Data communication transmission method in wireless communication network

Abstract

The invention relates to the technical field of wireless communication, and discloses a data communication transmission method in a wireless communication network, which comprises the following steps of: 1, deploying cognitive radio equipment, receiving a channel signal in the wireless communication network, extracting received signal sample data, calculating an energy value of a current channel and judging an occupation state of the channel; inputting the channel state as historical channel state data into the channel state prediction model so as to predict the channel state at a future moment; and step 2, combining the channel state predicted in the step 1 with the communication demand of the user, and determining the matching relationship between the user and the channel. Real-time sensing of the channel state is achieved through the cognitive radio technology, the channel occupancy state at the future moment is predicted in combination with the long and short-term memory network, intelligent matching of dynamically-allocated channel resources and user requirements is achieved, and the effects of improving the spectrum resource utilization rate and avoiding resource waste are achieved.

Description

Technical Field

[0001] The present invention relates to the technical field of wireless communications, and in particular to a data communication transmission method in a wireless communication network. Background Art

[0002] The rapid evolution of wireless communication technology, especially the promotion and application of 5G and future 6G communication networks, has led to an increasing demand for high-performance wireless data communications worldwide. As the representative of current wireless communication technology, 5G networks have the characteristics of high data rate and low latency, and are used in the fields of Internet of Things, cloud computing, virtual reality, and autonomous driving. However, with the rapid increase in user devices, the diversification of communication service types, and the explosive growth of data traffic, the existing wireless communication networks have many bottlenecks in spectrum management, task calculation, data transmission, and network load regulation, and cannot meet the future communication needs of intelligence, high efficiency, and low energy consumption.

[0003] The spectrum resources in wireless communication networks are limited and are an important basis for supporting data communication transmission. However, traditional spectrum management methods mainly rely on static allocation mechanisms and cannot adapt to complex and dynamically changing network environments. In addition, the uncertainty of user data demand and the dynamic changes in the wireless communication environment make it an important technical challenge to achieve real-time perception and intelligent matching of channel resources and further improve the utilization of spectrum resources.

[0004] As edge computing technology becomes more and more popular, the task computing and data transmission of user devices gradually rely on edge nodes for processing. However, there are differences in the computing amount of tasks, the computing resources of edge nodes, and the transmission paths of data, which leads to a significant increase in energy consumption and latency during task offloading. At the same time, how to achieve joint optimization of computing energy consumption and transmission latency during task offloading, improve task execution efficiency, and reduce power consumption of network nodes is an important direction for current wireless communication network optimization.

[0005] In the process of data communication transmission, path determination is an important factor in determining communication efficiency. However, existing technologies mostly use static path selection or path scheduling mechanisms based on preset rules, which cannot effectively adapt to the dynamic changes in network link status. Therefore, how to dynamically split the data stream and achieve network load balancing and real-time optimization of the transmission path through multi-path parallel transmission is the key to solving the bottleneck of data transmission performance.

[0006] In communication networks with high traffic loads, the network link status and path quality have obvious real-time fluctuations, which can easily lead to network congestion and instability of transmission paths. Traditional path scheduling algorithms usually rely on static or local optimization mechanisms and cannot perform global prediction and real-time regulation in complex dynamic network environments, resulting in reduced data transmission efficiency and network performance.

[0007] Therefore, those skilled in the art provide a data communication transmission method in a wireless communication network to solve the above-mentioned problems. Summary of the invention

[0008] In view of the deficiencies of the prior art, the present invention provides a data communication transmission method in a wireless communication network to solve the problems raised in the above background technology.

[0009] To achieve the above objectives, the present invention is implemented through the following technical solutions: A data communication transmission method in a wireless communication network, comprising:

[0010] Step 1: Deploy cognitive radio equipment to receive channel signals in the wireless communication network, extract received signal sample data, calculate the energy value of the current channel, determine the occupancy state of the channel, and input the channel state as historical channel state data into the channel state prediction model to predict the channel state at a future time.

