Intelligent signal optimization method of 5G router

Abstract

The embodiment of the invention provides an intelligent signal optimization method for a 5G router, and the method comprises the steps: determining a current channel load state according to a real-time user connection number statistical result; evaluating a network congestion level according to the channel load state; adjusting a channel bandwidth allocation strategy based on the network congestion level; the adjusted bandwidth allocation policy is applied to the router to optimize signal transmission. According to the scheme provided by the embodiment of the invention, the channel bandwidth can be regulated and controlled according to the real-time user connection number so as to solve the problem of network congestion.

Description

Technical Field

[0001] This application relates to the field of communication technologies, and specifically to an intelligent signal optimization method for a 5G router. Background Art

[0002] The intelligent signal optimization of 5G routers can dynamically adjust signal parameters in the 5G network to improve connection stability and efficiency, reduce interference, and optimize the user experience by combining artificial intelligence algorithms and real-time data monitoring. However, this method currently faces a problem, that is, how to flexibly adjust the channel bandwidth according to the real-time number of user connections to effectively solve the network congestion caused by a sudden increase in the number of users. This requires further improvement of the adjustment mechanism to achieve more accurate resource allocation and performance improvement. Summary of the Invention

[0003] In view of this, embodiments of the present disclosure provide an intelligent signal optimization method for a 5G router, which at least partially solves the problems existing in the prior art.

[0004] An intelligent signal optimization method for a 5G router includes:

[0005] Determine the current channel load status according to the statistical result of the real-time number of user connections;

[0006] Evaluate the network congestion level according to the channel load status;

[0007] Adjust the channel bandwidth allocation strategy based on the network congestion level;

[0008] Apply the adjusted bandwidth allocation strategy to the router to optimize signal transmission.

[0009] Preferably, adjusting the channel bandwidth allocation strategy based on the network congestion level includes the following steps:

[0010] Determine that the real-time number of user connections is N;

[0011] Calculate the current network bandwidth B, and set the maximum bandwidth as Bmax;

[0012] Dynamically adjust the bandwidth W per user according to the formula W = Bmax * exp(N / λ), where λ is a predefined load parameter;

[0013] When W is less than the minimum guaranteed bandwidth Wmin, set W = Wmin.

[0014] Preferably, it further includes the following steps:

[0015] Monitor the user-requested data traffic F, and record the average rate R = the amount of data within the time window of F / T;

[0016] Evaluate the current channel load status and classify it according to the following rules: If R > μB, the level is severe congestion; otherwise, it is light congestion, where μ is the system bandwidth utilization efficiency coefficient;

[0017] Input the classification result into the adjustment algorithm and dynamically update the number of channels K based on the channel congestion level g, K = K_base + ΔK * (1 - g), where ΔK is the increment value;

[0018] Set an upper limit of max(K × W) for the total bandwidth of all users and ensure that it does not exceed B.

[0019] Preferably, further adjusting the bandwidth allocation based on the number of channels K includes the following steps:

[0020] Recount the number of active users Na, excluding offline users whose idle time exceeds Ta;

[0021] Refine the allocation of the bandwidth Wi for each user using a new formula: Wi = B / ((Na - Ni) + δ), where Ni is the base number of users already allocated on the current channel and δ is a small positive value;

[0022] Assign a weighting factor ωi > 1 to high-priority users, and ωi = 1 for others;

[0023] Check whether Wi is evenly divided without overflow or being too low.

[0024] Preferably, further stipulating the user priority determination step includes the following content:

[0025] Obtain the Pi value sequence according to the QoS request category classification, reflecting the service importance;

[0026] If Pi >= η * Pavg is satisfied, enter the high-level queue, where Pavg represents the global average value;

[0027] Apply the bandwidth reduction coefficient α = β^((Fi - Fbar) / γF) to non-priority group users, where β is the attenuation rate and γF is the smoothing parameter;

[0028] Verify whether there is a fairness problem after reduction, that is, compare whether the ratio difference in each sub-band is less than ε, a pre-set small threshold.

