MUDR-based network capacity enhancement method, apparatus and device, and medium

Abstract

The invention relates to an MUDR-based network capacity enhancement method, device, equipment and medium, a plurality of user signals received by a receiving antenna in a multi-element antenna array are processed by adopting a minimum mean square error to obtain an independent signal of each user, and a dynamic route selection algorithm is adopted according to a current network state to enhance the network capacity of each user. A preliminary optimal transmission strategy is obtained for each single signal, when network transmission is carried out on the corresponding single signal according to the preliminary optimal transmission strategy, real-time optimization is carried out on the preliminary optimal transmission strategy according to dynamic changes of the network based on a Q learning algorithm, and a multi-path transmission control protocol is adopted to carry out real-time optimization on the preliminary optimal transmission strategy. And single signals are transmitted in the multiple optimized paths in the initial optimal transmission strategy after real-time optimization. By adopting the method, the network capacity can be effectively enhanced.

Description

Technical Field

[0001] This application relates to the field of wireless communication technology, and in particular to a method, apparatus, device and medium for enhancing network capacity based on MUDR. Background Technology

[0002] In modern communication systems, network capacity—the ability of a network to process data within a specific timeframe—is a crucial indicator of its performance. With the surge in mobile device users and the diversification of internet applications, traditional single-user reception technologies are increasingly unable to meet the demands for high data rates and low latency.

[0003] Meanwhile, traditional network technologies employ static or semi-static strategies for data flow routing and resource allocation. This results in insufficient flexibility in responding to dynamically changing communication demands, hindering the maximization of resource utilization efficiency. The network's scalability and flexibility are inadequate, making it difficult to support rapidly growing service types and data volumes. Especially during sudden surges in traffic, existing technologies struggle to quickly adjust resource allocation strategies to adapt to changing demands. Summary of the Invention

[0004] Therefore, it is necessary to provide a method, apparatus, device, and medium for enhancing network capacity based on MUDR that can handle high density and high data rate, in order to address the above-mentioned technical problems.

[0005] A network capacity enhancement method based on MUDR, the method comprising:

[0006] Acquire multi-user signals, which are received by receiving antennas in a multi-element antenna array, wherein each receiving antenna independently captures multiple user signals;

[0007] The multi-user signals are processed using the minimum mean square error method to obtain individual signals for each user;

[0008] A dynamic routing algorithm is used to obtain a preliminary optimal transmission strategy for each individual signal based on the current network status.

[0009] When transmitting the corresponding individual signals over the network according to the preliminary optimal transmission strategy, the preliminary optimal transmission strategy is optimized in real time according to the dynamic changes of the network using a Q-learning algorithm.

[0010] Based on the multipath transmission control protocol, the individual signal is transmitted through multiple optimized paths in the preliminary optimal transmission strategy after real-time optimization.

[0011] In one embodiment, processing the multi-user signals using the minimum mean square error to obtain a separate signal for each user includes:

[0012] The multi-user signal is subjected to automatic gain control and bandpass filtering to obtain a preprocessed multi-user signal.

[0013] Construct a corresponding signal matrix based on the preprocessed multi-user signals, and perform minimum mean square error processing on the signal matrix to obtain the processed signal matrix;

[0014] The processed signal matrix is ​​split to obtain preliminary individual signals for each user.

[0015] The individual signals are then reconstructed after interference cancellation, and the individual signals for each user are obtained through bit error rate evaluation.

[0016] In one embodiment, Dijkstra's algorithm is used when the dynamic routing algorithm obtains the preliminary optimal transmission strategy for each individual signal based on the current network state.

[0017] In one embodiment, when optimizing the preliminary optimal transmission strategy based on the Q-learning algorithm according to dynamic changes in the network:

[0018] The preliminary optimal path in the preliminary optimal transmission strategy is taken as the action, and the real-time dynamic changes in the network are taken as the state.

[0019] Calculate the value function obtained by selecting different actions in the stated state, and optimize the preliminary optimal transmission strategy based on the value function.

