Coal mine wireless communication system and method

Abstract

The invention belongs to the technical field of wireless communication networks, and discloses a coal mine wireless communication system and method. The method comprises the following steps: firstly, constructing a coal mine wireless communication model, wherein the coal mine wireless communication model comprises a plurality of servers and a plurality of sensors; then, based on a coal mine wireless communication model, constructing an end-to-end error probability minimization problem taking a sensor-server matching strategy, a code length and transmitting power as optimization variables; and finally, solving an end-to-end error probability minimization problem, determining a matching strategy, and obtaining a code length-transmitting power allocation optimization scheme. And performing coal mine wireless communication configuration based on the distribution optimization scheme. According to the invention, the end-to-end reliability in a coal mine wireless transmission scene can be improved.

Description

Technical Field

[0001] This invention belongs to the field of wireless communication network technology, and more specifically, relates to a wireless communication system and method for coal mines. Background Technology

[0002] In coal mine wireless communication scenarios, transmission reliability plays a crucial role in coal mine production, operation, and safety management, such as transmitting environmental information like temperature, humidity, and volume. However, existing coal mine wireless communication systems only focus on transmission rate, neglecting the reliability of the transmission process. Because sensor data collected in coal mines is typically small and latency requirements are high, the traditional assumption of error-free transmission under Shannon's theorem does not apply in this scenario, meaning data transmission is not always reliable. Therefore, improving the reliability of multi-sensor, multi-server coal mine wireless communication scenarios is of significant research importance for practical applications.

[0003] Thanks to the research of Polyanskiy et al. on the finite code length theorem, the decoding error probability in the wireless communication transmission process can be modeled well. However, Polyanskiy et al. only analyzed the decoding error probability in the point-to-point wireless communication transmission process, and research on end-to-end reliability is still lacking, especially research on end-to-end reliability improvement methods for transmission between multiple sensors and multiple servers. Summary of the Invention

[0004] The purpose of this invention is to improve the end-to-end reliability of wireless transmission in coal mine scenarios by providing a wireless communication system and method for coal mines.

[0005] This invention provides a wireless communication method for coal mines, comprising the following steps; A coal mine wireless communication model is constructed, which includes multiple servers and multiple sensors; Based on the aforementioned coal mine wireless communication model, an end-to-end error probability minimization problem is constructed with sensor-server matching strategy, code length, and transmission power as optimization variables. The problem of minimizing the end-to-end error probability is solved to determine the matching strategy and obtain an optimized allocation scheme for code length and transmit power; the coal mine wireless communication is configured based on the optimized allocation scheme.

[0006] Preferably, the coal mine wireless communication model includes N servers and K sensors located within a circular area of ​​radius R; the radius R ranges from 10m to 160m, the total number of servers ranges from 2 to 10, and the total number of sensors ranges from 2 to 100.

[0007] Preferably, the matching strategy is expressed as follows:

[0008] Each sensor can only be matched with one server, and the following relationships must be satisfied simultaneously:

[0009] In the formula, B represents the matching strategy matrix. This represents the matching relationship between the nth server and the kth sensor. , N represents the total number of servers, and K represents the total number of sensors; , When the value is 0, there is no match. When the value of is 1, it corresponds to a match; The code length is represented as follows:

[0010] In the formula, M represents the code length matrix; This indicates the code length used by the k-th sensor in the uplink from the k-th sensor to the n-th server.

[0011] Preferably, the end-to-end error probability of the entire system in a multi-sensor, multi-server coal mine wireless communication scenario is represented as follows:

[0012] In the formula, This represents the end-to-end error probability of the entire system; This represents the probability of a decoding error on the nth server when the sensor matches the nth server. in, It is expressed as follows:

[0013]

[0014] In the formula, This represents the probability of a decoding error on the nth server when the kth sensor matches the nth server. This represents the Q function in statistics. This represents the channel dispersion corresponding to the uplink from the k-th sensor to the n-th server. This represents the data packet size corresponding to the uplink from the k-th sensor to the n-th server. This represents the signal-to-noise ratio (SNR) of the uplink from the k-th sensor to the n-th server. This represents the Shannon capacity corresponding to the uplink from the k-th sensor transmitting its corresponding data packet to the n-th server.