[0029] Preferably, adding constraint restrictions to the bandwidth reduction conditions includes the following steps:

[0030] Introduce a congestion prediction index C, and through machine learning fitting C = f(B, F, Na), which is used to estimate in advance the possible peak value in the future time period;

[0031] When the prediction index indicates approaching the critical point pcr, i.e., |Cpcr| < σc, an emergency speed reduction measure is triggered; at this time, the limit value per user Ui = C × Vb * (1 - exp(-ηt)), where Vb and ηt are partial parameters of the speed adjustment function;

[0032] Continuously monitor the data packet loss ratio D% during the execution of this mode. If it is found that the allowable loss ratio Δd% is exceeded twice consecutively, switch to the stable mode;

[0033] Output the finally determined bandwidth amount to adapt to the current demand.

[0034] Preferably, it also includes:

[0035] Audit the allocation of each channel Uj according to the actual channel usage efficiency Ue, and adjust the weight wj of the jth channel through the following formula: wj = Ue_j / max_Ueff ^ (χj), where max_Ueff is the total optimal efficiency value and χ is the sensitivity factor;

[0036] Accumulate all the unfilled bandwidth shares ΔBi of the previous cycle and add them to the new round of basic configuration;

[0037] When multiple groups of requests compete for the same additional unit, the voting ranking method is enabled to select the winning group Gi, where Gi = SUM(wkj) / (|Lkj| ^ ρ), and ρ controls the distribution smoothing effect;

[0038] Verify that the conflict probability index CP < ζ after all new layouts. If the determination is successful, fix the allocation.

[0039] Preferably, it also includes:

[0040] When it is detected that the sudden interruption risk signal Sr is greater than the warning threshold Tr, i.e., Sr / Tr >= δr%, start the fast switching to the standby channel group Sc;

[0041] Quickly locate and match the alternative channel position number Pos_m through the pre-stored mapping table, so that the distance error from the original path Delta_Distance(Pos_m old_p) < ψ;

[0042] At the same time, adjust all the node routing parameters Zij involved, where Zij = λ × exp[γ * distance(m, n) * time], γ reflects the delay factor, and λ refers to the weight multiplication coefficient.

[0043] Preferably, it also includes:

[0044] Statistically count the identity identification code id_s of the original owners to whom the temporarily occupied resources belong, and count their respective remaining amounts to be transmitted L_remaining;

[0045] Construct a stepwise concession progress planning diagram SPM(id_s): Give appropriate refund compensation bandwidth Wr = id_s × θ(L_remaining / B_total) to each affected customer one by one;

[0046] Calculate the difference ΔT by comparing the overall throughput T_curr in the current state with the initial expected level T_target, and use the formula adjust_factor = log(T_curr + λ) to guide fine-tuning.

[0047] The embodiment of the present disclosure provides an intelligent signal optimization method for a 5G router, including: determining the current channel load status according to the statistical result of the real-time user connection number; evaluating the network congestion level according to the channel load status; adjusting the channel bandwidth allocation strategy based on the network congestion level; applying the adjusted bandwidth allocation strategy to the router to optimize signal transmission. Through the solution of the embodiment of the present disclosure, the channel bandwidth can be regulated according to the real-time user connection number to solve the problem of network congestion. BRIEF DESCRIPTION OF THE DRAWINGS

[0048] In the drawings, unless otherwise specified, the same reference numerals throughout the several views denote the same or similar components or elements. These drawings are not necessarily drawn to scale. It should be understood that these drawings only depict some embodiments disclosed in accordance with the present application and should not be regarded as limiting the scope of the present application.

