Wireless communication data efficient transmission method, system and product

Abstract

The invention belongs to the technical field of wireless communication, and discloses a wireless communication data efficient transmission method, system and product. Comprising the following steps: constructing m data transmission paths by adopting an improved path selection algorithm; collecting channel quality data of each data transmission path in real time, and performing channel quality evaluation; screening an optimal transmission path from the m data transmission paths according to the channel quality; collecting data flow attribute data; performing feature extraction on the data stream attribute data to obtain data stream feature data; dynamically allocating channel characteristic parameters based on the channel quality of the optimal transmission path and the data stream characteristic data by adopting an improved optimization algorithm, and performing data transmission according to the channel characteristic parameters; the method can effectively adapt to the change of a network environment, realizes optimal network resource scheduling, reduces the delay and packet loss rate of data transmission, and improves the reliability and energy efficiency of data transmission.

Description

Technical Field

[0001] The present invention relates to the field of wireless communication technologies, and more specifically, to a method, a system, and a product for efficiently transmitting wireless communication data. Background Art

[0002] With the rapid development of information technology, wireless communication technology has become a core component of modern communication systems; the wide application of wireless communication not only promotes the progress of fields such as personal communication, Internet of Things (IoT), and smart home, but also provides support for emerging application scenarios such as industrial automation, smart city, and remote healthcare; however, with the explosion of device numbers and the increase in data traffic, traditional wireless communication technologies are facing challenges such as insufficient bandwidth, high transmission latency, and excessive energy consumption; in this context, it is particularly important to develop efficient wireless communication data transmission methods; it can not only improve the speed and reliability of data transmission, but also effectively reduce energy consumption, thereby extending the service life of devices.

[0003] The patent with the publication number CN114466033A discloses a method for distributing and interconnecting data transmission of wireless signal acquisition nodes; including: S1: selection of data transmission mode; S2: deployment of hardware; S3: determination of data transmission path; S4: judgment of whether the data is trustworthy; during the process of wireless data transmission at the signal acquisition nodes, the selection of data transmission paths between multiple nodes and related logical judgments are added to ensure the integrity of the original data; when reselecting the data transmission path, logical judgments need to be made on communication protocols, data transmission length, transmission time, and the remaining receiving window of the node, and data forwarding and re-uploading can only be performed through other wireless signal acquisition nodes when all conditions are met; this invention is mainly used to ensure the data reliability and integrity of wireless transmission technology when making efficient connections in occasions with long standby time and long network transmission distance.

[0004] However, although the above method can select the transmission path through factors such as communication protocols, data transmission length, transmission time, and the remaining receiving window of the node, it only relies on preset logical judgments and lacks dynamic channel quality assessment and real-time consideration of data flow characteristics, that is, it does not fully utilize the dynamic changes of channel quality and data flow characteristics to optimize path selection and resource allocation; therefore, when the network environment changes or the data flow characteristics are unstable, optimal resource scheduling cannot be achieved, thus affecting the data transmission efficiency and reliability.

[0005] In view of this, the present invention proposes a method, a system, and a product for efficiently transmitting wireless communication data to solve the above problems. Summary of the Invention

[0006] To overcome the above-mentioned defects of the prior art and achieve the above object, the present invention provides the following technical solution: A method for efficient transmission of wireless communication data, including:

[0007] Adopt an improved path selection algorithm to construct m data transmission paths;

[0008] Collect the channel quality data of each data transmission path in real time and conduct channel quality evaluation;

[0009] According to the channel quality, screen out the best transmission path from the m data transmission paths;

[0010] Collect data stream attribute data;

[0011] Extract features from the data stream attribute data to obtain data stream feature data;

[0012] Adopt an improved optimization algorithm, based on the channel quality of the best transmission path and the data stream feature data, dynamically allocate channel characteristic parameters, and conduct data transmission according to the channel characteristic parameters.

[0013] Further, the step of constructing m data transmission paths includes:

[0014] Step S101: According to the starting node and the ending node, obtain the network topology structure, which includes the nodes between the starting node and the ending node and the connections between the nodes;

[0015] Step S102: Count the number of connections of each node in the network topology structure;

[0016] Step S103: Mark the nodes with 1 or 0 connections as boundary nodes, and delete each boundary node that is not the ending node from the network topology structure;

[0017] Step S104: Starting from the starting node, explore all the nodes adjacent to the starting node, randomly select a node as the successor node, and mark it as the selected node;

[0018] Step S105: Explore all the nodes adjacent to the successor node, randomly select a node as the new successor node, and mark it as the selected node;

[0019] Step S106: Loop step S105 until all the nodes adjacent to the successor node are marked as selected nodes, then the loop ends, and an initial transmission path is obtained;

[0020] Step S107: Judge whether the last node in the initial transmission path is the ending node. If so, take the initial transmission path as the data transmission path. If not, delete the initial data path;

[0021] Step S108: Loop through steps S104 to S107 to obtain m data transmission paths; where m is an integer greater than 1, and the m data transmission paths are all different.

[0022] Further, the channel quality data includes delay, packet loss rate, and signal-to-noise ratio;

[0023] The method for obtaining the signal-to-noise ratio corresponding to each data transmission path is as follows: Measure the corresponding signal-to-noise ratio at each node on each data transmission path and mark it as the unit signal-to-noise ratio; Count the number of nodes in each data transmission path and mark it as the number of nodes; Add up the unit signal-to-noise ratios corresponding to each data transmission path in sequence, and then divide by the corresponding number of nodes to obtain the signal-to-noise ratio corresponding to each data transmission path;

[0024] The method for evaluating channel quality includes:

[0025] Preset a sliding length, where the sliding length is the length of the sliding window; According to the sliding length, obtain the historical quality data of each data transmission path; The historical quality data is the channel quality data obtained at historical moments, and the acquisition moments corresponding to each channel quality data in the historical quality data are connected in sequence, and the latest acquisition moment corresponding to the historical quality data is adjacent to the acquisition moment of the real-time quality data, and the real-time quality data is the channel quality data collected in real time;

[0026] Increment the sliding length by one to obtain an extended length; Add up the same data in the historical quality data and the real-time quality data corresponding to each data transmission path in sequence, and then divide by the extended length to obtain the average quality data corresponding to each data transmission path; The average quality data includes average delay, average packet loss rate, and average signal-to-noise ratio; Preset a weight set, where the weight set includes the weight coefficients corresponding to each data in the average quality data; Multiply each data in the average quality data corresponding to each data transmission path by the corresponding weight coefficient and add them up in sequence to obtain the channel quality corresponding to each data transmission path.