[0015] Preferably, the problem of minimizing the end-to-end error probability is denoted as OP and expressed as follows:

[0016] In the formula, This represents the transmit power of the k-th sensor. Indicates the maximum transmit power limit. Indicates the maximum code length limit.

[0017] Preferably, solving the end-to-end error probability minimization problem includes: Determine the matching strategy under the current channel conditions; Based on the matching strategy under the current channel state, the problem of minimizing the end-to-end error probability is transformed into a decomposed optimization problem. The decomposed optimization problem is solved to obtain the allocation optimization scheme.

[0018] Preferably, the matching strategy under the current channel state is determined in the following way: Collect wireless channel information between the sensor and the server, as well as historical data corresponding to the server information; the wireless channel information includes bandwidth, modulation and coding scheme and channel gain, and the server information includes calculation frequency, server temperature and antenna elevation angle; Using the wireless channel information and the server information as parameters, and associating them with the existing matching results between sensors and servers during historical data transmission, a labeled historical matching dataset is constructed. Feature importance analysis is performed on the historical matching dataset, and different weights are assigned to each parameter based on the results of the feature importance analysis. In the LightGBM framework, average prior information based on historical data is introduced to train and obtain a nonlinear function model of each parameter and the matching strategy; the nonlinear function model is expressed as follows: In this context, X and Y are used as inputs, and B is used as the output. X represents wireless channel information, Y represents server information, and B represents the matching strategy. The real-time collected wireless channel information and server information are input into the nonlinear function model to obtain the matching strategy between each sensor and different servers under the current channel condition.

[0019] Preferably, the decomposed optimization problem is denoted as P1 and expressed as follows:

[0020] The second derivatives of the decoding error probability with respect to the code length and the transmit power are expressed as follows:

[0021] In the formula, L represents the number of sensors connected to the nth server; Indicates the first In the uplink from the nth sensor to the nth server, the nth The code length used by each sensor; Indicates the first The transmit power of each sensor; Indicates the first When a sensor is matched with the nth server, the probability of decoding error on the nth server; For P1, the allocation optimization scheme is obtained iteratively using the block coordinate descent method.

[0022] Preferably, the coal mine wireless communication method further includes: utilizing the closed-form relationship between optimal code length solutions, transforming P1 into a dimension-reduced decomposition optimization problem P2, where P2 is expressed as follows:

[0023] The closed-form relationship between the optimal code length solutions is expressed as follows:

[0024] In the formula, express The closed-form relationship between the corresponding optimal code length solutions; This indicates the code length used by the first sensor in the uplink from the first sensor to the nth server; a sensor with a clearly defined matching strategy is designated as the first sensor. Indicates the first The channel dispersion corresponding to the uplink from one sensor to the nth server. Indicates the first The size of the data packets corresponding to the uplink from one sensor to the nth server. Indicates the first The signal-to-noise ratio corresponding to the uplink of the data packets transmitted by each sensor to the nth server. Indicates the first The Shannon capacity corresponding to the uplink of each sensor transmitting the corresponding data packet to the nth server; This represents the channel dispersion corresponding to the uplink from the 1st sensor to the nth server. This represents the data packet size corresponding to the uplink from the 1st sensor to the nth server. This represents the signal-to-noise ratio (SNR) of the uplink from the first sensor transmitting its corresponding data packet to the nth server. This represents the Shannon capacity corresponding to the uplink from the first sensor transmitting the corresponding data packet to the nth server; For P2, the allocation optimization scheme is obtained iteratively using the block coordinate descent method.

[0025] On the other hand, the present invention provides a coal mine wireless communication system, comprising: A model building unit is used to build a coal mine wireless communication model, which includes multiple servers and multiple sensors. The minimization problem construction unit is used to construct an end-to-end error probability minimization problem based on the coal mine wireless communication model, with the sensor-server matching strategy, code length, and transmission power as optimization variables. The computing unit is used to solve the end-to-end error probability minimization problem, determine the matching strategy, and obtain the code length-transmit power allocation optimization scheme. The coal mine wireless communication system is used to perform the steps in the coal mine wireless communication method described above.