[0049] Figure 1 is a flowchart of an intelligent signal optimization method for a 5G router;

[0050] Figure 2 is a flowchart of adjusting the channel bandwidth allocation strategy based on the network congestion level, including the following steps:

[0051] Figure 3 is a flowchart of further including the following steps:

[0052] Figure 4 is a flowchart of further adjusting the bandwidth allocation based on the number of channels K, including the following steps:

[0053] Figure 5 is a flowchart of further specifying the user priority determination step, including the following content:

[0054] Figure 6 is a flowchart of adding constraint limitations to the bandwidth reduction conditions for, including the following steps:

[0055] Figure 7 is a flowchart of also;

[0056] Figure 8 is a flowchart of also;

[0057] Figure 9 It is also a flowchart. Specific implementation manners

[0058] In the following, only some exemplary embodiments are simply described. As those skilled in the art can recognize, the described embodiments can be modified in various different ways without departing from the spirit or scope of the present application. Therefore, the drawings and the description are considered to be exemplary in nature rather than restrictive.

[0059] Next, with reference to the drawings, an intelligent signal optimization method for a 5G router of the present invention is described. This method includes multiple key steps to achieve dynamic optimization of the channel bandwidth allocation strategy in the 5G network. These steps are aimed at improving network performance, solving network congestion problems that may occur in real-time connections, and providing users with a high-quality experience. The following specifically elaborates on the specific operations of the method and their functions.

[0060] S101: Determine the current channel load status according to the statistical result of the real-time user connection number. In the first step, it is necessary to statistically analyze the real-time connection number of users. In this stage, various data from connected users, such as device type, current rate, traffic request volume, etc., are collected through sensing devices deployed in the network. During this process, the system summarizes the historical records over a period of time and converts them into a form that is easy to analyze. For example, in one embodiment, the number of connected devices can be obtained every few seconds through a periodic data capture mechanism and stored in a memory or a distributed database in the cloud for subsequent processing. The purpose of this data sampling is to ensure that the channel state can be accurately reflected and to avoid the impact of errors caused by a single instantaneous sample on the accuracy of the entire system's decision-making.

[0061] S102: Evaluate the network congestion level according to the channel load status. Based on the previously collected user connection data, determine the current channel load status. This step requires comparing and analyzing all the relevant information collected previously with preset criteria to obtain a preliminary judgment on whether the channel is in a light load, balanced load, or high load situation. Specifically, different channel state level classification models can be defined according to the difference ratio between the maximum simultaneous connection number that each channel can support and its existing connection number, or based on the calculated result of the actual traffic load rate of the channel. If it is detected that a specific channel bears more than the set threshold, it can be considered that there is a potential network congestion hazard here, and immediate intervention measures need to be taken to alleviate it. For example, in some cases, if the average load of a single channel continuously exceeds 90% for more than three minutes, it is determined that this channel has entered a highly tense state.

[0062] S103: Based on the network congestion level, adjust the channel bandwidth allocation strategy. The third step is to further carry out the network congestion level assessment work based on the channel load status obtained previously. This process will combine multiple indicators to comprehensively consider the global network quality performance. In addition to simply counting the workload level of each independent path, it is also necessary to take into account the overall topology architecture design characteristics and other factors before a comprehensive balance can be given to give a final and reasonable evaluation conclusion. In other words, it is necessary not only to understand whether there are serious pressure problems in the local area, but also to see whether there is a coordinated deterioration in the operating status of other adjacent associated units; and to set up differentiated evaluation systems according to different types of business needs (such as broadcast streaming media transmission and download-type HTTP GET requests). For example, video call services may be more sensitive to slight delay changes, while web browsing may be able to tolerate a certain degree of ductility changes. Therefore, in actual application, the weight coefficient should be flexibly adjusted according to the actual situation to obtain an effective score value that is closer to the actual situation.