[0027] Further, the method for screening out the best transmission path from the m data transmission paths includes:

[0028] According to the network topology structure, obtain the transmission distance between every two adjacent nodes; According to the nodes in the m data transmission paths, obtain the transmission distance corresponding to each data transmission path; Add up the transmission distances corresponding to each data transmission path in sequence to obtain the path distance corresponding to each data transmission path;

[0029] Mark each node in each data transmission path as a transmission node; use the delay and packet loss rate of each data transmission path as the delay and packet loss rate of the corresponding transmission node; multiply the delay, packet loss rate, and unit signal-to-noise ratio corresponding to each transmission node by the corresponding weight coefficient, and add them up in sequence to obtain the node quality corresponding to each transmission node; use the node quality corresponding to each data transmission path as a set of nodes, and the set of nodes corresponds to the data transmission path one by one; count the number corresponding to different node qualities in each set of nodes and mark it as the quality number; divide each quality number by the number of nodes in the corresponding set of nodes to obtain the occurrence probability of different node qualities in each set of nodes; perform a natural logarithm operation on each occurrence probability, and then multiply it by the same occurrence probability to obtain the information amount corresponding to each occurrence probability; add up the information amounts corresponding to each set of nodes in sequence, and then take the opposite number to obtain the relative entropy corresponding to each set of nodes.

[0030] Preset an adjustment parameter, multiply the relative entropy of each set of nodes by the adjustment parameter and then add one to obtain the adjusted entropy of each set of nodes; use the reciprocal of the adjusted entropy corresponding to each set of nodes as the proportionality coefficient corresponding to the set of nodes; perform normalization processing on the channel quality and path distance corresponding to each data transmission path to obtain the quality standard and distance standard; multiply the quality standard of each data transmission path by the corresponding proportionality coefficient, and then subtract the corresponding path standard to obtain the screening coefficient of each data transmission path; sort the screening coefficients of each data transmission path from largest to smallest, and use the data transmission path corresponding to the screening coefficient ranked first as the best transmission path.

[0031] Further, the data stream attribute data includes the data stream size, data stream type, and data stream source.

[0032] The method for obtaining the data stream feature data includes:

[0033] Set different digital tags for different data stream types and mark them as type tags; set different digital tags for different data stream sources and mark them as source tags; use the type tags corresponding to the data stream type and the source tags corresponding to the data stream source in the data stream attribute data as analysis data, and input the analysis data into the trained priority prediction model to predict the priority of the corresponding data stream and use it as the data stream feature data; the priority prediction model is a deep neural network model.

[0034] The training process of the priority prediction model includes:

[0035] Pre-collect a set of analysis data, set corresponding priorities for each set of analysis data, where a is an integer greater than 1, and convert the analysis data and the corresponding priorities into a corresponding set of feature vectors; use each set of feature vectors as the input of the priority prediction model, and the priority prediction model outputs a set of predicted priorities corresponding to each set of analysis data, with the actual priority corresponding to each set of analysis data as the prediction target, and the actual priority is the pre-set priority corresponding to the analysis data; use minimizing the sum of prediction errors of all analysis data as the training target; train the priority prediction model until the sum of prediction errors converges and then stop training.

[0036] Further, the channel characteristic parameters include bandwidth, frequency, and time slot;

[0037] The steps of dynamically allocating channel characteristic parameters include:

[0038] Step S201: Obtain the parameter range and construct n sets of parameter collections;

[0039] Step S202: Set different digital labels for each of the n sets of parameter collections and mark them as set labels; construct a search interval [1, n] according to the set labels corresponding to the n sets of parameter collections;

[0040] Step S203: Define the population size b and the maximum number of iterations T max ;

[0041] Step S204: Construct an initial population according to the population size b, and the iteration number t of the initial population is 0; randomly generate the positions of each individual in the population, and each individual's position corresponds one-to-one with the set labels within the search interval;

[0042] Step S205: Define the optimization value function;

[0043] Step S206: Calculate the optimization value of each individual, and take the individual with the largest optimization value as the optimal individual;

[0044] Step S207: Determine the individual behavior according to the optimization value of each individual;

[0045] Step S208: Update the position of each individual according to the individual behavior and the position of the optimal individual;

[0046] Step S209: Let t = t + 1;

[0047] Step S210: Loop through steps S206 to S209 until t ≥ T max When the loop ends, enter step S211;

[0048] Step S211: Obtain the optimal individual in the population, use the set label corresponding to the position of the optimal individual as the optimal label, and dynamically allocate the channel characteristic parameters according to the parameter set corresponding to the optimal label.

[0049] Further, in the step S201, the parameter range includes the bandwidth range, frequency range, and time slot range of the best transmission path; the method for constructing n sets of parameter sets is: randomly select a value from each range within the parameter range to construct a set of parameter sets, and a total of n sets of parameter sets are constructed, and the n sets of parameter sets are all different;

[0050] In the step S204, the generation steps of the position of each individual include:

[0051] Step S301: Preset the number of levels d;

[0052] Step S302: Uniformly divide the search interval according to the number of levels to obtain d sub-search intervals, where the width of each sub-search interval is

[0053] Step S303: Generate the position of an individual from each sub-search interval in turn;

[0054] Step S304: Loop step S303 until the positions of b individuals are generated, and the loop ends;

[0055] In the step S303, the method for generating the individual position from the sub-search interval includes:

[0056] Mark the minimum value of the sub-search interval as the minimum search value, and mark the maximum value of the sub-search interval as the maximum search value; subtract the minimum search value from the maximum search value to obtain the search difference; multiply the search difference by a random number within the range of [0,1], and then add the minimum search value to obtain the position of the individual;

[0057] In the step S205, the expression of the optimization value function is f = cx + xz; where f is the optimization value, cx is the efficiency standard, and xz is the signal standard; the methods for obtaining the efficiency standard and the signal standard include:

[0058] Replace the data flow type in the data flow attribute data with the corresponding type label, and replace the data flow source with the corresponding source label; mark the data flow attribute data after replacement as replacement data, and mark the channel quality of the optimal transmission path as the optimal quality; use the replacement data, data flow feature data, optimal quality, and the corresponding set label of the individual as evaluation data; input the evaluation data into the trained efficiency prediction model to predict the corresponding transmission efficiency; input the evaluation data into the trained quality prediction model to predict the corresponding signal quality; wherein, the training processes of the efficiency prediction model and the quality prediction model are both the same as the training process of the priority prediction model, and both are deep neural network models; perform normalization processing on the transmission efficiency and the signal quality respectively to obtain the efficiency standard and the signal standard.