[0020] In one embodiment, the value function is calculated using the following formula:

[0021]

[0022] In the above formula, Q(s,a) is the value function for choosing action a in state s, α represents the learning rate, γ represents the discount factor, and r t +1 is the instant reward gained after taking action a.

[0023] In one embodiment, the network's real-time changes in bandwidth, latency, and packet loss rate are used as the state.

[0024] This application also provides a network capacity enhancement device based on MUDR, the device comprising:

[0025] A multi-user signal acquisition module is used to acquire multi-user signals, which are received by receiving antennas in a multi-element antenna array, wherein each receiving antenna independently captures multiple user signals.

[0026] The signal processing module is used to process the multi-user signals using the minimum mean square error to obtain a separate signal for each user.

[0027] The module for obtaining the preliminary optimal transmission strategy is used to obtain the preliminary optimal transmission strategy for each individual signal based on the current network state using a dynamic routing algorithm.

[0028] The Q-learning-based real-time optimization module is used to optimize the preliminary optimal transmission strategy in real time based on network dynamics when transmitting the corresponding individual signals according to the preliminary optimal transmission strategy.

[0029] The multipath transmission module is used to transmit the individual signal in multiple optimized paths in the preliminary optimal transmission strategy after real-time optimization, based on the multipath transmission control protocol.

[0030] A computer device includes a memory and a processor, the memory storing a computer program, and the processor executing the computer program performing the following steps:

[0031] Acquire multi-user signals, which are received by receiving antennas in a multi-element antenna array, wherein each receiving antenna independently captures multiple user signals;

[0032] The multi-user signals are processed using the minimum mean square error method to obtain individual signals for each user;

[0033] A dynamic routing algorithm is used to obtain a preliminary optimal transmission strategy for each individual signal based on the current network status.

[0034] When transmitting the corresponding individual signals over the network according to the preliminary optimal transmission strategy, the preliminary optimal transmission strategy is optimized in real time according to the dynamic changes of the network using a Q-learning algorithm.

[0035] Based on the multipath transmission control protocol, the individual signal is transmitted through multiple optimized paths in the preliminary optimal transmission strategy after real-time optimization.

[0036] A computer-readable storage medium having a computer program stored thereon, the computer program performing the following steps when executed by a processor:

[0037] Acquire multi-user signals, which are received by receiving antennas in a multi-element antenna array, wherein each receiving antenna independently captures multiple user signals;

[0038] The multi-user signals are processed using the minimum mean square error method to obtain individual signals for each user;

[0039] A dynamic routing algorithm is used to obtain a preliminary optimal transmission strategy for each individual signal based on the current network status.

[0040] When transmitting the corresponding individual signals over the network according to the preliminary optimal transmission strategy, the preliminary optimal transmission strategy is optimized in real time according to the dynamic changes of the network using a Q-learning algorithm.

[0041] Based on the multipath transmission control protocol, the individual signal is transmitted through multiple optimized paths in the preliminary optimal transmission strategy after real-time optimization.

[0042] The aforementioned network capacity enhancement method, apparatus, device, and medium based on MUDR process multiple user signals received by the receiving antennas in a multi-element antenna array using minimum mean square error (MME) to obtain a unique signal for each user. A dynamic routing algorithm is then used to derive a preliminary optimal transmission strategy for each unique signal based on the current network state. During network transmission of the corresponding unique signal according to the preliminary optimal transmission strategy, a Q-learning algorithm is employed to optimize the preliminary optimal transmission strategy in real time based on dynamic network changes. Finally, based on a multipath transmission control protocol, the unique signal is transmitted along multiple optimized paths within the real-time optimized preliminary optimal transmission strategy. This method effectively enhances network capacity. Attached Figure Description

[0043] Figure 1 This is a flowchart illustrating a network capacity enhancement method based on MUDR in one embodiment;

[0044] Figure 2 This is a flowchart illustrating multi-user signal processing in one embodiment;

[0045] Figure 3 This is a structural block diagram of a network capacity enhancement device based on MUDR in one embodiment;

[0046] Figure 4 This is an internal structural diagram of a computer device in one embodiment. Detailed Implementation

[0047] To make the objectives, technical solutions, and advantages of this application clearer, the following detailed description is provided in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative and not intended to limit the scope of this application.