[0011] Step 2: Combining the channel state predicted in step 1 with the user's communication needs, determining the matching relationship between the user and the channel, dynamically allocating available channel resources in the wireless communication network, and forming a channel allocation plan;

[0012] Step 3: According to the channel allocation scheme formed in step 2, a network communication topology is constructed, and the wireless communication network is modeled as a network model including a node set and a link set. The optimal transmission path for data communication is determined by combining the delay weight and channel allocation status of each link, and the transmission path for user data is dynamically determined through a path scheduling module.

[0013] Step 4: Determine the offloading path of the user task carried on the transmission path after dynamic scheduling in step 3 according to the computing amount of the task and the computing resources of the edge computing node, and offload the task to the edge node for computing and processing. After the edge node completes the calculation, the calculation result is returned to the target user node through the network transmission path;

[0014] Step 5: According to the data and link status returned by the edge computing node on the transmission path in step 4, the link quality is judged, and under the premise of meeting the signal-to-noise ratio condition of the communication link, the transmission power of the communication node is controlled, and the data stream to be transmitted is divided into multiple data streams, and the data streams are transmitted in parallel along the optimal transmission path determined in step 3 through the multi-path transmission mechanism to complete the data communication transmission in the wireless communication network.

[0015] Preferably, when calculating the energy value of the current channel in step 1, energy detection is performed based on the received signal sample data, and the energy value is determined in the following manner:

[0016]

[0017] Wherein, E is the energy value of the current channel, N is the number of sampling points, and y(n) is the received signal sample data.

[0018] Preferably, when predicting the channel state at a future time in step 1, a long short-term memory network is used as a channel state prediction model, and the model is trained by historical channel state data to minimize the mean square error, and the loss function of the mean square error is:

[0019]

[0020] Among them, L is the loss value of the prediction model, T is the number of time steps, is the predicted channel state, S(t) is the actual channel state, and t represents the time step.

[0021] Preferably, when dynamically allocating available channel resources of the wireless communication network in step 2, the allocation relationship between users and channels is determined based on the communication needs of users and channel states and in accordance with the goal of maximizing the data transmission rate, and the relationship between the data transmission rate and the signal-to-noise ratio satisfies:

[0022] R i，j =B·log2(1+SNR i，j ),

[0023] Among them, R i，j is the transmission rate of user i on channel j, B is the channel bandwidth,

[0024] SNR i，j is the signal-to-noise ratio of user i on channel j.

[0025] Preferably, when determining the optimal transmission path for data communication in step 3, a weighted shortest path algorithm is used to determine the path based on the network communication topology and the delay weight of each link to minimize the total transmission delay, and the total transmission delay is expressed as:

[0026] T total =∑ (i，j)∈E w i，j x i，j ,

[0027] Among them, T total is the total transmission delay, w i，j is the delay weight of link (i, j),

[0028] x i，j is a link variable, a value of 1 indicates selected, and a value of 0 indicates unselected;

[0029] The dynamic scheduling of the transmission path in step 3 is adjusted in combination with the congestion status of the link, and the path quality is predicted and updated in real time through a deep reinforcement learning model.

[0030] Preferably, when the task is offloaded to the edge node in step 4, the optimal task offloading path is dynamically determined according to the task computing amount and the edge node computing resources, and the offloading time and the transmission delay are taken as the optimization target;

[0031] The execution time of the task offloading in step 4 is determined by the task computation amount L and the processing capacity f of the edge computing node. edge It is determined that the execution time satisfies the following relationship:

[0032]

[0033] Among them, T edge is the edge processing time of the task, L is the task computation amount, f edge The processing power of the edge node.

[0034] Preferably, when the task is offloaded to the edge node in step 4, in order to further reduce energy consumption and task execution time, an energy consumption optimization model is used to jointly optimize the computational amount and transmission power of the task, and the optimization goal is to minimize the total energy consumption E total , the total energy consumption includes computing energy consumption and transmission energy consumption, expressed as:

[0035]

[0036] Among them, E total is the total energy consumption of task offloading, P transmit is the transmission power during the task transmission process, P transmit is the data transmission time of the task, k is the energy consumption coefficient of the edge computing node, and f edge is the processing capacity of the edge node, T edge Processing time for the edge of the task.

