: Divide the SNR into several segments according to the SNR coverage and resolution in the data pool, and generate an SNR-MCS dynamic mapping table based on the SNR coverage and MCS prediction model of each segment.

[0118] In this embodiment, the method for determining the coverage and resolution of the signal-to-noise ratio in the data pool includes:

[0119] By statistically analyzing the signal-to-noise ratio in the data pool, the mean of the signal-to-noise ratio is calculated and standard deviation ;

[0120] Set the lower limit of the signal-to-noise ratio coverage to , the upper limit is set to ;

[0121] Set the SNR resolution based on the SNR coverage.

[0122] In this embodiment, the signal-to-noise ratio is divided into several segments, and based on the signal-to-noise ratio coverage of each segment and the MCS prediction model, an SNR-MCS dynamic mapping table is generated, including:

[0123] The signal-to-noise ratio is divided into several segments according to its coverage and resolution;

[0124] Input the lower bound of the signal-to-noise ratio coverage range of each segment into the MCS prediction model to obtain the corresponding MCS prediction value;

[0125] By constructing a mapping between the coverage range of the signal-to-noise ratio and the corresponding MCS prediction value, an SNR-MCS dynamic mapping table is generated.

[0126] Step 3.8: Use the SNR-MCS dynamic mapping table for interaction between the user and the base station, and collect the signal-to-noise ratio, modulation and coding mode, and transmission feedback to update the data pool.

[0127] Step 4: Repeat step 3 and retrain the probability prediction model through the updated data pool to achieve dynamic optimization and iterative update of the SNR-MCS mapping table. The data pool update mechanism is as follows: Figure 2 shown.

[0128] Figure 4 1 is a schematic diagram of a simulation comparing the spectrum efficiency performance of the present invention and the outer loop link adaptation algorithm under different signal-to-noise ratios provided by an embodiment of the present invention. Figure 5 : This is a schematic diagram of a simulation comparing the bit error rate performance of the present invention and the outer loop link adaptation algorithm under different signal-to-noise ratios provided by an embodiment of the present invention, wherein the parameters are set as follows:

[0129] In this embodiment, the target communication system has 32 antennas at both the transmitter and receiver ends, uses a Rayleigh channel model, and has 2 paths. LDPC coding is used as the channel coding method, and the Belief Propagation (BP) algorithm is used as the decoding method with a code length of 1024. The data pool size is 100.

[0130] In this embodiment, the learning rate of the probability prediction model is set to 0.005, and the retraining cycle is 20.

[0131] In this embodiment, the modulation order and mode for initializing the MCS candidate set include QPSK, 16QAM, and 64QAM; the upper limit of the code rate is 0.9, the lower limit is 0.1, the resolution is 0.05, and the number of MCS candidate set entries is set to 48.

[0132] In this embodiment, the SNR lower limit covered by the SNR-MCS dynamic mapping table is set to 0 dB, the upper limit is set to 30 dB, and the resolution is 2 dB.

[0133] Example 3

[0134] Based on the same inventive concept as other embodiments, this embodiment introduces a computer-readable storage medium on which computer instructions are stored. When the computer instructions are executed by a processor, the steps of the method of the above-mentioned embodiment 1 or 2 are implemented.

[0135] Example 4

[0136] Based on the same inventive concept as other embodiments, this embodiment introduces a computer program product, including computer instructions. When the computer instructions are executed by a processor, the steps of the method in the above-mentioned embodiment 1 or 2 are implemented.

[0137] In summary, the present invention can capture changes in channel quality in real time, and intelligently screen the optimal MCS scheme based on a trained probability prediction model, so that the target communication system can dynamically select the modulation and coding method with the highest matching degree with the current SNR while ensuring transmission reliability. Compared with the traditional static mapping table scheme, the present invention can improve the spectrum efficiency in scenarios where channel conditions change rapidly. The present invention also ensures that the target communication system always operates at the optimal balance point of spectrum efficiency and transmission reliability through a continuously updated SNR-MCS dynamic mapping table. The present invention solves the problem of low transmission efficiency and reliability caused by the inability of existing technologies to adapt to constantly changing channel conditions.