[0063] S104: Apply the adjusted bandwidth allocation strategy to the router to optimize signal transmission. Based on all the above-mentioned evaluation criteria, officially start to re-set a new target planning diagram presentation scheme that is adapted to the requirements of the current environmental conditions and is oriented to the optimal efficiency. In other words, formulate a corresponding appropriate adjustment strategy based on the network congestion level index that has been clearly recognized, and then send the updated configuration plan after the modification to the corresponding hardware device entity to implement it, and finally achieve the expected ideal improvement goal. For example, assuming that after all the previous analysis programs have been calculated, it is concluded that the current situation of multiple users competing fiercely for scarce resources in some hot spots of popular communities is very prominent, then the dynamic spectrum division and reuse algorithm can be activated to automatically reduce the share of other low-priority tasks and free up more available capacity for emergency priority channels, thereby reducing interference factors and improving overall operation fluency and stability.

[0064] Next, refer to Figure 2 , describing the channel bandwidth allocation strategy based on network congestion level adjustment of the present invention, including the following steps: S201, determining the number of real-time user connections as N; S202, calculating the current network bandwidth B, and setting the maximum bandwidth as Bmax; S203, dynamically adjusting the bandwidth W of each user according to the formula W=Bmax*exp(N / λ), where λ is a predefined load parameter; S204, when W is less than the minimum guaranteed bandwidth Wmin, setting W=Wmin.

[0065] In the above process, first, the number of users N accessing the current router is detected through statistical methods or sensors. This step is the basic data collection part for understanding the current network load situation. N may be an instantaneously changing value, but it is usually updated at fixed intervals to ensure stability and accuracy. Next, the current available network bandwidth B needs to be measured, and a maximum bandwidth Bmax is preset. This is a numerical range determined by the physical performance and configuration of the system, used to evaluate the total bandwidth upper limit. Then, the formula W = Bmax * exp(N / λ) is introduced. In this formula, N is the number of real-time user connections, Bmax is the theoretical maximum network bandwidth, exp is the natural exponential function, indicating an exponential decreasing change law as the number of user connections increases, and the load parameter λ is an adjustable factor that can be optimized. Its value range is generally an integer between 5 and 100, and the optimal value depends on the results of different usage environments and test data. The setting of this formula reflects the demand logic that bandwidth allocation is sensitive to the actual number of connections, avoiding resource waste or uneven allocation problems in extreme cases. If the dynamically calculated W is less than the pre-determined minimum guaranteed bandwidth Wmin, the bandwidth is forcibly adjusted to the Wmin level to ensure the user experience, reflecting the importance of the bottom line principle.

[0066] For example, in one embodiment, a 5G router is connected to 30 devices (i.e., N = 30), the total bandwidth B is set at 800 Mbps, and it is assumed that Bmax is 1000 Mbps, and λ is experimentally selected as 50. After calculation using the formula, W is approximately equal to 296 Mbps (the specific number may vary slightly according to the rounding rule). If the minimum guaranteed bandwidth is required to be Wmin = 150 Mbps, then no further adjustment is needed for the result. However, if W is less than Wmin, a correction mechanism will immediately be adopted to increase the bandwidth to 150 Mbps. Specifically, deploying such an algorithm in some high-density scenarios such as conference rooms or large shopping mall environments is particularly beneficial because in such environments, it is often necessary to balance the competition between various types of terminals and service demands.

[0067] Next, refer to Figure 3 , to describe the process steps of another embodiment of the present invention, including:

[0068] S301: Monitor the user request data traffic F and record the average rate R = the data volume within the time window of F / T;

[0069] S302: Evaluate the current channel load status and classify the level according to the following rules: If R > μB, the level is severe congestion, otherwise it is mild congestion, where μ is the system bandwidth utilization efficiency coefficient;

[0070] S303: Input the classification result into the adjustment algorithm, and dynamically update the number of channels K based on the channel congestion level g, where K = K_base + ΔK * (1 - g), and ΔK is the incremental value;

[0071] S304: Set an upper limit of max(K × W) for the total bandwidth of all users and ensure it does not exceed B.