[0026] One or more technical solutions provided in this invention have at least the following technical effects or advantages: This invention first constructs a coal mine wireless communication model (including multiple servers and multiple sensors); then, based on the coal mine wireless communication model, it constructs an end-to-end error probability minimization problem with sensor-server matching strategy, code length, and transmission power as optimization variables; finally, it solves the end-to-end error probability minimization problem, determines the matching strategy, and obtains an optimized allocation scheme for code length and transmission power; and configures the coal mine wireless communication based on the optimized allocation scheme. In other words, this invention, targeting the characteristics of coal mine wireless communication scenarios and aiming to improve the end-to-end reliability of transmission between multiple sensors and multiple servers, proposes a novel and complete wireless communication scheme. This invention fills the research gap in improving the end-to-end reliability of transmission between multiple sensors and multiple servers in coal mine wireless communication scenarios, and can improve the end-to-end reliability of wireless transmission in coal mine scenarios. Attached Figure Description

[0027] Figure 1 This is a flowchart illustrating a wireless communication method for coal mines provided in Embodiment 1 of the present invention. Figure 2 This is a schematic diagram of a coal mine wireless communication model in a coal mine wireless communication method provided in Embodiment 1 of the present invention; Figure 3 This is a performance comparison diagram between a coal mine wireless communication method provided in Embodiment 1 of the present invention and a traditional method. Detailed Implementation

[0028] To better understand the above technical solutions, the following will provide a detailed explanation of the technical solutions in conjunction with the accompanying drawings and specific implementation methods.

[0029] Example 1: Example 1 provides a wireless communication method for coal mines, see [link to example]. Figure 1 This includes the following steps; S1. Construct a coal mine wireless communication model, which includes multiple servers and multiple sensors; S2. Based on the aforementioned coal mine wireless communication model, construct an end-to-end error probability minimization problem with sensor-server matching strategy, code length, and transmission power as optimization variables; S3. Solve the end-to-end error probability minimization problem, determine the matching strategy, and obtain the code length-transmit power allocation optimization scheme; configure the coal mine wireless communication based on the allocation optimization scheme.

[0030] The following is a detailed explanation of each step.

[0031] In step S1, considering the characteristics of wireless communication scenarios in coal mines, this invention constructs a wireless communication model for coal mines that includes N servers and K sensors. (See [link to relevant documentation]). Figure 2 .

[0032] Specifically, the coal mine wireless communication model includes N servers and K sensors located within a circular area of ​​radius R, where n and k represent the server and sensor indices, respectively, and satisfy the following conditions: and .

[0033] For example, the radius R ranges from 10m to 160m, the total number of servers ranges from 2 to 10, and the total number of sensors ranges from 2 to 100.

[0034] In step S2, based on the constructed coal mine wireless communication model, a sensor-server matching strategy B is introduced to determine the correspondence between each sensor and the server. The finite code length theory is used to characterize the decoding error probability of a single sensor transmission, deriving a closed-form expression for the decoding error probability on each server, and thus obtaining the end-to-end service error probability. Based on this, an end-to-end error probability minimization problem is constructed, with matching strategy B, code length, and transmission power as optimization variables.

[0035] Specifically, the matching relationship between each sensor and the server is represented by B, and the matching strategy is expressed as follows: (1) Each sensor can only be matched with one server, and the following relationships must be satisfied simultaneously: (2) In the formula, B represents the matching strategy matrix. This represents the matching relationship between the nth server and the kth sensor. , N represents the total number of servers, and K represents the total number of sensors; , When the value is 0, there is no match. When the value is 1, it corresponds to a match.

[0036] Since a sensor can only be paired with one server, The sum should be equal to the maximum number of users (i.e., the total number of sensors).

[0037] Accordingly, the code length is represented as follows: (3) In the formula, M represents the code length matrix; This indicates the code length used by the k-th sensor in the uplink from the k-th sensor to the n-th server.

[0038] The corresponding decoding error probability can be expressed as: (4) In the formula, This represents the probability of a decoding error on the nth server when the kth sensor matches the nth server. This represents the Q function in statistics. This represents the channel dispersion corresponding to the uplink from the k-th sensor to the n-th server. This represents the data packet size corresponding to the uplink from the k-th sensor to the n-th server. This represents the signal-to-noise ratio (SNR) of the uplink from the k-th sensor to the n-th server. This represents the Shannon capacity corresponding to the uplink from the k-th sensor transmitting its corresponding data packet to the n-th server.