[0059] Further, in the step S207, the method for determining the individual behavior includes:

[0060] Add up the optimization values of each individual in sequence, and then divide by b to obtain the optimization mean value; compare the optimization value of each individual with the optimization mean value in sequence, mark the individuals with optimization values greater than or equal to the optimization mean value as excellent individuals, and mark the individuals with optimization values less than the optimization mean value as ordinary individuals; the individual behavior corresponding to the excellent individual is the approaching behavior, and the behavior corresponding to the ordinary individual is the exploring behavior;

[0061] In the step S208, the method for updating the position of each individual includes:

[0062] If the individual behavior is the approaching behavior, the corresponding method for updating the position includes:

[0063] Preset a step size factor and an influence factor; mark the position of the optimal individual as the optimal position, and mark the position of the current individual as the current position; subtract the current position from the optimal position, and then multiply by the step size factor to obtain the moving step size; mark all individuals other than the optimal individual and the current individual as the remaining individuals; mark the positions of the remaining individuals as the remaining positions, subtract the current position from each remaining position in sequence to obtain the position difference; add up all the position differences in sequence, and then multiply by the influence factor to obtain the influence step size; add the moving step size to the current position, and then add the influence step size to obtain the first updated position, and update the current position according to the first updated position;

[0064] If the individual behavior is the exploring behavior, the corresponding method for updating the position includes:

[0065] A preset avoidance factor; compare the optimization values of all individuals respectively, mark the individual with the smallest optimization value as the worst individual, and mark the position of the worst individual as the worst position; subtract the worst position from the current position, and then multiply by the avoidance factor to obtain the avoidance step length; add the avoidance step length to the current position to obtain the second updated position, and update the current position according to the second updated position.

[0066] A wireless communication data efficient transmission system, implementing the described wireless communication data efficient transmission method, includes:

[0067] A path construction module, used to construct m data transmission paths by adopting an improved path selection algorithm;

[0068] A quality evaluation module, used to collect the channel quality data of each data transmission path in real time and conduct channel quality evaluation;

[0069] A path screening module, used to screen out the best transmission path from m data transmission paths according to the channel quality;

[0070] A data collection module, used to collect data stream attribute data;

[0071] A feature extraction module, used to extract features from the data stream attribute data to obtain data stream feature data;

[0072] A characteristic allocation module, used to adopt an improved optimization algorithm, dynamically allocate channel characteristic parameters based on the channel quality of the best transmission path and the data stream feature data, and conduct data transmission according to the channel characteristic parameters.

[0073] A wireless communication data efficient transmission product, integrating the described wireless communication data efficient transmission system.

[0074] The technical effects and advantages of the wireless communication data efficient transmission method, system and product of the present invention:

[0075] Construct multiple data transmission paths by adopting an improved path selection algorithm, and evaluate the channel quality of each path in real time to select the best transmission path; at the same time, use deep learning technology to extract the features of the data stream, and combine the channel quality of the best transmission path, and apply an optimization algorithm to dynamically allocate channel characteristic parameters, including bandwidth, frequency and time slot; based on the dynamic optimization mechanism that combines path quality and data stream features, it can effectively adapt to the changes in the network environment, achieve optimal network resource scheduling, reduce the latency and packet loss rate of data transmission, improve the reliability and energy efficiency of data transmission, so as to meet the high-efficiency transmission requirements in complex wireless network environments. Description of the Drawings

[0076] Figure 1Schematic diagram of a high - efficiency wireless communication data transmission system according to Embodiment 1 of the present invention;

[0077] Figure 2 Flowchart of a high - efficiency wireless communication data transmission method according to Embodiment 2 of the present invention. Detailed implementation manners

[0078] Next, the technical solutions in the embodiments of the present invention will be clearly and completely described in conjunction with the accompanying drawings in the embodiments of the present invention. Obviously, the described embodiments are only a part of the embodiments of the present invention, rather than all the embodiments. All other embodiments obtained by those of ordinary skill in the art based on the embodiments of the present invention without creative efforts shall fall within the protection scope of the present invention.

[0079] Embodiment 1

[0080] Please refer to Figure 1 As shown, a high - efficiency wireless communication data transmission system in this embodiment includes a path construction module, a quality assessment module, a path screening module, a data acquisition module, a feature extraction module, and a characteristic assignment module; each module is connected by wired and / or wireless means to achieve data transmission between modules.

[0081] The path construction module is used to construct m data transmission paths by using an improved path selection algorithm.

[0082] The steps of constructing m data transmission paths include:

[0083] Step S101: According to the start node and the end node, obtain the network topology structure, which includes the nodes (such as routers, switches, etc.) between the start node and the end node and the connections between the nodes; the start node is the source in the data transmission process, responsible for initiating the data sending, usually a sending device or a server; the end node is the destination in the data transmission process, responsible for receiving the data, usually a receiving device or a client; the network topology structure is obtained through network monitoring tools (such as Nagios, Zabbix, SolarWinds, etc.).

[0084] Step S102: Count the number of connections of each node in the network topology structure, where the number of connections is the number of connections between a node and other nodes.

[0085] Step S103: Mark the nodes with 1 or 0 connections as boundary nodes, and delete each boundary node that is not the end node from the network topology structure.

[0086] Step S104: Starting from the start node, explore all the nodes adjacent to the start node, randomly select a node as the successor node, and mark it as the selected node.