[0072] First, monitor the data traffic requested by users, represented by the parameter F, and record the data volume within a fixed time window T. Calculate the average rate R based on this information, defined as R = F / T. This process aims to obtain the real-time change in user traffic on the channel, facilitating the subsequent steps to judge the load status of the current channel. Generally, the range of T can be between 1 second and 30 seconds to adapt to the requirements of different application scenarios, and its optimal value depends on the router performance and traffic fluctuation characteristics. Through this formula, the data volume flowing through the channel per unit time can be quantified.

[0073] Then, evaluate the load status of the current channel based on the average rate in the previous step, and use the system bandwidth utilization efficiency coefficient μ and the bandwidth B to judge whether the channel is congested. If the rate R exceeds μB, then mark the channel level as severely congested; otherwise, mark it as lightly congested. For example, in an actual scenario, assume the system bandwidth utilization efficiency coefficient μ = 0.85, and the total bandwidth supported by the system B = 10 Gbps. If the observed R is greater than or equal to 8.5 Gbps at a specific moment, the system will determine it to be in a severely congested state.

[0074] The next step is to input the channel level into the dynamic adjustment algorithm to dynamically adjust the number of available channels K. The specific implementation is by adjusting the contribution of the incremental value ΔK. Set the initial number of channels as K_base, and the dynamically adjusted K value follows the following rule: K = K_base + ΔK * (1 - g). In this process, g refers to the mapping value of the channel congestion level determined from the previous step. For example, severe congestion maps to a higher g value and reduces the number of available channels. For example, if ΔK takes the optimal empirical value of 3 and g is set to 0.4 in the case of light congestion, it can be obtained that the number of channels increases to nearly 2 times the original base to optimize the network environment configuration.

[0075] Finally, set the total bandwidth limit for all users as max(K × W), ensuring it does not exceed the system maximum bandwidth B. Here, W represents the standard bandwidth size provided by a single channel, and determine the final allocation scheme based on the previous steps. For example, in an actual scenario, if K is finally determined to be 8, and each channel provides a standard bandwidth of 1 Gbps, then check the total calculated value (such as 8 × 1 Gbps) to ensure it does not exceed the preset total bandwidth boundary limit of 10 Gbps.

[0076] In one embodiment, consider a 5G router that mainly serves high-speed network download applications. If multiple people in the user group are performing large file transfer tasks, a high F / T calculation result will be detected. After the system automatically determines it as a heavily congested state, the number of available channels is reduced to save the overall bandwidth and prevent some users from preempting resources and causing the overall network to collapse. Specifically, after reducing the number of available channels to a lower value such as 6, it is still possible to reasonably arrange the priorities of each service to ensure the basic communication experience without wasting the hardware resource capacity.

[0077] Next, refer to Figure 4 , to describe the process of adjusting bandwidth allocation based on the number of channels K of the present invention, including:

[0078] S401: Re-statistics the number of active users Na, excluding offline users whose idle time exceeds Ta. The first step is to re-statistics the number of active users, specifically excluding offline users whose idle time exceeds the threshold Ta and obtaining the accurate number of active users Na. The determination criterion for active users lies in whether they are in a long-term idle state. The purpose of this operation is to dynamically reflect the real load in the system and lay a foundation for subsequent resource allocation. For example, in one embodiment, Ta may be set to 30 seconds. If a device does not upload or download data within this time, the device is considered idle.