[0138] The present invention continuously updates the SNR-MCS mapping relationship through a closed-loop feedback mechanism, enabling the target communication system to automatically select the optimal modulation and coding scheme based on real-time channel conditions, maximizing spectrum utilization efficiency while ensuring transmission reliability. By adopting a screening strategy that maximizes the product of spectrum efficiency and transmission success rate, the present invention overcomes the efficiency bottleneck of traditional MCS selection schemes and achieves better spectrum resource allocation under the same bit error rate indicator.

[0139] Those skilled in the art will appreciate that embodiments of the present invention may be provided as methods, systems, or computer program products. Thus, the present invention may take the form of an entirely hardware embodiment, an entirely software embodiment, or an embodiment combining software and hardware aspects. Furthermore, the present invention may take the form of a computer program product implemented on one or more computer-usable storage media (including but not limited to magnetic disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0140] The present invention is described with reference to flowcharts and / or block diagrams of methods, devices (systems), and computer program products according to embodiments of the present invention. It should be understood that each process and / or block in the flowcharts and / or block diagrams, as well as combinations of processes and / or blocks in the flowcharts and / or block diagrams, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, a special-purpose computer, an embedded processor, or other programmable data processing device to produce a machine, so that the instructions executed by the processor of the computer or other programmable data processing device generate instructions for implementing the processes in the flowcharts and / or block diagrams. Figure 1 a process or multiple processes and / or boxes Figure 1 A device that provides the functions specified in a block or multiple blocks.

[0141] These computer program instructions may also be stored in a computer readable memory that can direct a computer or other programmable data processing device to work in a specific manner, so that the instructions stored in the computer readable memory produce an article of manufacture comprising an instruction device, which implements the process Figure 1a process or multiple processes and / or boxes Figure 1 The function specified in one or more boxes.

[0142] These computer program instructions can also be loaded onto a computer or other programmable data processing device so that a series of operational steps are executed on the computer or other programmable device to produce a computer-implemented process, thereby providing the instructions executed on the computer or other programmable device for implementing the process. Figure 1 a process or multiple processes and / or boxes Figure 1 A step that specifies a function in one or more boxes.

[0143] The embodiments of the present invention are described above in conjunction with the accompanying drawings, but the present invention is not limited to the above-mentioned specific implementation methods. The above-mentioned specific implementation methods are merely illustrative and not restrictive. Under the guidance of the present invention, ordinary technicians in this field can also make many forms without departing from the scope of protection of the purpose of the present invention and the claims, which are all protected by the present invention.

Claims

1. A method for designing a dynamic SNR-MCS mapping table, characterized in that: include: Acquire communication data between users and base stations and build a data pool, wherein the communication data includes signal-to-noise ratio, modulation and coding mode, and transmission feedback; Use the data pool to train a probabilistic prediction model; Repeat the following steps to continuously update the SNR-MCS dynamic mapping table: According to the signal-to-noise ratio and modulation and coding mode, a real-time prediction is performed based on a trained probability prediction model to output the probability of successful transmission of the communication data; Establish an MCS candidate set based on the modulation and coding scheme in the data pool; Screening an MCS candidate set according to the transmission success probability of the communication data to build an MCS prediction model; Collect the signal-to-noise ratio in the actual operating environment and input it into the MCS prediction model, and output the corresponding modulation and coding scheme; Use the corresponding modulation and coding method to interact between users and base stations and collect transmission feedback; Update the data pool according to the signal-to-noise ratio, the corresponding modulation and coding scheme, and the transmission feedback in the data pool; The SNR in the data pool is divided into several segments according to its coverage and resolution. Based on the coverage of the SNR in each segment and the MCS prediction model, an SNR-MCS dynamic mapping table is generated. The SNR-MCS dynamic mapping table is used for interaction between users and base stations, and the signal-to-noise ratio, modulation and coding scheme, and transmission feedback are collected to update the data pool; The probability prediction model is retrained using the updated data pool.

2. The SNR-MCS dynamic mapping table design method according to claim 1, characterized in that: The probability prediction model includes a binary classification neural network with two hidden layers and a Hard-Sigmoig function. After the binary classification neural network receives the signal-to-noise ratio and modulation coding mode for processing, it inputs the Hard-Sigmoig function for processing and outputs the probability of successful transmission.