[0048] In existing wireless communication technologies, single-user multiple-input multiple-output (SU-MIMO) and multiple-user multiple-input multiple-output (MU-MIMO) have low spectral efficiency in high-density user environments, failing to fully utilize spatial diversity and leading to insufficient spectrum resource utilization and network congestion. These technologies also exhibit significant interference between user equipment in multi-user environments, particularly in signal coverage, resulting in signal attenuation and high packet loss rates in edge users or obstructed areas.

[0049] Furthermore, existing technologies require complex system configurations and high maintenance costs, increasing the economic burden of implementation and limiting their adoption in cost-sensitive applications. Their data flow management and resource allocation methods lack flexibility, making it difficult to quickly adapt to changes in network demands, particularly in response to sudden traffic surges or new service deployments, thus limiting system scalability and long-term development.

[0050] To address the aforementioned issues, this application, such as Figure 1 As shown, a network capacity enhancement method based on MUDR is provided, including the following steps:

[0051] Step S100: Acquire multi-user signals. These multi-user signals are received by receiving antennas in a multi-element antenna array, wherein each receiving antenna independently captures multiple user signals.

[0052] Step S110: The multi-user signals are processed using the minimum mean square error to obtain individual signals for each user.

[0053] Step S120: Using a dynamic routing algorithm, a preliminary optimal transmission strategy is obtained for each individual signal based on the current network status.

[0054] Step S130: When transmitting the corresponding individual signals over the network according to the preliminary optimal transmission strategy, the preliminary optimal transmission strategy is optimized in real time according to the dynamic changes of the network using a Q-learning algorithm.

[0055] Step S140: Based on the multipath transmission control protocol, the individual signal is transmitted in multiple optimized paths in the preliminary optimal transmission strategy after real-time optimization.

[0056] This application addresses the issue of network capacity enhancement by considering two aspects that cause this problem: network congestion and capacity limitations, and inefficient resource classification. It proposes a method using multiple-mode multi-antenna arrays (MIMO) combined with the minimum mean square error (MMSE) algorithm to receive and process multi-user signals, known as Multi-User Diversity Receiver (MUDR), to solve the network congestion and capacity limitations problems. Based on MUDR, a dynamic routing algorithm is first used to obtain a preliminary optimal transmission strategy according to the current network state. During data transmission according to this preliminary optimal strategy, a Q-learning algorithm is used to optimize the strategy in real time based on the dynamically changing network state. Furthermore, data is transmitted through multiple optimized paths using a multipath transmission control protocol. This effectively enhances network capacity and data throughput while reducing complexity and cost, enabling efficient utilization in high-density, high-data-rate communication environments.

[0057] In steps S100 and S110, a multi-electrode multi-array antenna system (MIMO) combined with the minimum mean square error (MMSE) algorithm is used to receive and process multi-user signals. Each receiving antenna independently captures signals from different users, and the MMSE algorithm optimizes the power ratio between signal and noise to reduce received signal errors. This algorithm constructs a joint covariance matrix of the signals and uses linear algebra techniques to calculate the optimal weight vector, thereby weighting and combining the received signals to effectively separate and recover the data streams of each user. This process not only improves the signal reception quality but also significantly reduces interference between users, further enhancing the system's data processing capabilities and the network's data throughput.

[0058] like Figure 2 As shown, the process of processing multi-user signals using minimum mean square error to obtain individual signals for each user includes: performing automatic gain control and bandpass filtering on the multi-user signals to obtain preprocessed multi-user signals; then, constructing corresponding signal matrices based on the preprocessed multi-user signals, and performing minimum mean square error processing on the signal matrices to obtain processed signal matrices; splitting the processed signal matrices to obtain preliminary individual signals for each user; performing interference cancellation on each preliminary individual signal before reconstruction; and obtaining the individual signals for each user through bit error rate evaluation.