[0039] For example, the parameter settings are as follows: In a coal mine wireless transmission scenario, the radius R = 150 m, the user's transmit power is 20 dBm, the bandwidth is 5 MHz, and the path loss issue is... The duration of each symbol is set to... The given maximum error probability and time limit are 0.01 and 25ms, respectively. Based on the above parameters, once the transmission distance d between the sensor and the server is determined, the path loss can be directly obtained, and the corresponding signal-to-noise ratio can be calculated according to the formula. The decoding error probability can be calculated using formula (4).

[0040] Specifically, when the k-th sensor is not connected to the n-th server, the decoding error probability can be considered infinite, and therefore is not considered here. Therefore, when the sensor is matched with the n-th server, the decoding error probability on the n-th server is expressed as: (5) In the formula, This represents the probability of a decoding error on the nth server when the sensor matches the nth server.

[0041] That is, Formula (5) is a closed-form expression of the decoding error probability on a single server. The transmission system corresponding to a single server is considered reliable if and only if the data transmitted by all sensors is successfully decoded on the server.

[0042] In summary, the end-to-end error probability of the entire system in a multi-sensor, multi-server coal mine communication scenario can be expressed as: (6) In the formula, This represents the end-to-end error probability of the entire system.

[0043] That is, Formula (6) is a closed-form expression of the end-to-end decoding error probability of the entire system. The system is considered to have high reliability if and only if no decoding error occurs during the transmission process.

[0044] Therefore, an end-to-end error probability minimization problem is constructed, with matching strategy B, code length, and transmit power as optimization variables. That is, this invention denotes the end-to-end error probability minimization problem as OP, and expresses it as follows: (7) In the formula, This represents the transmit power of the k-th sensor. Indicates the maximum transmit power limit. Indicates the maximum code length limit.

[0045] Step S3, solving the end-to-end error probability minimization problem includes: S31. Determine the matching strategy under the current channel condition; S32. Based on the matching strategy under the determined current channel state, the end-to-end error probability minimization problem is transformed into a decomposed optimization problem; S33. Solve the decomposed optimization problem to obtain the allocation optimization scheme.

[0046] Specifically, in step S31, the matching strategy under the current channel state is determined in the following way: Collect wireless channel information between the sensor and the server, as well as historical data corresponding to the server information; the wireless channel information includes bandwidth, modulation and coding scheme and channel gain, and the server information includes calculation frequency, server temperature and antenna elevation angle; Using the wireless channel information and the server information as parameters, and associating them with the existing matching results between sensors and servers during historical data transmission, a labeled historical matching dataset is constructed. Feature importance analysis is performed on the historical matching dataset, and different weights are assigned to each parameter based on the results of the feature importance analysis. In the LightGBM framework, average prior information based on historical data is introduced to train and obtain a nonlinear function model of each parameter and the matching strategy; the nonlinear function model is expressed as follows: In this context, X and Y are used as inputs, and B is used as the output. X represents wireless channel information, Y represents server information, and B represents the matching strategy. The real-time collected wireless channel information and server information are input into the nonlinear function model to obtain the matching strategy between each sensor and different servers under the current channel condition.

[0047] In other words, this invention considers the highly nonlinear relationship between the matching strategies of each sensor and different servers, which makes the aforementioned optimization problem unsolvable directly. Therefore, this invention combines historical data recorded by the system, cleans the data, and obtains reconstructed historical data containing the correlations between each sensor and different servers. Based on the reconstructed historical data, an importance analysis of the parameters is conducted, providing a theoretical basis for setting subsequent model training parameters. To further improve model performance, average prior information is introduced into LightGBM, and a nonlinear function model of each parameter and matching strategy is trained accordingly.

[0048] Specifically, the wireless channel information includes bandwidth, modulation and coding scheme, and channel gain, respectively represented by... , and The server information includes the calculation frequency, server temperature, and antenna elevation angle, respectively represented by... , and This indicates that a dataset D is constructed based on the acquired historical data, and the sliding window is always set to [value] during the training process. (The sliding window is a parameter during the training process. Since LightGBM adds average prior information during training, the length of this information and the amount of historical information it is based on are determined by the sliding window here.) The average value is added to LightGBM as prior information for training, resulting in a nonlinear function model of each parameter and the matching strategy. Based on the constructed historical data, the matching strategy between each sensor and different servers can be obtained in the following way: (8) Formula (8) means that, based on network parameters, the matching strategy under the current channel condition can be obtained through a trained model. Model F is obtained by training an improved LightGBM model based on historical data.