[0079] S402: Refined allocation of the bandwidth Wi for each user using a new formula: Wi = B / ((Na * Ni)+δ), where Ni is the base number of users already allocated on the current channel, and δ prevents division by zero. The second step is to re-allocate the bandwidth Wi for each user using a refined formula. The bandwidth formula is defined as Wi = B / ((Na - Ni)+δ), where B is the total available bandwidth, Na represents the current number of active users, Ni is the base number of users already allocated on the current channel, and δ is a small positive value to prevent the denominator from approaching zero and causing a calculation overflow problem. The parameter range is usually that B belongs to the available frequency width interval in the system design, Na and Ni are generated according to the statistical results, and the preferred range of δ is a smaller real number between (0,1] to maintain the numerical stability of the formula. The core idea of this formula setting is to reasonably balance the competitive demands between multiple channels and multiple users and avoid allocation imbalance caused by certain variables. By considering the channel load (Na + Ni) in the allocation model and introducing a small constant to ensure safe operating conditions, fairness can be improved and the system stability can be protected.

[0080] S403: Assign a weighting factor ωi > 1 to high-priority users, and ωi = 1 for others. Subsequently, provide a special weighting mechanism for specific-priority users and apply the weight ωi when allocating bandwidth. For high-priority users, their weighting factor satisfies ωi > 1, and for general users, ωi remains 1. This method is used to balance the experience of users at different levels. The setting of the weight reflects different tolerance requirements for service quality. Specifically, in an application scenario involving video conferencing, compared with low-latency-sensitive services such as just browsing the web, the former is assigned a higher ω value to ensure high-quality network services first.

[0081] S404: Check whether Wi is evenly divided without overflow or being too low. Finally, execute an inspection program to verify whether there is excessive skewness or other unreasonable phenomena after the bandwidth is reallocated. This ensures that each part of users obtains a reasonable proportion of the share while also controlling the overall resource utilization within a normal range. For example, check and confirm that the sum of the bandwidths of all individual users is close to the total amount B and no single user is significantly higher or lower than the predetermined level. Such verification guarantees the feasibility and effectiveness of the bandwidth allocation scheme in the actual application environment.

[0082] Next, refer to Figure 5 , to describe the steps for determining the user priority of the present invention, including:

[0083] S501: Classify to obtain a sequence of Pi values according to the QoS request category, reflecting the importance of the service. This step aims to classify the QoS requests of all users according to different service requirements and assign corresponding priority values Pi. Pi represents the importance of each user request in the service level, with a range of [0, 1], where the higher the value, the higher the demand for real-time performance, stability, or throughput. For example, in a 5G router scenario, the priority of a video conferencing application may be set higher than that of a background file download task.

[0084] S502: When the condition Pi ≥ η * Pavg is satisfied, the user enters the high-level queue. This step determines whether a user is classified into the high-level queue by calculating whether the Pi value of the user is greater than or equal to a dynamic threshold based on the global average priority Pavg. The parameter η is a tuning factor, usually taking values between [0.8, 1.2], used to balance the proportion of high-priority users and the network resource allocation efficiency. The formula sets this condition to ensure that users with critical task requirements can obtain higher-quality services in a timely manner. In one embodiment, assume that a router serves 100 users, the calculated result of Pavg is 0.6, and η is set to 0.9, then only when Pi reaches 0.54 or higher will the corresponding user be placed in the high-level queue.

[0085] S503: For users who do not meet the high - level standards, the bandwidth reduction coefficient α = β^((Fi - Fbar) / γF) is used to dynamically adjust the available bandwidth. In this process, β and γF are two main parameters. The reasonable value range of the β decay rate is [0.5, 0.9], and a smaller value can reduce the resource occupation of non - priority users faster; γF is a smoothing parameter, generally taking values between [1, 3] to ensure that the change curve is more stable and continuous. The numerator part (Fi - Fbar) of the formula is used to measure the deviation degree of the traffic of the i - th user relative to the average traffic Fbar. Finally, the dynamic reduction factor of non - priority group users is determined through this exponential expression. For example, if the current traffic of a non - priority user significantly exceeds the system average level, it will be subject to a higher bandwidth reduction to release resources to support the requirements of high - priority queues.