[0037] Preferably, when the path quality is predicted and updated in real time by the deep reinforcement learning model in step 3, the path determination process is optimized by using the state-action value function Q, and the optimal decision of the path determination satisfies the following update relationship:

[0038]

[0039] Among them, Q(s, a) is the value function of taking action a in state s, R is the immediate reward after performing action a, γ is the discount factor, s ′ is the next state after the current state s executes action a, a ′ is the optimal action in the next state.

[0040] Preferably, when dynamically allocating available channel resources of the wireless communication network in step 2, a multi-objective optimization model is adopted to dynamically allocate channel resources by considering user service quality constraints and spectrum resource utilization, and the channel conflict probability is related to the number of users and the allocation scheme, which is expressed as:

[0041]

[0042] Among them, P conflict is the channel collision probability, p i，j is the probability of collision between user i and channel j, and M is the number of users in the wireless communication network.

[0043] Preferably, when the data stream is divided into multiple data streams for multi-path transmission in step 5, in order to improve the reliability and delay sensitivity of transmission, the allocation ratio α of each path is dynamically determined. i , satisfying the following constraints:

[0044] Among them, α i is the data flow allocation ratio of path i, N is the number of available transmission paths, L is the total amount of data to be transmitted, and R i is the data transmission rate on path i, T total is the total delay of data transmission.

[0045] The present invention provides a data communication transmission method in a wireless communication network. It has the following beneficial effects:

[0046] 1. The present invention realizes real-time perception of channel status through cognitive radio technology, and predicts the channel occupancy status at future moments in combination with long short-term memory network, realizes intelligent matching of dynamic allocation of channel resources and user needs, and achieves the effect of improving spectrum resource utilization and avoiding resource waste.

[0047] 2. The present invention establishes a joint optimization model of energy consumption and delay by comprehensively considering the task computing amount, edge node computing resources and transmission path conditions, thereby minimizing the computing energy consumption and transmission energy consumption of the task, thereby reducing the energy consumption of network nodes and the task execution delay.

[0048] 3. The present invention divides the data stream into multiple data streams and uses a multi-path transmission mechanism for parallel transmission, thereby achieving dynamic path allocation and real-time update, thereby reducing data transmission delay, balancing network load and improving transmission reliability.

[0049] 4. The present invention introduces a deep reinforcement learning model to perform real-time prediction and dynamic update of network link status and path quality, thereby realizing intelligent regulation of path selection and dynamic adjustment of congestion status, thereby alleviating network congestion and improving data transmission continuity and network performance. BRIEF DESCRIPTION OF THE DRAWINGS

[0050] Figure 1 It is a flow chart of the present invention. DETAILED DESCRIPTION

[0051] In order to make the technical personnel in the technical field understand the scheme of the present invention, the technical scheme in the embodiment of the present invention will be clearly and completely described below in combination with the drawings in the embodiment of the present invention. Obviously, the described embodiment is a partial embodiment of the present invention, not a complete embodiment. Based on the embodiment of the present invention, other embodiments obtained by ordinary technicians in the field without creative work should fall within the scope of protection of the present invention.

[0052] The present invention is described in detail below in conjunction with the accompanying drawings:

[0053] Example:

[0054] Please see attached Figure 1 , an embodiment of the present invention provides a data communication transmission method in a wireless communication network, comprising:

[0055] Step 1: Deploy cognitive radio equipment to receive channel signals in the wireless communication network, extract received signal sample data, calculate the energy value of the current channel, determine the occupancy state of the channel, and input the channel state as historical channel state data into the channel state prediction model to predict the channel state at a future time.

[0056] Step 2: Combining the channel state predicted in step 1 with the user's communication needs, determining the matching relationship between the user and the channel, dynamically allocating available channel resources in the wireless communication network, and forming a channel allocation plan;

[0057] Step 3: According to the channel allocation scheme formed in step 2, a network communication topology is constructed, and the wireless communication network is modeled as a network model including a node set and a link set. The optimal transmission path for data communication is determined by combining the delay weight and channel allocation status of each link, and the transmission path for user data is dynamically determined through a path scheduling module.