[0049] Based on the obtained matching strategy B, this invention decomposes the original optimization problem into several sub-problems, and the optimal solution to the original problem can be obtained by solving these sub-problems. To solve the sub-problems, this invention analyzes and obtains the closed-form relationship between code lengths, further reducing the dimensionality of the optimization problem, and proposes an efficient code length-power allocation scheme, effectively improving the end-to-end reliability of coal mine wireless communication services.

[0050] Specifically, the subproblems can be obtained through a defined matching strategy B. At this point, the matching strategies between each sensor and the server are determined one by one. Therefore, the optimization problem can be viewed as a single-server scenario. Let L be the number of sensors connected to the nth server, denoted as the set. Correspondingly, in steps S32 and S33, the decomposed optimization problem is denoted as P1 and expressed as follows: (9) In the formula, L represents the number of sensors connected to the nth server; Indicates the first In the uplink from the nth sensor to the nth server, the nth The code length used by each sensor; Indicates the first The transmit power of each sensor.

[0051] Formula (9) is a sub-problem obtained by decomposing the problem based on the obtained matching strategy. At this time, the optimization variables of the optimization problem are reduced to code length and transmit power.

[0052] At this point, the second derivatives of the decoding error probability with respect to the code length and the transmit power are expressed as follows: (10) In the formula, Indicates the first The probability of decoding error on the nth server when a sensor is matched with the nth server.

[0053] Formula (10) is used to determine the convex function relationship between the objective function of the decomposed subproblem and the code length and the transmit power by solving the sign of the second derivative of the objective function with respect to the code length and the transmit power.

[0054] For P1, the allocation optimization scheme is obtained iteratively using the block coordinate descent method.

[0055] That is, at this point, the end-to-end decoding error probability of the nth server holds a convex functional relationship with respect to the code length and transmit power, which can be obtained iteratively using the block coordinate descent method. However, directly using the block coordinate descent method cannot guarantee reliable performance well, and requires a large number of iterations.

[0056] Therefore, we further analyze the relationship between code lengths. Since the decoding error probability decreases monotonically with respect to the code length, when the code length used for one of the code lengths is determined, the closed-form relationship between the other code lengths with respect to that determined code length, i.e., the closed-form relationship between the optimal code length solutions, is expressed as: (11) In the formula, express The closed-form relationship between the corresponding optimal code length solutions; This indicates the code length used by the first sensor in the uplink from the first sensor to the nth server; a sensor with a clearly defined matching strategy is designated as the first sensor. Indicates the first The channel dispersion corresponding to the uplink from one sensor to the nth server. Indicates the first The size of the data packets corresponding to the uplink from one sensor to the nth server. Indicates the first The signal-to-noise ratio corresponding to the uplink of the data packets transmitted by each sensor to the nth server. Indicates the first The Shannon capacity corresponding to the uplink of each sensor transmitting the corresponding data packet to the nth server; This represents the channel dispersion corresponding to the uplink from the 1st sensor to the nth server. This represents the data packet size corresponding to the uplink from the 1st sensor to the nth server. This represents the signal-to-noise ratio (SNR) of the uplink from the first sensor transmitting its corresponding data packet to the nth server. This represents the Shannon capacity corresponding to the uplink from the first sensor transmitting the corresponding data packet to the nth server.

[0057] At this point, the subproblem (P1) can be further reduced in dimensionality and expressed as: (12) That is, in the preferred scheme, step S33 can be further optimized, that is, after completing step S32, it can also include: using the closed relationship between the optimal solutions of code length, P1 can be transformed into an optimization problem P2 after dimensionality reduction decomposition.

[0058] Finally, for P2, the allocation optimization scheme is obtained iteratively using the block coordinate descent method.

[0059] Formula (12) is a subproblem obtained by further reducing the dimensionality of Formula (11). At this time, the optimization variables are further reduced (to half of the original subproblem). Therefore, when using alternating iteration, the number of iterations required is greatly reduced, and the result is closer to the optimal performance.