[0059] Specifically, taking a simple digital signal transmission as an example, suppose the signals sent by the users are: s1 = 1 + 0i (user 1), s2 = 1 + 1i (user 2), s3 = 0 + 1i (user 3), then the channel matrix H and the received multi-user signal y are expressed as follows:

[0060]

[0061] The received signal y is calculated using the channel matrix and expressed as:

[0062] y = H·s + n

[0063] Here, the noise n is assumed to be slight, for example:

[0064]

[0065] The calculation yielded:

[0066]

[0067] Next, MMSE signal estimation is performed: the signal estimate is calculated using the MMSE formula. Assuming noise power

[0068]

[0069] The above calculation example accurately demonstrates how the MMSE algorithm processes the actual received signal and recovers the user's original signal. It illustrates the effectiveness and calculation methods of signal processing techniques in practical applications.

[0070] Next, in steps S120 to S140, the transmission strategy employed when transmitting individual signals sent by each user is further explained. In this embodiment, the signal transmission quality is determined by evaluating metrics such as signal-to-noise ratio (SNR) and bit error rate (BER) of each signal path in real time. The transmission efficiency of each path is calculated using a path loss model and real-time network state information. Using a Q-learning-based reinforcement learning algorithm, the system can dynamically select the optimal signal path. This algorithm continuously updates its value function, learns and adapts to environmental changes, and optimizes a path selection strategy that maximizes long-term gains.

[0071] In step S120, a cost-based routing decision method is used in the dynamic routing algorithm to obtain a preliminary optimal transmission strategy. Specifically, Dijkstra's algorithm, a shortest path algorithm from graph theory, is employed. This algorithm selects the node with the smallest "shortest distance" from the unprocessed nodes each time and processes all of that node's neighbors. Mathematical representation:

[0072] D(v) = min(D(v), D(u) + c(u, v))

[0073] In the above formula, D(v) represents the shortest distance from the source node to node v, and c(u,v) is the direct distance from node u to v.

[0074] The dynamic routing algorithm obtains a preliminary optimal transmission strategy based on the current network state, which includes multiple transmission paths adapted to the current network state.

[0075] Given that network conditions are constantly changing, the initially optimal transmission strategy obtained may not be the optimal path during data transmission due to changes in network conditions. For example, in a wireless communication network in an urban environment, the network condition may change rapidly due to obstructions from buildings, interference from electronic devices, or fluctuations in the number of users.

[0076] Therefore, in step S130, the preliminary optimal transmission strategy is optimized based on the Q-learning algorithm according to the dynamic changes of the network. That is, the Q-learning algorithm is used to optimize each path in the preliminary optimal transmission strategy in real time according to the network status during the signal transmission process. For example, in a certain path, a routing node whose performance has degraded due to congestion is bypassed, thereby maintaining the communication quality and speed of that path.

[0077] Specifically, Q-learning is a model-free reinforcement learning algorithm used to learn the value function of actions in the absence of an environmental model. In communication networks, Q-learning is mainly used to dynamically select the optimal data transmission path to adapt to changes in network conditions and improve overall data transmission efficiency and reliability.

[0078] Furthermore, when optimizing the preliminary optimal transmission strategy based on the Q-learning algorithm according to the dynamic changes of the network: the preliminary optimal path in the preliminary optimal transmission strategy is taken as the action, the real-time changes of the network are taken as the state, the value function obtained by choosing different actions in the state is calculated, and the preliminary optimal transmission strategy is optimized according to the value function.

[0079] Specifically, the value function is calculated using the following formula:

[0080]

[0081] In the above formula, Q(s,a) is the value function for choosing action a in state s, α represents the learning rate, γ represents the discount factor, and r t +1 is the immediate reward gained after taking action 'a'. t and s t+1 These represent the current state and the next state, respectively.