[0058] Step 4: Determine the offloading path of the user task carried on the transmission path after dynamic scheduling in step 3 according to the computing amount of the task and the computing resources of the edge computing node, and offload the task to the edge node for computing and processing. After the edge node completes the calculation, the calculation result is returned to the target user node through the network transmission path;

[0059] Step 5: According to the data and link status returned by the edge computing node on the transmission path in step 4, the link quality is judged, and under the premise of meeting the signal-to-noise ratio condition of the communication link, the transmission power of the communication node is controlled, and the data stream to be transmitted is divided into multiple data streams, and the data streams are transmitted in parallel along the optimal transmission path determined in step 3 through the multi-path transmission mechanism to complete the data communication transmission in the wireless communication network.

[0060] Benefits of step 1: Real-time perception of channel status is achieved through cognitive radio technology, and the current channel occupancy is accurately determined to ensure the availability of data transmission. A channel status prediction model is introduced to effectively predict the future channel occupancy status, plan channel resources in advance, avoid resource waste caused by static channel allocation, and provide data support for subsequent dynamic allocation of channel resources.

[0061] Step 2 Benefits: Combine user needs with channel status to achieve dynamic allocation of channel resources and improve the utilization of spectrum resources. The dynamic nature of the channel allocation scheme enables the network to adapt to real-time communication needs and meet the transmission requirements of different users in different time periods. It provides an efficient channel allocation mechanism, improves network transmission rate, and alleviates resource conflicts during communication.

[0062] Step 3 Benefits: By building a network communication topology, the communication relationship between network nodes and links is fully described, providing a basis for path scheduling. Considering the delay weight and channel allocation status of the link, the optimal transmission path for data communication is determined, effectively reducing the transmission delay. The dynamic path scheduling module is used to adjust the path in real time according to the link status, reduce network congestion, and improve the stability and reliability of data transmission.

[0063] Benefits of step 4: By deploying edge computing nodes, the task computing load is dynamically offloaded to the appropriate edge nodes, reducing the computing burden of the terminal devices. Combined with the computing resources of the edge nodes, the optimal task offloading path is dynamically determined, effectively improving the task processing efficiency, shortening the task computing delay, and reducing the response time of data return, meeting the application scenarios with high real-time requirements;

[0064] Benefits of step 5: By judging the link quality and dynamically controlling the transmission power of the communication nodes, the network energy consumption can be effectively reduced to meet the needs of green communication. The data stream to be transmitted can be divided into multiple data streams, which can be transmitted in parallel through a multi-path transmission mechanism to further improve the data transmission rate and network bandwidth utilization. In multi-path transmission, dynamic path allocation and real-time update enhance the reliability and stability of data transmission and reduce network congestion and data loss.

[0065] When calculating the energy value of the current channel in step 1, energy detection is performed based on the received signal sample data, and the energy value is determined in the following way:

[0066]

[0067] Wherein, E is the energy value of the current channel, N is the number of sampling points, and y(n) is the received signal sample data.

[0068] When predicting the channel state at the future time in step 1, the long short-term memory network is used as the channel state prediction model. The model is trained with historical channel state data to minimize the mean square error. The loss function of the mean square error is:

[0069]

[0070] Among them, L is the loss value of the prediction model, T is the number of time steps, is the predicted channel state, S(t) is the actual channel state, and t represents the time step.

[0071] In step 1, energy detection is performed through the received signal sample data, which can determine the occupancy status of the current channel in real time and determine whether the channel is idle. The calculation of the specific energy value is based on the cumulative energy of the received signal sample data, which effectively avoids the complex channel estimation process and has the advantages of simple calculation and strong real-time performance. In addition, the energy detection method is applicable to most channel environments and can achieve rapid detection under low signal-to-noise ratio conditions, ensuring the accuracy and reliability of channel status perception.