[0060] That is, the functional relationship in formula (10) also holds for subproblem (P2), and the optimization variables are reduced to half that of subproblem (P1). At this time, due to the use of the closed relationship between the optimal solutions of code length, the number of iterations required is greatly reduced, and the result is closer to the optimal performance.

[0061] Figure 3 The performance comparison chart shows that the method proposed in this invention is superior to the traditional method (fixed matching strategy). The data shows that the end-to-end reliability of the method proposed in this invention is always better than that of the traditional method. This invention can improve the end-to-end reliability in the wireless transmission scenario of coal mines.

[0062] Example 2: Example 2 provides a wireless communication system for coal mines, comprising: A model building unit is used to build a coal mine wireless communication model, which includes multiple servers and multiple sensors. The minimization problem construction unit is used to construct an end-to-end error probability minimization problem based on the coal mine wireless communication model, with the sensor-server matching strategy, code length, and transmission power as optimization variables. The computing unit is used to solve the end-to-end error probability minimization problem, determine the matching strategy, and obtain the code length-transmit power allocation optimization scheme. The coal mine wireless communication system is used to perform the steps in the coal mine wireless communication method as described in Example 1.

[0063] Since the functions of each unit in the coal mine wireless communication system provided in Embodiment 2 correspond to the steps in the coal mine wireless communication method provided in Embodiment 1, Embodiment 2 can be understood by referring to the description of Embodiment 1, and will not be repeated here.

[0064] In summary, addressing the current lack of research on end-to-end reliability performance in coal mine wireless communication scenarios, this invention provides a coal mine wireless transmission system and method, aiming to improve end-to-end reliability in multi-sensor, multi-server coal mine wireless communication processes. Simulation results also demonstrate that this invention effectively improves the end-to-end reliability of coal mine wireless communication services.

[0065] Finally, it should be noted that the above specific embodiments are only used to illustrate the technical solutions of the present invention and not to limit it. Although the present invention has been described in detail with reference to examples, those skilled in the art should understand that modifications or equivalent substitutions can be made to the technical solutions of the present invention without departing from the spirit and scope of the technical solutions of the present invention, and all such modifications or substitutions should be covered within the scope of the claims of the present invention.

Claims

1. A wireless communication method for coal mines, characterized in that, Includes the following steps; A coal mine wireless communication model is constructed, which includes multiple servers and multiple sensors; Based on the aforementioned coal mine wireless communication model, an end-to-end error probability minimization problem is constructed with sensor-server matching strategy, code length, and transmission power as optimization variables. The end-to-end error probability minimization problem is solved to determine the matching strategy and obtain the code length-transmit power allocation optimization scheme; Configure wireless communication in coal mines based on the aforementioned allocation optimization scheme.

2. The coal mine wireless communication method according to claim 1, characterized in that, The coal mine wireless communication model includes N servers and K sensors located within a circular area of ​​radius R; the radius R ranges from 10m to 160m, the total number of servers ranges from 2 to 10, and the total number of sensors ranges from 2 to 100.

3. The coal mine wireless communication method according to claim 1, characterized in that, The matching strategy is expressed as follows: Each sensor can only be matched with one server, and the following relationships must be satisfied simultaneously: In the formula, B represents the matching strategy matrix. This represents the matching relationship between the nth server and the kth sensor. , N represents the total number of servers, and K represents the total number of sensors; , When the value is 0, there is no match. When the value of is 1, it corresponds to a match; The code length is represented as follows: In the formula, M represents the code length matrix; This indicates the code length used by the k-th sensor in the uplink from the k-th sensor to the n-th server.

4. The coal mine wireless communication method according to claim 3, characterized in that, The end-to-end error probability of the entire system in a multi-sensor, multi-server coal mine wireless communication scenario is represented as follows: In the formula, This represents the end-to-end error probability of the entire system; This represents the probability of a decoding error on the nth server when the sensor matches the nth server. in, It is expressed as follows: In the formula, This represents the probability of a decoding error on the nth server when the kth sensor matches the nth server. This represents the Q function in statistics. This represents the channel dispersion corresponding to the uplink from the k-th sensor to the n-th server. This represents the data packet size corresponding to the uplink from the k-th sensor to the n-th server. This represents the signal-to-noise ratio (SNR) of the uplink from the k-th sensor to the n-th server. This represents the Shannon capacity corresponding to the uplink from the k-th sensor transmitting its corresponding data packet to the n-th server.