[0086] S504: Verify whether there is a fairness problem in the reduced bandwidth distribution. That is, it is necessary to compare whether the bandwidth ratio difference within each sub - band is less than a predefined small threshold ε (usually a very small value between [0.01, 0.1]). This is to ensure that even with the high - low priority stratification mechanism, each user in the low - priority group can still obtain a basic fair minimum quality - of - service guarantee. Specifically, if it is found that the ratio differences in some sub - bands exceed the preset threshold, α should be appropriately adjusted or the Fi value should be re - planned until the fairness constraint holds. Through this operation, the goal of consistent optimization of the global user quality of service is achieved.

[0087] Next, refer to Figure 6 , describe the addition of constraint restrictions on the bandwidth reduction conditions of the present invention, including the following steps:

[0088] S601: Introduce a congestion prediction index C, and through machine learning fitting C = f(B, F, Na) to estimate the possible peak value in the future period in advance;

[0089] S602: When the prediction index indicates approaching the critical point pcr, that is, |Cpcr| < σc, trigger an emergency speed - reduction measure; at this time, the per - user limit value Ui = C×Vb*(1 - exp(ηt)), where Vb and ηt are parameters of the speed - adjustment function part;

[0090] S603: Continuously monitor the data packet loss ratio D% during the execution of this mode. If it is found that the allowable loss ratio Δd% is exceeded twice in a row, switch to the stable mode;

[0091] S604: Output the finally determined bandwidth amount to adapt to the current demand.

[0092] Introduce the congestion prediction index C and fit it as C = f(B, F, Na) through machine learning methods. B represents the total bandwidth, F is the traffic model vector (which can include time characteristics and distribution parameters), and Na is the number of active users or the statistical situation of user traffic distribution. The value range of C is in [0, 2], and the optimal prediction value should approximate the measured load trend and be less than 1, indicating that the system is operating normally. This step estimates the network pressure in the next period by analyzing historical data to construct a mapping relationship, ensuring accurate prediction and making decisions in advance.

[0093] In the second step, when C approaches or equals the predefined critical point pcr (i.e., |C - pcr| < σc, and σc is usually set in the range of 0.1 - 0.2), an emergency speed reduction action is triggered. At the same time, the per-user limit value Ui = C × Vb * (1 - exp(-ηt)) is calculated according to the formula, where Vb and ηt represent the initial bandwidth reference value (for example, Vb can be set to a maximum of 80% of the standard total bandwidth) and the speed adjustment factor that changes with time respectively. As the execution duration increases, the rate decreases to prevent excessive resource contention. For example, when a network congestion warning occurs during the peak period from 10 am to 11 am, the maximum uplink speed of all online video conferencing users will be dynamically reduced according to the traffic characteristics at this time, but it will not be immediately reduced to the minimum.

[0094] The third stage requires continuously observing whether the packet loss ratio D% during the execution process exceeds the set tolerance value Δd% twice (assuming Δd% = 5%, indicating that the average packet loss cannot exceed this percentage). Once it is found that this index is not met, switch to a more conservative and stable transmission protocol and service mode to maintain the quality of user experience. Specifically, if signal interference occurs at multiple points in a certain area, resulting in an increase in data loss, immediately adopt a higher redundancy mechanism to ensure that the transmission of key information is not affected.

[0095] The last step will summarize and output the final bandwidth amount adjusted according to the actual situation based on the results of the previous stages. This result is not only used to guide the adjustment of parameters for existing connections but also serves as a feedback to optimize the previous modeling basis and continuously improve efficiency. For example, after completing a complete network load regulation cycle, it is determined to allocate 4 Gbps of shared bandwidth to a specific high-priority Internet of Things device group, and the remaining part is rebalanced to other low-latency service nodes.

[0096] Next, refer to Figure 7 , and describe another embodiment of the intelligent signal optimization method of the 5G router of the present invention, including: calculating channel weights, bandwidth reconfiguration, selecting the winning party in the competition, and verifying the conflict probability index.