[0072] By using the long short-term memory network as the channel state prediction model, we can make full use of historical channel state data, capture the long-term dependency of channel states, and predict the channel state at future moments. The long short-term memory network model trains and optimizes the channel state by minimizing the mean square error loss function, thereby improving the accuracy of the prediction results. The prediction results provide accurate data support for subsequent channel resource allocation, realize advance planning and dynamic scheduling of channel resources, and further improve the communication efficiency of the network.

[0073] When dynamically allocating available channel resources of the wireless communication network in step 2, the allocation relationship between users and channels is determined based on the communication needs of users and channel status and in accordance with the goal of maximizing the data transmission rate. The relationship between the data transmission rate and the signal-to-noise ratio satisfies:

[0074] R i，j =B·log2(1+SNR i，j ),

[0075] Among them, R i，j is the transmission rate of user i on channel j, B is the channel bandwidth,

[0076] SNR i，j is the signal-to-noise ratio of user i on channel j.

[0077] When dynamically allocating available channel resources in a wireless communication network, the present invention comprehensively considers the user's communication needs and the current channel status to ensure that channel resources can be reasonably allocated to users and meet the communication needs of different users. Compared with the traditional static channel allocation method, the dynamic allocation mechanism can make adjustments according to changes in the real-time network environment, effectively avoiding resource waste and further improving the utilization rate of spectrum resources.

[0078] This step determines the allocation relationship between users and channels by associating the user data transmission rate with the channel signal-to-noise ratio, with the goal of maximizing the data transmission rate. Allocating more user resources on channels with higher signal-to-noise ratios fully utilizes the transmission capacity of high-quality channels, increases the user's data transmission rate, and improves communication efficiency.

[0079] When determining the optimal transmission path for data communication in step 3, a weighted shortest path algorithm is used to determine the path based on the network communication topology and the delay weight of each link to minimize the total transmission delay. The total transmission delay is expressed as:

[0080] T total =∑ (i，j)∈E w i，j x i，j ,

[0081] Among them, T total is the total transmission delay, w i，j is the delay weight of link (i, j),

[0082] x i，j is a link variable, a value of 1 indicates selected, and a value of 0 indicates unselected;

[0083] In step 3, the dynamic scheduling of the transmission path is adjusted based on the congestion status of the link, and the path quality is predicted and updated in real time through a deep reinforcement learning model.

[0084] The present invention uses the weighted shortest path algorithm to determine the optimal data transmission path through the network communication topology and the delay weight of each link, with the goal of minimizing the total transmission delay. In the process of optimizing the link weight, the real-time status of the link is comprehensively considered, effectively reducing the transmission time of data in the network, improving communication efficiency, and meeting application scenarios with high requirements for low delay.

[0085] During data transmission, the state of the network link fluctuates dynamically, which can easily lead to path congestion. The present invention dynamically schedules the path based on the congestion state of the link, adjusts the data transmission path in real time, avoids the use of congested links, improves the transmission stability of network data, and ensures the balanced distribution of network load.

[0086] The present invention introduces a deep reinforcement learning model in the dynamic scheduling of transmission paths, and realizes intelligent path selection by real-time monitoring of network link status and dynamic prediction of path quality. Through the self-learning and real-time update mechanism of reinforcement learning, the optimal path can be quickly found in a complex network environment, adapting to the dynamic changes of link status, and effectively improving the continuity of data transmission and network performance.

[0087] Through the intelligent decision-making and dynamic optimization of the path scheduling module, the utilization rate of network resources is further improved. At the same time, the dynamic determination and real-time update of the optimal path ensure the transmission stability of data communication, reduce the impact of transmission path failure or performance degradation, and enhance the robustness of the network.

[0088] When the task is offloaded to the edge node in step 4, the optimal task offloading path is dynamically determined according to the task computing amount and the edge node computing resources, and the offloading time and transmission delay are taken as the optimization target;

[0089] The execution time of the task offloading in step 4 is determined by the task computation amount L and the processing capacity of the edge computing node f edge Determine that the execution time satisfies the following relationship:

[0090]

[0091] Among them, T edge is the edge processing time of the task, L is the task computation amount, f edge The processing power of the edge node.