[0087] Step S105: Explore all the nodes adjacent to the successor node, randomly select a node as the new successor node, and mark it as the selected node;

[0088] Step S106: Loop step S105 until all the nodes adjacent to the successor node are marked as selected nodes, then the loop ends and an initial transmission path is obtained;

[0089] Step S107: Determine whether the last node in the initial transmission path is the termination node. If so, take the initial transmission path as the data transmission path. If not, delete the initial data path;

[0090] Step S108: Loop steps S104 to S107 to obtain m data transmission paths; where m is an integer greater than 1, and the m data transmission paths are all different.

[0091] It should be noted that the purpose of deleting the boundary nodes is that the boundary nodes cannot effectively participate in data transmission. Deleting the boundary nodes that contribute nothing to data transmission can reduce the complexity of the network, simplify the network topology structure, thereby reducing interference and unnecessary path selection, improving the efficiency of constructing data transmission paths, and further ensuring the effectiveness and stability of data transmission paths.

[0092] The quality assessment module is used to collect the channel quality data of each data transmission path in real time and conduct channel quality assessment.

[0093] The channel quality data includes delay, packet loss rate, and signal-to-noise ratio;

[0094] The delay is the time required for data to travel from the start node to the termination node; the higher the delay, the longer the response time of data transmission, which affects the performance of real-time applications, such as video conferencing or online games, thus resulting in a decline in channel quality, and vice versa.

[0095] The packet loss rate is the proportion of data packets lost during data transmission; the higher the packet loss rate, the more data packets are lost during transmission, affecting the integrity of the data, thus resulting in a decline in channel quality, and vice versa.

[0096] The signal-to-noise ratio is the ratio of the signal to the background noise; the lower the signal-to-noise ratio, the weaker the signal intensity and the greater the background noise, resulting in the signal being vulnerable to interference, reducing the reliability of data transmission, thus resulting in a decline in channel quality, and vice versa.

[0097] The latency and packet loss rate of each data transmission path are obtained by monitoring the traffic on each data transmission path through network monitoring tools (such as Wireshark, SolarWinds, etc.); the method for obtaining the signal-to-noise ratio corresponding to each data transmission path is as follows: use a wireless signal analyzer or Wi-Fi analysis tool to measure the corresponding signal-to-noise ratio at each node on each data transmission path, and mark it as the unit signal-to-noise ratio; count the number of nodes in each data transmission path and mark it as the number of nodes; add up the unit signal-to-noise ratios corresponding to each data transmission path in sequence, and then divide by the corresponding number of nodes to obtain the signal-to-noise ratio corresponding to each data transmission path.

[0098] The methods for evaluating the channel quality include:

[0099] Preset a sliding length, where the sliding length is the length of the sliding window, and the sliding length is preset by those skilled in the art according to the actual situation; according to the sliding length, obtain the historical quality data of each data transmission path from the system-built database; the historical quality data is the channel quality data obtained at historical moments, and the acquisition moments corresponding to each channel quality data in the historical quality data are connected in sequence, and the latest acquisition moment corresponding to the historical quality data is adjacent to the acquisition moment of the real-time quality data, and the real-time quality data is the channel quality data collected in real time;

[0100] Add one to the sliding length to obtain an extended length; add up the same-kind data in the historical quality data and the real-time quality data corresponding to each data transmission path in sequence, and then divide by the extended length to obtain the average quality data corresponding to each data transmission path; the average quality data includes average latency, average packet loss rate, and average signal-to-noise ratio; preset a weight set, where the weight set includes the weight coefficients corresponding to each data in the average quality data, and the weight set is preset by those skilled in the art according to the influence degree of each data in the average quality data on the channel quality; multiply each data in the average quality data corresponding to each data transmission path by the corresponding weight coefficient, and add them up in sequence to obtain the channel quality corresponding to each data transmission path.

[0101] A path screening module, configured to screen out the best transmission path from m data transmission paths according to the channel quality.

[0102] The method for screening out the best transmission path from m data transmission paths includes:

[0103] According to the network topology structure, obtain the transmission distance between every two adjacent nodes; according to the nodes in the m data transmission paths, obtain the transmission distance corresponding to each data transmission path; add up the transmission distances corresponding to each data transmission path in sequence to obtain the path distance corresponding to each data transmission path.

[0104] Mark each node in each data transmission path as a transmission node; use the delay and packet loss rate of each data transmission path as the delay and packet loss rate of the corresponding transmission node; multiply the delay, packet loss rate, and unit signal-to-noise ratio corresponding to each transmission node by the corresponding weight coefficient and add them up in sequence to obtain the node quality corresponding to each transmission node; use the node quality corresponding to each data transmission path as a set of nodes, and the set of nodes corresponds to the data transmission path one by one; count the number corresponding to different node qualities in each set of nodes and mark it as the quality quantity; divide each quality quantity by the number of nodes in the corresponding set of nodes to obtain the occurrence probability of different node qualities in each set of nodes; perform a natural logarithm operation on each occurrence probability and then multiply it by the same occurrence probability to obtain the information amount corresponding to each occurrence probability; add up the information amounts corresponding to each set of nodes in sequence and then take the opposite number to obtain the relative entropy corresponding to each set of nodes.

[0105] Preset adjustment parameters, which are preset by those skilled in the art according to the actual situation; multiply the relative entropy of each set of nodes by the adjustment parameter and then add one to obtain the adjusted entropy of each set of nodes; use the reciprocal of the adjusted entropy corresponding to each set of nodes as the proportionality coefficient corresponding to the set of nodes; perform normalization processing (such as z-score normalization, min-max normalization, etc.) on the channel quality and path distance corresponding to each data transmission path to obtain the quality standard and distance standard; multiply the quality standard of each data transmission path by the corresponding proportionality coefficient and then subtract the corresponding path standard to obtain the screening coefficient of each data transmission path; sort the screening coefficients of each data transmission path from large to small, and use the data transmission path corresponding to the screening coefficient ranked first as the best transmission path.

[0106] A data acquisition module for acquiring data stream attribute data.

[0107] The data stream attribute data includes the data stream size, data stream type, and data stream source.

[0108] The data stream size is the total amount of data to be transmitted, which is obtained through a network monitoring tool; the data stream type is the type of data carried by the data stream, such as text, image, audio, video, etc., which is obtained through the metadata of the transmission protocol, such as the Content-Type field in the HTTP request and response headers; the data stream source is the generating device or system of the data stream, such as a sensor, user device, server, etc., which is obtained through the metadata in the data packet.