[0082] In this embodiment, the real-time changes in network bandwidth, latency, and packet loss rate are used as the status.

[0083] Furthermore, the reward function is calculated using the following formula:

[0084] r t+1=α×Improvement in Speed-β×Increase in Delay-γ×Increase inPacket Loss Rate

[0085] In the above formula, "Improvement in Speed" represents the speed improvement compared to the previous state. "Increase in Delay" represents the delay increase compared to the previous state, quantified using (new delay - 1-day delay). "Increase in Packet Loss Rate" represents the packet loss rate increase compared to the previous state, quantified using (new packet loss rate - old packet loss rate). α, β, and γ are the corresponding weighting factors, representing the importance of different metrics.

[0086] Employing the Q-learning algorithm to optimize data transmission paths based on network state changes can further reduce signal interference: by selecting the optimal signal transmission path, interference between different users is reduced, especially in multi-user environments; data throughput is enhanced: by optimizing signal transmission paths, data transmission rates and network capacity are increased, particularly in areas with complex network conditions or high user density; and network stability is improved: by dynamically adapting to network changes, such as channel quality variations or user movement, the stability and reliability of network connections are maintained.

[0087] Furthermore, in step S140, during signal transmission, the Multipath Transmission Control Protocol (MPTCP) is also incorporated. A Q-learning-based path optimization method dynamically selects the optimal data transmission path for MPTCP. In this combination, MPTCP not only utilizes multiple paths to increase transmission redundancy and reliability, but each path selection is also learned and optimized in real time, adjusted based on immediate feedback from network conditions. Simultaneously, the paths dynamically evaluated and optimized based on the Q-learning algorithm can be directly applied to MPTCP's path management. Thus, MPTCP can maintain its multipath advantages while achieving higher data transmission efficiency and lower latency by selecting the optimal path combination.

[0088] Data flow management and intelligent resource allocation:

[0089] Introducing Q-learning-based optimization algorithms into DFMT enables more precise resource allocation strategies. For example, when dynamically adjusting bandwidth allocation and load balancing strategies in response to changes in real-time network load and user demand, the first step is to deploy monitoring tools in the network to collect key performance indicators in real time, such as traffic load, bandwidth usage, latency, and user data demand. SNMP can be used to obtain traffic statistics from routers and switches, while tools like NetFlow or sFlow can be used to collect fine-grained traffic samples. Next, state definitions and updates are performed. The state vector should include various indicators that express the current network status, such as: state = [total_bandwidth_usage, average_delay, peak_load, user_demand_change]. These state indicators need to be updated periodically or based on events, such as whenever new traffic data becomes available, and the value function and reward function are dynamically calculated. Based on the real-time calculated reward and the current state vector, the Q-value is calculated. Based on the data in the Q-table, actions that maximize the expected reward are selected to adjust the bandwidth allocation and load balancing strategies.

[0090] Q-learning-based optimization algorithms can predict and respond to changes in network conditions, providing decision support for DFMT. This enables data flow management to not only respond to the current situation but also predict future changes in demand, thereby optimizing the overall utilization of network resources.

[0091] In this embodiment, the initial handshake between the client and server at the start of the MPTCP session extends the traditional TCP three-way handshake process by adding MPTCP-compatible options to negotiate the use of MPTCP. For example:

[0092] Client→Server:SYN,MP C APABLE

[0093] Server→Client:SYN-ACK,MP C APABLE

[0094] Client→Server:ACK

[0095] Furthermore, additional MPTCP options, such as 'aDD_ADDR', are used to notify the other party of new available paths (IP addresses and ports). These options are dynamically exchanged during data transmission to add or delete paths, for example:

[0096] Client→Server:ADD A DDR, AddrID, IP A ddr,Port

[0097] MPTCP uses comprehensive metrics such as Round Trip Time (RTT), bandwidth, and historical packet loss rate to evaluate the quality of each path. The path scoring formula is as follows:

[0098]

[0099] In the above formula, ω1, ω2, and ω3 represent weight parameters used to adjust the influence of each indicator.