[0092] When the task is offloaded to the edge node in step 4, in order to further reduce energy consumption and task execution time, the energy consumption optimization model is used to jointly optimize the task's computational workload and transmission power. The optimization goal is to minimize the total energy consumption E total , the total energy consumption includes computing energy consumption and transmission energy consumption, which can be expressed as:

[0093]

[0094] Among them, E total is the total energy consumption of task offloading, P transmit is the transmission power during the task transmission process, P transmit is the data transmission time of the task, k is the energy consumption coefficient of the edge computing node, and f edge is the processing capacity of the edge node, T edgeProcessing time for the edge of the task.

[0095] The present invention dynamically determines the optimal task offloading path by comprehensively considering the task computation amount and the computing resources of the edge node, ensuring that the task is executed on the edge node with stronger computing power. The task execution time is dynamically matched by the task computation amount and the edge node processing power, and the computing resources are reasonably allocated to minimize the execution delay of the task, improve the computing processing efficiency, and meet the demand for low-latency tasks.

[0096] In the process of task offloading, the present invention introduces an energy consumption optimization model, comprehensively considers the task computing amount and transmission power, and jointly minimizes the computing energy consumption and transmission energy consumption. By dynamically adjusting the transmission power and task execution path, balancing computing resources and transmission energy consumption, the total energy consumption of network nodes in the process of task offloading is further reduced, and the energy-saving performance of the system is enhanced.

[0097] During the task offloading process, the present invention optimizes both the task transmission time and the calculation execution time, ensuring that the task completes the calculation and feedback in the shortest time. In the transmission stage, the transmission power during the task transmission process is controlled to balance the transmission energy consumption and task delay; in the calculation stage, the task calculation is completed quickly through the efficient processing capability of the edge node, thereby achieving rapid response and execution of the task.

[0098] By dynamically offloading tasks to edge nodes for processing, the computing burden of terminal devices can be effectively reduced, the energy consumption and temperature of the devices can be reduced, and the service life of the terminal devices can be extended. At the same time, the terminal devices can use more resources for other key functions, improving the efficiency and performance of the overall system.

[0099] The present invention can flexibly adapt to the dynamic changes of network environment and node computing resources by dynamically determining the task offloading path and combining the real-time joint optimization of energy consumption and delay. In the case of high load or network status fluctuations, it can effectively determine the optimal offloading path to ensure the efficient execution of tasks.

[0100] In step 3, when the path quality is predicted and updated in real time through the deep reinforcement learning model, the state-action value function Q is used to optimize the path determination process. The optimal decision for path determination satisfies the following update relationship:

[0101]

[0102] Among them, Q(s, a) is the value function of taking action a in state s, R is the immediate reward after performing action a, γ is the discount factor, s ′ is the next state after the current state s executes action a, a ′ is the optimal action in the next state.

[0103] By introducing the deep reinforcement learning model in step 3 and using the state-action value function to perform real-time prediction and optimization of path quality, the present invention achieves the following key advantages:

[0104] The reinforcement learning model is used to monitor link status in real time, dynamically predict path quality, and ensure the optimality and real-time performance of path determination.

[0105] When the network status changes dynamically, the path decision can be adjusted quickly to avoid congested links and improve the accuracy of path selection.

[0106] Reinforcement learning continuously optimizes path decisions through self-learning mechanisms, improving network transmission efficiency and system performance.

[0107] Optimize the path based on the state-action value function to reduce data transmission delay and meet the needs of low-latency applications.

[0108] It has strong robustness and anti-interference capabilities in complex network environments, ensuring the continuity and reliability of data transmission.

[0109] In summary, step 3 realizes intelligent prediction and real-time optimization of path decisions through the introduction of deep reinforcement learning models, effectively reduces data transmission delays, alleviates network congestion, and improves the stability and reliability of data communications, providing intelligent and efficient path scheduling solutions for data transmission in wireless communication networks.