3. The SNR-MCS dynamic mapping table design method according to claim 2, characterized in that: The transmission success probability is expressed as: ； Where, Indicates the signal-to-noise ratio based on the interaction between the user and the base station and modulation and coding schemes , the transmission success probability is obtained, where Indicates that the transmission is successful. represents the Hard-Sigmoig function, Represents the signal-to-noise ratio of the binary classification neural network receiving the interaction between the user and the base station and modulation and coding schemes Output.

4. The SNR-MCS dynamic mapping table design method according to claim 2, characterized in that: The loss function for training the probability prediction model is expressed as: ； Where, Represents the loss value of the binary classification neural network weight, Indicates the number of samples of training data in the data pool, Represents the first training data in the data pool samples, Indicates the Samples were successfully transferred. represents the logarithmic function, Indicates the first The signal-to-noise ratio of the samples and modulation and coding schemes , and obtain the transmission success probability.

5. The SNR-MCS dynamic mapping table design method according to claim 1, characterized in that: Establish an MCS candidate set based on the modulation and coding schemes in the data pool, including: Determine the modulation order setting range of the modulation coding scheme according to the minimum modulation order and the maximum modulation order of the MCS in the data pool; Determine the code rate setting range of the modulation and coding mode based on the minimum code rate and maximum code rate of the MCS in the data pool; Selecting multiple modulation order values within the modulation order setting range, selecting multiple code rate values within the code rate setting range, cross-combining the selected modulation order values with the selected code rate values to generate candidate MCS entries; Traverse the candidate MCS entries: If the spectrum efficiency of one candidate MCS entry is higher than the spectrum efficiency of another candidate MCS entry, and the modulation order is lower than the modulation order of the other candidate MCS entry, then the other candidate MCS entry is eliminated; After the traversal is completed, the MCS candidate set is obtained.

6. The SNR-MCS dynamic mapping table design method according to claim 5, characterized in that: The MCS candidate set includes several MCS entries, and the MCS entry includes an index identifier, a modulation mode / order, a code rate, and a spectrum efficiency, wherein the index identifier and the spectrum efficiency have a monotonically increasing relationship.

7. The SNR-MCS dynamic mapping table design method according to claim 6, characterized in that: Screening an MCS candidate set according to the transmission success probability of the communication data and constructing an MCS prediction model includes: Traversing the MCS candidate set; For each MCS entry, calculate the product of spectrum efficiency and transmission success probability; The MCS entry with the maximum product of spectrum efficiency and transmission success probability is selected as the MCS prediction model.

8. The SNR-MCS dynamic mapping table design method according to claim 7, characterized in that: The MCS entry with the maximum product of spectrum efficiency and transmission success probability is expressed as: ； Where, The MCS entry that represents the maximum value of the product of spectrum efficiency and transmission success probability, where represents the signal-to-noise ratio, Indicates Get the MCS entry with the maximum value, Indicates the signal-to-noise ratio based on the interaction between the user and the base station and modulation and coding schemes , the transmission success probability is obtained, Modulation and coding scheme spectral efficiency.

9. The SNR-MCS dynamic mapping table design method according to claim 1, characterized in that: The method for determining the coverage and resolution of the signal-to-noise ratio in the data pool includes: By statistically analyzing the signal-to-noise ratio in the data pool, the mean of the signal-to-noise ratio is calculated and standard deviation ; Set the lower limit of the signal-to-noise ratio coverage to , the upper limit is set to ; Set the SNR resolution based on the SNR coverage.

10. The SNR-MCS dynamic mapping table design method according to claim 9, characterized in that: The SNR is divided into several segments. Based on the SNR coverage and MCS prediction model of each segment, a dynamic SNR-MCS mapping table is generated, including: The signal-to-noise ratio is divided into several segments according to its coverage and resolution; Input the lower bound of the signal-to-noise ratio coverage range of each segment into the MCS prediction model to obtain the corresponding MCS prediction value; By constructing a mapping between the coverage range of the signal-to-noise ratio and the corresponding MCS prediction value, an SNR-MCS dynamic mapping table is generated.

Images