[0109] It should be noted that the purpose of collecting data stream attribute data is to optimize network performance and transmission efficiency. Among them, obtaining the data stream size helps to reasonably allocate bandwidth and ensure the effective allocation of network resources. Identifying the data stream type and data stream source helps to adjust the transmission strategy and determine the processing priority of data.

[0110] A feature extraction module is used to extract features from the data stream attribute data to obtain data stream feature data.

[0111] The methods for obtaining data stream feature data include:

[0112] Set different digital tags for different data stream types and mark them as type tags. Set different digital tags for different data stream sources and mark them as source tags. Use the type tags corresponding to the data stream type and the source tags corresponding to the data stream source in the data stream attribute data as analysis data, and input the analysis data into the trained priority prediction model to predict the priority of the corresponding data stream, which is used as the data stream feature data.

[0113] The training process of the priority prediction model includes:

[0114] Pre-collect a set of analysis data, and set corresponding priorities for the a set of analysis data. a is an integer greater than 1. Convert the analysis data and the corresponding priorities into a corresponding set of feature vectors. The priorities corresponding to the analysis data are determined by those skilled in the art in the process of determining the data stream priorities in history. Collect a set of analysis data, and according to the data stream type and data stream source corresponding to each set of analysis data, set corresponding priorities in combination with the actual situation, and sequentially set corresponding priorities for the a set of analysis data.

[0115] Use each set of feature vectors as the input of the priority prediction model. The priority prediction model outputs a set of predicted priorities corresponding to each set of analysis data, and uses the actual priority corresponding to each set of analysis data as the prediction target. The actual priority is the priority preset corresponding to the analysis data. Use minimizing the sum of the prediction errors of all analysis data as the training target. Among them, the calculation formula of the prediction error is η w =(φ w -ε w ) 2 , where η w is the prediction error, w is the group number of the feature vector corresponding to the analysis data, θ w is the predicted priority corresponding to the w-th set of analysis data, and ε w is the actual priority corresponding to the w-th set of analysis data. Train the priority prediction model until the sum of the prediction errors reaches convergence and then stop training.

[0116] The above-mentioned priority prediction model is specifically a deep neural network model, which includes an input layer, a hidden layer, and an output layer. Each hidden layer contains multiple neurons, and there are connections between each neuron and the neurons in the next layer. These connections include weights, which determine the importance and influence of data transmission in the neural network. An activation function is applied to each neuron between the hidden layer and the output layer. The activation function introduces non-linearity and allows the network to learn more complex patterns and features.

[0117] A feature allocation module is used to dynamically allocate channel characteristic parameters based on the channel quality of the optimal transmission path and the data stream feature data by adopting an improved optimization algorithm, and perform data transmission according to the channel characteristic parameters.

[0118] The channel characteristic parameters include bandwidth, frequency, and time slot.

[0119] The bandwidth is the maximum amount of data that can be transmitted per unit time on the data transmission path, representing the transmission capacity of the network. The frequency is the number of times the signal changes per unit time, determining the fluctuation rate of the signal. The time slot is a unit of time division for multiplexing, which divides the transmission bandwidth into multiple time periods, allowing multiple signals to be transmitted sequentially on the same data transmission path, and each time slot is occupied by a specific data stream.

[0120] The steps for dynamically allocating channel characteristic parameters include:

[0121] Step S201: Obtain the parameter range and construct n groups of parameter sets.

[0122] Step S202: Set different digital labels for each of the n groups of parameter sets and mark them as set labels. According to the set labels corresponding to the n groups of parameter sets, construct a search interval [1, n].

[0123] Step S203: Define the population size b and the maximum number of iterations T max ;

[0124] Step S204: Construct an initial population according to the population size b, and the iteration number t of the initial population is 0. Randomly generate the positions of each individual in the population, and each individual's position corresponds one-to-one with the set labels within the search interval.

[0125] Step S205: Define the optimization value function.

[0126] Step S206: Calculate the optimization value of each individual, and take the individual with the largest optimization value as the optimal individual.

[0127] Step S207: Determine the individual behavior according to the optimization value of each individual.

[0128] Step S208: Update the position of each individual according to the individual behavior and the position of the optimal individual;

[0129] Step S209: Let t = t + 1;

[0130] Step S210: Loop steps S206 to S209 until t ≥ T max When the loop ends, proceed to step S211;

[0131] Step S211: Obtain the optimal individual in the population, take the set label corresponding to the position of the optimal individual as the optimal label, and dynamically allocate the channel characteristic parameters according to the parameter set corresponding to the optimal label.

[0132] In the above step S201, the parameter range includes the bandwidth range, frequency range, and time slot range of the optimal transmission path; the parameter range is obtained by those skilled in the art through channel measurement of the optimal transmission path in the actual environment and in combination with literature research; the method for constructing n groups of parameter sets is: randomly select a value from each range within the parameter range to construct a group of parameter sets, and a total of n groups of parameter sets are constructed, and the n groups of parameter sets are all different.

[0133] In the above step S203, the population size b and the maximum number of iterations T max are both preset by those skilled in the art according to the algorithm accuracy requirements.

[0134] In the above step S204, the generation steps of the position of each individual include:

[0135] Step S301: Preset the number of levels d, and the number of levels is preset by those skilled in the art according to the actual situation;

[0136] Step S302: Uniformly divide the search interval according to the number of levels to obtain d sub-search intervals, where the width of each sub-search interval is

[0137] Step S303: Generate the position of an individual from each sub-search interval in turn;

[0138] Step S304: Loop step S303 until the positions of b individuals are generated, and the loop ends.

[0139] In the above step S303, the method for generating the individual position from the sub-search interval includes:

[0140] Mark the minimum value of the sub-search interval as the minimum search value, and mark the maximum value of the sub-search interval as the maximum search value; subtract the minimum search value from the maximum search value to obtain the search difference; multiply the search difference by a random number within the interval [0, 1], and then add the minimum search value to obtain the position of the individual.