[0100] The optimal path for data transmission is dynamically selected based on path scores. The path selection algorithm takes into account the current network conditions and automatically adjusts to ensure even distribution of data load and maximize total throughput.

[0101] MPTCP divides the data stream into multiple sub-streams and dynamically allocates data traffic based on the current performance of each path:

[0102]

[0103] In the above formula, F i Total_Data is the amount of data allocated to path i, and Total_Data is the total amount of data.

[0104] Each sub-stream can adjust its window size based on the performance of the path to adapt to changing network conditions, thereby optimizing transmission efficiency and reducing latency.

[0105] When path performance degrades, the system will reduce the data load on that path, or even temporarily suspend data transmission on that path until performance recovers.

[0106] In the aforementioned MUDR-based network capacity enhancement method, multi-user diversity reception technology is first used for receiving and processing multi-user signals. During data transmission, a dynamic routing algorithm first provides initial path decisions based on the current network conditions. Subsequently, a Q-learning-based reinforcement learning algorithm dynamically adjusts and optimizes these decisions, ensuring that path selection adapts to changes in network conditions. Finally, MPTCP technology performs actual data transmission on these optimized paths, leveraging the advantages of multi-path transmission to enhance the reliability and efficiency of data transmission. The entire process demonstrates a complete link from routing decision to data transmission, with each technical step contributing significantly to the optimization of network communication.

[0107] It should be understood that, although Figure 1The steps in the flowchart are shown sequentially as indicated by the arrows, but these steps are not necessarily executed in the order indicated by the arrows. Unless otherwise specified herein, there is no strict order in which these steps are executed, and they can be performed in other orders. Figure 1 At least some of the steps in the process may include multiple sub-steps or multiple stages. These sub-steps or stages are not necessarily completed at the same time, but can be executed at different times. The execution order of these sub-steps or stages is not necessarily sequential, but can be executed in turn or alternately with other steps or at least some of the sub-steps or stages of other steps.

[0108] In one embodiment, such as Figure 3 As shown, a network capacity enhancement device based on MUDR is provided, including: module A, module B, and module C, wherein:

[0109] The multi-user signal acquisition module 200 is used to acquire multi-user signals, which are received by receiving antennas in a multi-element antenna array, wherein each receiving antenna independently captures multiple user signals.

[0110] Signal processing module 210 is used to process the multi-user signals using minimum mean square error to obtain individual signals for each user;

[0111] The module 220 for obtaining the preliminary optimal transmission strategy is used to obtain the preliminary optimal transmission strategy for each individual signal based on the current network state using a dynamic routing algorithm.

[0112] The Q-learning-based real-time optimization module 230 is used to optimize the preliminary optimal transmission strategy in real time according to the dynamic changes of the network when transmitting the corresponding individual signals according to the preliminary optimal transmission strategy.

[0113] The multipath transmission module 240 is used to transmit the individual signal in multiple optimized paths in the preliminary optimal transmission strategy after real-time optimization, based on the multipath transmission control protocol.

[0114] Specific limitations regarding the MUDR-based network capacity enhancement device can be found in the limitations of the MUDR-based network capacity enhancement method described above, and will not be repeated here. Each module in the aforementioned MUDR-based network capacity enhancement device can be implemented entirely or partially through software, hardware, or a combination thereof. These modules can be embedded in or independent of the processor in a computer device in hardware form, or stored in the memory of a computer device in software form, so that the processor can call and execute the operations corresponding to each module.