5. The coal mine wireless communication method according to claim 4, characterized in that, The problem of minimizing the end-to-end error probability is denoted as OP and expressed as follows: In the formula, This represents the transmit power of the k-th sensor. Indicates the maximum transmit power limit. Indicates the maximum code length limit.

6. The coal mine wireless communication method according to claim 5, characterized in that, Solving the end-to-end error probability minimization problem includes: Determine the matching strategy under the current channel conditions; Based on the matching strategy under the current channel state, the problem of minimizing the end-to-end error probability is transformed into a decomposed optimization problem. The decomposed optimization problem is solved to obtain the allocation optimization scheme.

7. The coal mine wireless communication method according to claim 6, characterized in that, The matching strategy under the current channel condition is determined as follows: Collect wireless channel information between the sensor and the server, as well as historical data corresponding to the server information; the wireless channel information includes bandwidth, modulation and coding scheme and channel gain, and the server information includes calculation frequency, server temperature and antenna elevation angle; Using the wireless channel information and the server information as parameters, and associating them with the existing matching results between sensors and servers during historical data transmission, a labeled historical matching dataset is constructed. Feature importance analysis is performed on the historical matching dataset, and different weights are assigned to each parameter based on the results of the feature importance analysis. In the LightGBM framework, average prior information based on historical data is introduced to train and obtain a nonlinear function model of each parameter and the matching strategy; the nonlinear function model is expressed as follows: In this context, X and Y are used as inputs, and B is used as the output. X represents wireless channel information, Y represents server information, and B represents the matching strategy. The real-time collected wireless channel information and server information are input into the nonlinear function model to obtain the matching strategy between each sensor and different servers under the current channel condition.

8. The coal mine wireless communication method according to claim 6, characterized in that, Let the decomposed optimization problem be denoted as P1, and expressed as follows: The second derivatives of the decoding error probability with respect to the code length and the transmit power are expressed as follows: In the formula, L represents the number of sensors connected to the nth server; Indicates the first In the uplink from the nth sensor to the nth server, the nth The code length used by each sensor; Indicates the first The transmit power of each sensor; Indicates the first When a sensor is matched with the nth server, the probability of decoding error on the nth server; For P1, the allocation optimization scheme is obtained iteratively using the block coordinate descent method.

9. The coal mine wireless communication method according to claim 8, characterized in that, Also includes: Using the closed-form relationship between optimal code length solutions, P1 is transformed into a dimension-reduced optimization problem P2, which is represented as follows: The closed-form relationship between the optimal code length solutions is expressed as follows: In the formula, express The closed-form relationship between the corresponding optimal code length solutions; This indicates the code length used by the first sensor in the uplink from the first sensor to the nth server; a sensor with a clearly defined matching strategy is designated as the first sensor. Indicates the first The channel dispersion corresponding to the uplink from one sensor to the nth server. Indicates the first The size of the data packets corresponding to the uplink from one sensor to the nth server. Indicates the first The signal-to-noise ratio corresponding to the uplink of the data packets transmitted by each sensor to the nth server. Indicates the first The Shannon capacity corresponding to the uplink of each sensor transmitting the corresponding data packet to the nth server; This represents the channel dispersion corresponding to the uplink from the 1st sensor to the nth server. This represents the data packet size corresponding to the uplink from the 1st sensor to the nth server. This represents the signal-to-noise ratio (SNR) of the uplink from the first sensor transmitting its corresponding data packet to the nth server. This represents the Shannon capacity corresponding to the uplink from the first sensor transmitting the corresponding data packet to the nth server; For P2, the allocation optimization scheme is obtained iteratively using the block coordinate descent method.

10. A wireless communication system for coal mines, characterized in that, include: A model building unit is used to build a coal mine wireless communication model, which includes multiple servers and multiple sensors. The minimization problem construction unit is used to construct an end-to-end error probability minimization problem based on the coal mine wireless communication model, with the sensor-server matching strategy, code length, and transmission power as optimization variables. The computing unit is used to solve the end-to-end error probability minimization problem, determine the matching strategy, and obtain the code length-transmit power allocation optimization scheme. The coal mine wireless communication system is used to perform the steps in the coal mine wireless communication method as described in any one of claims 1-9.

Images