[0141] In the above step S205, the expression of the optimized value function is f = cx + xz; where f is the optimized value, cx is the efficiency criterion, and xz is the signal criterion; the methods for obtaining the efficiency criterion and the signal criterion include:

[0142] Replace the data stream type in the data stream attribute data with the corresponding type label, and replace the data stream source with the corresponding source label; mark the data stream attribute data after replacement as replacement data, and mark the channel quality of the optimal transmission path as the optimal quality; use the replacement data, data stream feature data, optimal quality, and the corresponding set label of the individual as evaluation data; input the evaluation data into the trained efficiency prediction model to predict the corresponding transmission efficiency; input the evaluation data into the trained quality prediction model to predict the corresponding signal quality; among them, the training processes of the efficiency prediction model and the quality prediction model are both the same as the training process of the priority prediction model, and both are deep neural network models; perform normalization processing on the transmission efficiency and the signal quality respectively to obtain the efficiency criterion and the signal criterion.

[0143] In the above step S207, the method for determining the individual behavior includes:

[0144] Add up the optimized values of each individual in turn, and then divide by b to obtain the optimized mean value; compare the optimized value of each individual with the optimized mean value in turn, mark the individual with an optimized value greater than or equal to the optimized mean value as an excellent individual, and mark the individual with an optimized value less than the optimized mean value as an ordinary individual; the corresponding individual behavior of the excellent individual is the approaching behavior, and the corresponding behavior of the ordinary individual is the exploring behavior; where the approaching behavior is the behavior of the excellent individual approaching the optimal individual, and the exploring behavior is the behavior of the ordinary individual searching for an individual with a higher optimized value in the search interval.

[0145] In the above step S208, the method for updating the position of each individual includes:

[0146] If the individual behavior is the approaching behavior, the corresponding method for position update includes:

[0147] Preset a step factor and an influence factor, which are preset by those skilled in the art according to the required algorithm convergence speed and algorithm accuracy; mark the position of the optimal individual as the optimal position, and mark the position of the current individual as the current position; subtract the current position from the optimal position, and then multiply by the step factor to obtain the movement step; mark all individuals other than the optimal individual and the current individual as the remaining individuals; mark the positions of the remaining individuals as the remaining positions, subtract the current position from each remaining position in turn to obtain the position difference; add up all the position differences in turn, and then multiply by the influence factor to obtain the influence step; add the movement step to the current position, and then add the influence step to obtain the first updated position, and update the current position according to the first updated position.

[0148] If the individual behavior is a exploration behavior, the method for updating the corresponding position includes:

[0149] Preset an avoidance factor, which is preset by those skilled in the art according to the actual situation; compare the optimization values of all individuals respectively, mark the individual with the smallest optimization value as the worst individual, and mark the position of the worst individual as the worst position; subtract the worst position from the current position, and then multiply by the avoidance factor to obtain the avoidance step; add the avoidance step to the current position to obtain the second updated position, and update the current position according to the second updated position.

[0150] In this embodiment, an improved path selection algorithm is used to construct multiple data transmission paths, and the channel quality of each path is evaluated in real time, and the best transmission path is selected from them; at the same time, deep learning technology is used to extract the characteristics of the data stream, and combined with the channel quality of the best transmission path, an optimization algorithm is applied to dynamically allocate channel characteristic parameters, including bandwidth, frequency and time slot; based on the dynamic optimization mechanism that combines path quality and data stream characteristics, it can effectively adapt to the changes in the network environment, achieve optimal network resource scheduling, reduce the latency and packet loss rate of data transmission, improve the reliability and energy efficiency of data transmission, so as to meet the high-efficiency transmission requirements in complex wireless network environments.

[0151] Embodiment 2

[0152] Please refer to Figure 2 As shown, for the parts not described in detail in this embodiment, refer to the description content of Embodiment 1. A method for efficient transmission of wireless communication data is provided, and the method includes:

[0153] Adopt an improved path selection algorithm to construct m data transmission paths;

[0154] Collect the channel quality data of each data transmission path in real time and conduct channel quality evaluation;

[0155] According to the channel quality, screen out the best transmission path from the m data transmission paths;

[0156] Collect data stream attribute data;

[0157] Extract features from the data stream attribute data to obtain data stream feature data;

[0158] Adopt the improved optimization algorithm, dynamically allocate channel characteristic parameters based on the channel quality of the best transmission path and the data stream feature data, and perform data transmission according to the channel characteristic parameters.

[0159] Embodiment 3

[0160] This embodiment provides a product for efficient transmission of wireless communication data, integrating a system for efficient transmission of wireless communication data in Embodiment 1.

[0161] Embodiment 4

[0162] This application also provides an electronic device. The electronic device may include one or more processors and one or more memories. Among them, the memory stores computer-readable code, and when the computer-readable code is run by one or more processors, it can execute a method for efficient transmission of wireless communication data as described above.

[0163] The method or system according to the embodiment of the present application can also be implemented by means of the architecture of the electronic device shown in the present application. The electronic device may include a bus, one or more CPUs, ROM, RAM, a communication port connected to the network, input / output, a hard disk, etc. The storage device in the electronic device, such as ROM or hard disk, can store a method for efficient transmission of wireless communication data provided by the present application. Further, the electronic device may also include a user interface. Of course, the architecture shown in the present application is only exemplary, and when implementing different devices, one or more components shown in the electronic device of the present application can be omitted according to actual needs.

[0164] Embodiment 5

[0165] One embodiment of the present application discloses a computer-readable storage medium. Computer-readable instructions are stored on the computer-readable storage medium. When the computer-readable instructions are run by a processor, a method for efficient transmission of wireless communication data according to the embodiment of the present application described with reference to the above drawings can be executed. The storage medium includes but is not limited to, for example, volatile memory and / or non-volatile memory. Volatile memory may include, for example, random access memory (RAM) and cache memory, etc. Non-volatile memory may include, for example, read-only memory (ROM), hard disk, flash memory, etc.

[0166] In addition, according to an embodiment of the present application, the process described above with reference to the flowchart can be implemented as a computer software program. For example, the present application provides a non-transitory machine-readable storage medium storing machine-readable instructions that can be run by a processor to execute instructions corresponding to the method steps provided by the present application, such as: a method for efficient transmission of wireless communication data. When the computer program is executed by a central processing unit (CPU), the above functions defined in the method of the present application are executed.