[0115] In one embodiment, a computer device is provided, which may be a terminal, and its internal structure diagram may be as follows: Figure 4 As shown, the computer device includes a processor, memory, network interface, display screen, and input devices connected via a system bus. The processor provides computing and control capabilities. The memory includes non-volatile storage media and internal memory. The non-volatile storage media stores the operating system and computer programs. The internal memory provides an environment for the operation of the operating system and computer programs stored in the non-volatile storage media. The network interface is used to communicate with external terminals via a network connection. When the computer program is executed by the processor, it implements a network capacity enhancement method based on MUDR. The display screen can be an LCD screen or an e-ink display screen. The input devices can be a touch layer covering the display screen, buttons, a trackball, or a touchpad mounted on the computer device casing, or an external keyboard, touchpad, or mouse.

[0116] Those skilled in the art will understand that Figure 4 The structure shown is merely a block diagram of a portion of the structure related to the present application and does not constitute a limitation on the computer device to which the present application is applied. Specific computer devices may include more or fewer components than those shown in the figure, or combine certain components, or have different component arrangements.

[0117] In one embodiment, a computer device is provided, including a memory and a processor, wherein the memory stores a computer program, and the processor executes the computer program to perform the following steps:

[0118] Acquire multi-user signals, which are received by receiving antennas in a multi-element antenna array, wherein each receiving antenna independently captures multiple user signals;

[0119] The multi-user signals are processed using the minimum mean square error method to obtain individual signals for each user;

[0120] A dynamic routing algorithm is used to obtain a preliminary optimal transmission strategy for each individual signal based on the current network status.

[0121] When transmitting the corresponding individual signals over the network according to the preliminary optimal transmission strategy, the preliminary optimal transmission strategy is optimized in real time according to the dynamic changes of the network using a Q-learning algorithm.

[0122] Based on the multipath transmission control protocol, the individual signal is transmitted through multiple optimized paths in the preliminary optimal transmission strategy after real-time optimization.

[0123] In one embodiment, a computer-readable storage medium is provided having a computer program stored thereon, the computer program performing the following steps when executed by a processor:

[0124] Acquire multi-user signals, which are received by receiving antennas in a multi-element antenna array, wherein each receiving antenna independently captures multiple user signals;

[0125] The multi-user signals are processed using the minimum mean square error method to obtain individual signals for each user;

[0126] A dynamic routing algorithm is used to obtain a preliminary optimal transmission strategy for each individual signal based on the current network status.

[0127] When transmitting the corresponding individual signals over the network according to the preliminary optimal transmission strategy, the preliminary optimal transmission strategy is optimized in real time according to the dynamic changes of the network using a Q-learning algorithm.

[0128] Based on the multipath transmission control protocol, the individual signal is transmitted through multiple optimized paths in the preliminary optimal transmission strategy after real-time optimization.

[0129] Those skilled in the art will understand that all or part of the processes in the methods of the above embodiments can be implemented by a computer program instructing related hardware. The computer program can be stored in a non-volatile computer-readable storage medium, and when executed, it can include the processes of the embodiments of the above methods. Any references to memory, storage, databases, or other media used in the embodiments provided in this application can include non-volatile and / or volatile memory. Non-volatile memory can include read-only memory (ROM), programmable ROM (PROM), electrically programmable ROM (EPROM), electrically erasable programmable ROM (EEPROM), or flash memory. Volatile memory can include random access memory (RAM) or external cache memory. By way of illustration and not limitation, RAM is available in various forms, such as static RAM (SRAM), dynamic RAM (DRAM), synchronous DRAM (SDRAM), dual data rate SDRAM (DDRSDRAM), enhanced SDRAM (ESDRAM), synchronous link DRAM (SLDRAM), Rambus direct RAM (RDRAM), direct memory bus dynamic RAM (DRDRAM), and memory bus dynamic RAM (RDRAM), etc.

[0130] The technical features of the above embodiments can be combined in any way. For the sake of brevity, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this specification.

[0131] The above-described embodiments merely represent several implementation methods of the present application. While the descriptions are relatively specific and detailed, they should not be construed as limiting the scope of the present invention. It should be noted that a person skilled in the art could make various modifications and improvements without departing from the spirit of the present application, all of which fall within the scope of protection of the present application. Therefore, the scope of protection of the present patent application shall be determined by the appended claims.