[0110] When dynamically allocating available channel resources of the wireless communication network in step 2, a multi-objective optimization model is used to dynamically allocate channel resources by considering user service quality constraints and spectrum resource utilization. The channel conflict probability is related to the number of users and the allocation scheme, which is expressed as:

[0111]

[0112] Among them, P conflict is the channel collision probability, p i，j is the probability of collision between user i and channel j, and M is the number of users in the wireless communication network.

[0113] When the data stream is divided into multiple data streams for multi-path transmission in step 5, in order to improve the reliability and delay sensitivity of transmission, the allocation ratio α of each path is dynamically determined i , satisfying the following constraints:

[0114]

[0115] Among them, α i is the data flow allocation ratio of path i, N is the number of available transmission paths, L is the total amount of data to be transmitted, and R i is the data transmission rate on path i, Ttotal is the total delay of data transmission.

[0116] Through the dynamic channel resource allocation in step 2 and the dynamic allocation of multi-path data streams in step 5, the present invention achieves the following key advantages:

[0117] Through the multi-objective optimization model, channel resources are dynamically allocated to reduce the probability of channel conflicts and improve the utilization efficiency of spectrum resources.

[0118] Split the data stream and transmit it in parallel through multiple paths to reduce the impact of a single link failure and improve the stability of data transmission.

[0119] By dynamically allocating the traffic ratio of each path, high-quality links are determined, transmission delays are effectively reduced, and the needs of delay-sensitive applications are met.

[0120] Through the dynamic distribution of traffic through the multi-path mechanism, the network load is balanced, the network bandwidth utilization is improved, and the continuity and efficiency of data transmission are guaranteed.

[0121] In summary, steps 2 and 5 solve the problems of low spectrum resource utilization, poor data transmission reliability, high transmission delay and unbalanced network load by combining resource optimization with dynamic path allocation, and provide an efficient, stable and low-latency wireless communication network data transmission solution.

[0122] Although embodiments of the present invention have been shown and described, it will be appreciated by those skilled in the art that various changes, modifications, substitutions and variations may be made to the embodiments without departing from the principles and spirit of the present invention, and that the scope of the present invention is defined by the appended claims and their equivalents.

Claims

1. A data communication transmission method in a wireless communication network, characterized in that: include: Step 1: Deploy cognitive radio equipment to receive channel signals in the wireless communication network, extract received signal sample data, calculate the energy value of the current channel, determine the occupancy state of the channel, and input the channel state as historical channel state data into the channel state prediction model to predict the channel state at a future time. Step 2: Combining the channel state predicted in step 1 with the user's communication needs, determining the matching relationship between the user and the channel, dynamically allocating available channel resources in the wireless communication network, and forming a channel allocation plan; Step 3: According to the channel allocation scheme formed in step 2, a network communication topology is constructed, and the wireless communication network is modeled as a network model including a node set and a link set. The optimal transmission path for data communication is determined by combining the delay weight and channel allocation status of each link, and the transmission path for user data is dynamically determined through a path scheduling module. Step 4: Determine the offloading path of the user task carried on the transmission path after dynamic scheduling in step 3 according to the computing amount of the task and the computing resources of the edge computing node, and offload the task to the edge node for computing and processing. After the edge node completes the calculation, the calculation result is returned to the target user node through the network transmission path; Step 5: According to the data and link status returned by the edge computing node on the transmission path in step 4, the link quality is judged, and under the premise of meeting the signal-to-noise ratio condition of the communication link, the transmission power of the communication node is controlled, and the data stream to be transmitted is divided into multiple data streams, and the data streams are transmitted in parallel along the optimal transmission path determined in step 3 through the multi-path transmission mechanism to complete the data communication transmission in the wireless communication network.

2. A method for data communication transmission in a wireless communication network according to claim 1, characterized in that: When calculating the energy value of the current channel in step 1, energy detection is performed based on the received signal sample data, and the energy value is determined in the following manner: Wherein, E is the energy value of the current channel, N is the number of sampling points, and y(n) is the received signal sample data.