[0167] The above is only a specific embodiment of the present invention, but the protection scope of the present invention is not limited thereto. Any person skilled in the art within the technical scope disclosed by the present invention can easily think of changes or substitutions, which should be covered by the protection scope of the present invention. Therefore, the protection scope of the present invention shall be subject to the protection scope of the claims.

[0168] Finally, the above is only the preferred embodiment of the present invention and is not used to limit the present invention. Any modifications, equivalent substitutions, improvements, etc. made within the spirit and principle of the present invention shall be included in the protection scope of the present invention.

Claims

1. A method for efficient transmission of wireless communication data, characterized in that: include: Using the improved path selection algorithm, m data transmission paths are constructed; Collect channel quality data of each data transmission path in real time and perform channel quality assessment; According to the channel quality, the best transmission path is selected from m data transmission paths; Collect data stream attribute data; Extract features from data stream attribute data to obtain data stream feature data; The improved optimization algorithm is used to dynamically allocate channel characteristic parameters based on the channel quality of the best transmission path and data stream characteristic data, and data is transmitted according to the channel characteristic parameters.

2. A wireless communication data efficient transmission method according to claim 1, characterized in that: The step of constructing m data transmission paths comprises: Step S101: acquiring a network topology structure according to a starting node and an ending node, wherein the network topology structure includes nodes between the starting node and the ending node and connections between the nodes; Step S102: Counting the number of connections of each node in the network topology; Step S103: marking nodes with a connection number of 1 or 0 as boundary nodes, and deleting each boundary node that is not a terminal node from the network topology structure; Step S104: starting from the starting node, explore all nodes adjacent to the starting node, randomly select a node as a successor node, and mark it as a screening node; Step S105: Explore all nodes adjacent to the successor node, randomly select a node as a new successor node, and mark it as a selected node; Step S106: looping step S105 until all nodes adjacent to the successor node are marked as screening nodes, the loop ends, and an initial transmission path is obtained; Step S107: determining whether the last node in the initial transmission path is a termination node, if so, using the initial transmission path as the data transmission path, if not, deleting the data initial path; Step S108: looping steps S104 to S107 to obtain m data transmission paths; wherein m is an integer greater than 1, and the m data transmission paths are all different.

3. A wireless communication data efficient transmission method according to claim 2, characterized in that: The channel quality data includes delay, packet loss rate and signal-to-noise ratio; The method for obtaining the signal-to-noise ratio corresponding to each data transmission path is as follows: the corresponding signal-to-noise ratio is measured at each node on each data transmission path and marked as a unit signal-to-noise ratio; the number of nodes in each data transmission path is counted and marked as the number of nodes; the unit signal-to-noise ratios corresponding to each data transmission path are added in sequence, and then divided by the corresponding number of nodes to obtain the signal-to-noise ratio corresponding to each data transmission path; Methods for evaluating channel quality include: A sliding length is preset, and the sliding length is the length of the sliding window; according to the sliding length, historical quality data of each data transmission path is obtained; the historical quality data is the channel quality data obtained at a historical moment, and the collection moments corresponding to each channel quality data in the historical quality data are sequentially connected, and the latest collection moment corresponding to the historical quality data is adjacent to the collection moment of the real-time quality data, and the real-time quality data is the channel quality data collected in real time; Add one to the sliding length to obtain the extended length; add the historical quality data corresponding to each data transmission path and the same data in the real-time quality data in sequence, and then divide by the extended length to obtain the average quality data corresponding to each data transmission path; the average quality data includes the average delay, the average packet loss rate and the average signal-to-noise ratio; a preset weight set is set, and the weight set includes the weight coefficient corresponding to each data in the average quality data; multiply each data in the average quality data corresponding to each data transmission path by the corresponding weight coefficient, and add them in sequence to obtain the channel quality corresponding to each data transmission path.

4. The method for efficient wireless communication data transmission according to claim 3, characterized in that: The method for selecting the best transmission path from m data transmission paths includes: According to the network topology, the transmission distance between every two adjacent nodes is obtained; according to the nodes in the m data transmission paths, the transmission distance corresponding to each data transmission path is obtained; the transmission distance corresponding to each data transmission path is added in sequence to obtain the path distance corresponding to each data transmission path; Mark all nodes in each data transmission path as transmission nodes; take the delay and packet loss rate of each data transmission path as the delay and packet loss rate of the corresponding transmission node; multiply the delay, packet loss rate and unit signal-to-noise ratio corresponding to each transmission node by the corresponding weight coefficient, and add them up in sequence to obtain the node quality corresponding to each transmission node; take the node quality corresponding to each data transmission path as a set of node sets, and the node sets correspond to the data transmission paths one by one; count the number of different node qualities in each node set and mark them as quality numbers; divide each quality number by the number of nodes in the corresponding node set to obtain the probability of occurrence of different node qualities in each node set; perform natural logarithm operation on each occurrence probability, and then multiply it by the same occurrence probability to obtain the amount of information corresponding to each occurrence probability; add the amount of information corresponding to each node set in sequence, and then take the inverse to obtain the relative entropy corresponding to each node set; Preset adjustment parameters, multiply the relative entropy of each node set by the adjustment parameter and add one to obtain the adjustment entropy of each node set; use the inverse of the adjustment entropy corresponding to each node set as the proportional coefficient of the corresponding node set; normalize the channel quality and path distance corresponding to each data transmission path to obtain the quality standard and distance standard; multiply the quality standard of each data transmission path by the corresponding proportional coefficient, and then subtract the corresponding path standard to obtain the screening coefficient of each data transmission path; sort the screening coefficients of each data transmission path from large to small, and use the data transmission path corresponding to the front screening coefficient as the optimal transmission path.

5. The method for efficient wireless communication data transmission according to claim 4, characterized in that: The data stream attribute data includes data stream size, data stream type and data stream source; The method for obtaining data stream characteristic data comprises: Different digital labels are set for different data flow types and marked as type labels; different digital labels are set for different data flow sources and marked as source labels; the type label corresponding to the data flow type in the data flow attribute data and the source label corresponding to the data flow source are used as analysis data, and the analysis data is input into the trained priority prediction model to predict the priority of the corresponding data flow and use it as the data flow feature data; the priority prediction model is a deep neural network model; The training process of the priority prediction model includes: Collect a group of analysis data in advance, set corresponding priorities for the a group of analysis data, a is an integer greater than 1, and convert the analysis data and the corresponding priorities into a corresponding set of feature vectors; use each set of feature vectors as the input of a priority prediction model, the priority prediction model uses a set of prediction priorities corresponding to each group of analysis data as output, and uses the actual priority corresponding to each group of analysis data as a prediction target, and the actual priority is the pre-set priority corresponding to the analysis data; minimize the sum of prediction errors of all analysis data as a training target; train the priority prediction model until the sum of prediction errors converges and stops training.