Claims

1. A network capacity enhancement method based on MUDR, characterized in that: The method comprises: Acquire a multi-user signal, wherein the multi-user signal is received by a receiving antenna in a multi-element antenna array, wherein each receiving antenna independently captures a plurality of user signals; Processing the multi-user signal by using minimum mean square error to obtain a separate signal for each user; A dynamic routing algorithm is used to obtain the initial optimal transmission strategy for each individual signal according to the current network status; When the corresponding individual signal is transmitted over the network according to the preliminary optimal transmission strategy, the preliminary optimal transmission strategy is optimized in real time according to the dynamic changes of the network using a Q-learning algorithm; Based on a multi-path transmission control protocol, the individual signal is transmitted in a plurality of optimized paths in a preliminary optimal transmission strategy after real-time optimization.

2. The network capacity enhancement method according to claim 1, characterized in that: The adopting minimum mean square error to process the multi-user signal to obtain a separate signal for each user comprises: Performing automatic gain control and bandpass filtering on the multi-user signal to obtain a pre-processed multi-user signal; Constructing a corresponding signal matrix according to the preprocessed multi-user signal, and performing minimum mean square error processing on the signal matrix to obtain a processed signal matrix; Performing signal splitting on the processed signal matrix to obtain preliminary individual signals of each user; The preliminary individual signals are reconstructed after interference elimination, and the individual signals corresponding to the individual users are obtained by bit error rate evaluation.

3. The network capacity enhancement method according to claim 2, characterized in that: When using a dynamic routing selection algorithm to obtain a preliminary optimal transmission strategy for each of the individual signals according to the current network status, a Dijkstra algorithm is used.

4. The network capacity enhancement method according to claim 3, characterized in that: When the preliminary optimal transmission strategy is optimized according to the dynamic changes of the network based on the Q learning algorithm: Taking the preliminary optimal path in the preliminary optimal transmission strategy as an action and taking the real-time dynamic changes of the network as a state; A value function obtained by selecting different actions under the state is calculated, and the preliminary optimal transmission strategy is optimized according to the value function.

5. The network capacity enhancement method according to claim 4, characterized in that: The value function is calculated using the following formula: Q(s t ,a t )←Q(s t ,a t )+α[r t+1 +γm a axQ(s t+1 ,a)-Q(s t ,a t )] In the above formula, Q(s,a) is the value function of selecting action a in state s, α represents the learning rate, γ represents the discount factor, and r t +1 is the immediate reward you get for taking action a.

6. The network capacity enhancement method according to claim 5, characterized in that: The real-time changing bandwidth, delay and packet loss rate of the network are used as the status.

7. A network capacity enhancement device based on MUDR, characterized in that: The device comprises: A multi-user signal acquisition module, used to acquire multi-user signals, wherein the multi-user signals are received by receiving antennas in a multi-element antenna array, wherein each receiving antenna independently captures multiple user signals; A signal processing module, used to process the multi-user signal by using minimum mean square error to obtain a separate signal for each user; A preliminary optimal transmission strategy obtaining module is used to obtain a preliminary optimal transmission strategy for each individual signal according to the current network status by using a dynamic routing selection algorithm; A Q-learning-based real-time optimization module is used to optimize the preliminary optimal transmission strategy in real time according to the dynamic changes of the network by using a Q-learning algorithm when the corresponding individual signal is transmitted over the network according to the preliminary optimal transmission strategy; The multi-path transmission module is used to transmit the individual signal in multiple optimized paths in the preliminary optimal transmission strategy after real-time optimization based on the multi-path transmission control protocol.

8. A computer device comprising a memory and a processor, wherein the memory stores a computer program, wherein: When the processor executes the computer program, the steps of the method according to any one of claims 1 to 6 are implemented.

9. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, the steps of the method according to any one of claims 1 to 6 are implemented.

Images