3. The method for data communication transmission in a wireless communication network according to claim 1, characterized in that: When predicting the channel state at a future time in step 1, a long short-term memory network is used as a channel state prediction model. The model is trained by historical channel state data to minimize the mean square error. The loss function of the mean square error is: Among them, L is the loss value of the prediction model, T is the number of time steps, is the predicted channel state, S(t) is the actual channel state, and t represents the time step.

4. The method for data communication transmission in a wireless communication network according to claim 1, characterized in that: When dynamically allocating available channel resources of the wireless communication network in step 2, the allocation relationship between users and channels is determined based on the communication requirements of users and channel status and in accordance with the goal of maximizing the data transmission rate, and the relationship between the data transmission rate and the signal-to-noise ratio satisfies: R i，j =B·log2(1+SNR i，j ), Among them, R i，j is the transmission rate of user i on channel j, B is the channel bandwidth, SNR i，j is the signal-to-noise ratio of user i on channel j.

5. The method for data communication transmission in a wireless communication network according to claim 1, characterized in that: When determining the optimal transmission path for data communication in step 3, a weighted shortest path algorithm is used to determine the path based on the network communication topology and the delay weight of each link to minimize the total transmission delay, which is expressed as: T total =∑ (i，j)∈E w i，j x i，j , Among them, T total is the total transmission delay, w i，j is the delay weight of link (i, j), x i，j is a link variable, a value of 1 indicates selected, and a value of 0 indicates unselected; The dynamic scheduling of the transmission path in step 3 is adjusted in combination with the congestion status of the link, and the path quality is predicted and updated in real time through a deep reinforcement learning model.

6. The method for data communication transmission in a wireless communication network according to claim 1, characterized in that: When the task is unloaded to the edge node in step 4, the optimal task unloading path is dynamically determined according to the task computing amount and the edge node computing resources, and the unloading time and the transmission delay are taken as the optimization target; The execution time of the task offloading in step 4 is determined by the task computation amount L and the processing capacity f of the edge computing node. edge It is determined that the execution time satisfies the following relationship: Among them, T edge is the edge processing time of the task, L is the task computation amount, f edge The processing power of the edge node.

7. A method for data communication transmission in a wireless communication network according to claim 6, characterized in that: When the task is offloaded to the edge node in step 4, in order to further reduce energy consumption and task execution time, the energy consumption optimization model is used to jointly optimize the task's computational workload and transmission power. The optimization goal is to minimize the total energy consumption E total , the total energy consumption includes computing energy consumption and transmission energy consumption, expressed as: Among them, E total is the total energy consumption of task offloading, P transmit is the transmission power during the task transmission process, P transmit is the data transmission time of the task, k is the energy consumption coefficient of the edge computing node, and f edge is the processing capacity of the edge node, T edge Processing time for the edge of the task.

8. The method for data communication transmission in a wireless communication network according to claim 5, characterized in that: When the path quality is predicted and updated in real time by the deep reinforcement learning model in step 3, the path determination process is optimized by using the state-action value function Q. The optimal decision for path determination satisfies the following update relationship: Among them, Q(s, a) is the value function of taking action a in state s, R is the immediate reward after performing action a, γ is the discount factor, s ′ is the next state after the current state s executes action a, a ′ is the optimal action in the next state.

9. The method for data communication transmission in a wireless communication network according to claim 4, characterized in that: When dynamically allocating available channel resources of the wireless communication network in step 2, a multi-objective optimization model is used to dynamically allocate channel resources by considering user service quality constraints and spectrum resource utilization. The channel conflict probability is related to the number of users and the allocation scheme, and is expressed as: Among them, P conflict is the channel collision probability, p i，j is the probability of collision between user i and channel j, and M is the number of users in the wireless communication network.

10. The method for data communication transmission in a wireless communication network according to claim 7, characterized in that: When the data stream is divided into multiple data streams for multi-path transmission in step 5, in order to improve the reliability and delay sensitivity of transmission, the allocation ratio α of each path is dynamically determined. i , satisfying the following constraints: Among them, α i is the data flow allocation ratio of path i, N is the number of available transmission paths, L is the total amount of data to be transmitted, and R i is the data transmission rate on path i, T total is the total delay of data transmission.

Images