6. A method for efficient wireless communication data transmission according to claim 5, characterized in that: The channel characteristic parameters include bandwidth, frequency and time slot; The step of dynamically allocating channel characteristic parameters comprises: Step S201: Obtain parameter ranges and construct n sets of parameter sets; Step S202: setting different digital labels for n sets of parameter sets and marking them as set labels; constructing a search interval [1, n] according to the set labels corresponding to the n sets of parameter sets; Step S203: Define the population size b and the maximum number of iterations T max ; Step S204: construct an initialization population according to the population size b, and the number of iterations of the initialization population is t=0; randomly generate the position of each individual in the population, and the position of each individual corresponds one-to-one to the set label in the search interval; Step S205: defining an optimization value function; Step S206: Calculate the optimization value of each individual, and take the individual with the largest optimization value as the optimal individual; Step S207: Determine individual behavior according to the optimization value of each individual; Step S208: updating the position of each individual according to the individual behavior and the position of the optimal individual; Step S209: let t=t+1; Step S210: loop through steps S206 to S209 until t ≥ T max When , the loop ends and goes to step S211; Step S211: Acquire the best individual in the population, take the set label corresponding to the position of the best individual as the best label, and dynamically allocate channel characteristic parameters according to the parameter set corresponding to the best label.

7. A wireless communication data efficient transmission method according to claim 6, characterized in that: In step S201, the parameter range includes the bandwidth range, frequency range and time slot range of the optimal transmission path; the method for constructing n sets of parameter sets is: randomly selecting a value from each range in the parameter range to construct a set of parameter sets, and constructing a total of n sets of parameter sets, and the n sets of parameter sets are all different; In step S204, the step of generating the position of each individual includes: Step S301: Preset the number of levels d; Step S302: Evenly divide the search interval according to the number of levels to obtain d sub-search intervals, where the width of each sub-search interval is Step S303: Generate an individual position from each sub-search interval in turn; Step S304: loop step S303 until the positions of b individuals are generated, and the loop ends; In step S303, the method of generating individual positions from the sub-search interval includes: Mark the minimum value of the sub-search interval as the minimum search value, and mark the maximum value of the sub-search interval as the maximum search value; subtract the minimum search value from the maximum search value to obtain the search difference; multiply the search difference by a random number in the interval [0,1], and add the minimum search value to obtain the position of the individual; In step S205, the expression of the optimization value function is f=cx+xz; wherein f is the optimization value, cx is the efficiency standard, and xz is the signal standard; the method for obtaining the efficiency standard and the signal standard includes: The data stream type in the data stream attribute data is replaced with the corresponding type label, and the data stream source is replaced with the corresponding source label; the replaced data stream attribute data is marked as replacement data, and the channel quality of the optimal transmission path is marked as optimal quality; the replacement data, data stream feature data, optimal quality and the corresponding set label of the individual are used as evaluation data; the evaluation data is input into the trained efficiency prediction model to predict the corresponding transmission efficiency; the evaluation data is input into the trained quality prediction model to predict the corresponding signal quality; wherein, the training process of the efficiency prediction model and the quality prediction model are consistent with the training process of the priority prediction model, and both are deep neural network models; the transmission efficiency and signal quality are normalized respectively to obtain the efficiency standard and signal standard.

8. The method for efficient wireless communication data transmission according to claim 7, characterized in that: In step S207, the method for determining individual behavior includes: Add the optimization values ​​of each individual in turn, and then divide by b to obtain the optimization mean; compare the optimization value of each individual with the optimization mean in turn, mark the individuals with optimization values ​​greater than or equal to the optimization mean as excellent individuals, and mark the individuals with optimization values ​​less than the optimization mean as ordinary individuals; the individual behavior corresponding to the excellent individuals is approach behavior, and the behavior corresponding to the ordinary individuals is exploration behavior; In step S208, the method for updating the location of each individual includes: If the individual behavior is a close behavior, the corresponding location update method includes: Preset the step factor and the influence factor; mark the position of the best individual as the best position, and mark the position of the current individual as the current position; subtract the current position from the best position, and then multiply it by the step factor to obtain the moving step; mark all individuals except the best individual and the current individual as the remaining individuals; mark the positions of the remaining individuals as the remaining positions, and subtract the current position from each remaining position in turn to obtain the position difference; add all the position differences in turn, and then multiply them by the influence factor to obtain the influence step; add the moving step to the current position, and then add the influence step to obtain the first updated position, and update the current position according to the first updated position; If the individual behavior is an exploration behavior, the corresponding location update method includes: Preset the avoidance factor; compare the optimized values ​​of all individuals respectively, mark the individual with the smallest optimized value as the worst individual, and mark the position of the worst individual as the worst position; subtract the worst position from the current position, and then multiply it by the avoidance factor to obtain the avoidance step length; add the avoidance step length to the current position to obtain the second updated position, and update the current position according to the second updated position.

9. A wireless communication data efficient transmission system, implementing a wireless communication data efficient transmission method according to any one of claims 1 to 8, characterized in that: include: A path construction module is used to construct m data transmission paths using an improved path selection algorithm; The quality assessment module is used to collect the channel quality data of each data transmission path in real time and perform channel quality assessment; A path screening module is used to screen out the best transmission path from m data transmission paths according to channel quality; A data collection module, used for collecting data stream attribute data; A feature extraction module is used to extract features from data stream attribute data and obtain data stream feature data; The characteristic allocation module is used to adopt the improved optimization algorithm, dynamically allocate channel characteristic parameters based on the channel quality of the best transmission path and the data stream characteristic data, and transmit data according to the channel characteristic parameters.

10. A wireless communication data efficient transmission product, characterized in that: An efficient wireless communication data transmission system integrating claim 9.

Images