[0097] S701: Audit the allocation situation Uj of each channel according to the actual channel utilization efficiency Ue, and adjust the weight wj of the jth channel through the following formula: wj = Ue_j / max_Ueff ^ (χj), where max_Ueff is the total best efficiency value, and χ is the sensitivity factor. First, audit the allocation situation of each channel according to the actual channel utilization efficiency Ue and adjust the weight wj of the jth channel. The formula is wj = Ue_j / max_Ueff ^ (χj). Among them, Ue_j is the utilization efficiency of the jth channel; max_Ueff is the maximum value of the best efficiency values of all channels in the entire system, used as a normalization factor; χ is the sensitivity factor, usually taking values in the range of [0, 1], and the recommended optimal value is 0.7. This formula is used to dynamically allocate the weight of a channel according to its actual performance to avoid high-load or inefficient channels occupying too many resources.

[0098] S702: Accumulate all the unloaded bandwidth shares ΔBi of the previous cycle to the new round of basic configuration. Secondly, accumulate all the unloaded bandwidth shares ΔBi of the previous cycle into the new round of basic configuration. This operation is to make full use of the previously unused bandwidth resources, thereby reducing resource waste and providing a better starting condition for subsequent allocation.

[0099] S703: When multiple groups of requests compete for the same additional unit, enable the voting and sorting method to select the winning group Gi, Gi = SUM(wkj) / (|Lkj| ^ ρ), where ρ controls the distribution smoothing effect. Thirdly, when multiple groups of requests compete for the same additional unit, enable the voting and sorting method to select the winning group Gi. The formula is Gi = SUM(wkj) / (|Lkj| ^ ρ), where wkj is the weight value corresponding to the candidate request k on the jth channel; |Lkj| represents the absolute value of the request volume of the k request covering the jth channel, used to evaluate the competition pressure; ρ is a parameter that controls the distribution smoothing effect, adjusted within the range of [1, 3], and the optimal value is usually set to 2. By incorporating the weight sum and competition impact into a unified formula for comprehensive evaluation, a more fair decision result can be achieved.

[0100] S704: Verify that the conflict probability index CP < ζ after all new layouts. If the determination is successful, fix the allocation. Finally, verify that the conflict probability index CP < ζ after all new layouts, where CP is the potential resource contention frequency statistically obtained based on simulation data; ζ is a preset threshold value, and the specific value depends on the scenario requirements. For example, for an ordinary indoor environment, ζ = 0.05 can be set; for a dense public place, more stringent constraints are required (such as ζ = 0.01). If the verification is successful, fix the allocation; otherwise, re-optimize.

[0101] In one embodiment, assume that a 5G router has 10 available channels. After the first round of operation, the average utilization efficiency of each channel is measured as {0.92, 0.76,...}. Combine the above weight calculation formula and determine an appropriate χ = 0.7, then adjust the weights of each channel. Then summarize the remaining ΔBi in the early stage and reallocate them to the initial configuration pool. At the same time, for traffic competition requests from multiple terminals, select the priority group through Gi decision sorting. Finally, perform conflict detection before the implementation of the new plan to ensure the stability of the plan. This entire process ensures the optimal system performance while improving the user service quality.

[0102] Next, refer to Figure 8 , the signal optimization method of the present invention when a sudden interruption risk occurs includes the following steps:

[0103] S801: Detect that the sudden interruption risk signal Sr exceeds the warning threshold Tr, that is, meet the condition of Sr / Tr≥δr%, and then immediately start to quickly switch to the standby channel group Sc;

[0104] S802: Use the pre-stored mapping table to quickly locate the position number Pos_m of the adapted candidate channel, and ensure that the position error between it and the original path is within the specified range, that is, meet Delta_Distance(Pos_m old_p)<ψ; S801: S803: Adjust the value of the routing parameter Zij of all relevant nodes to Zij = λ×exp[γ*distance(m,n)*time], where γ is used to reflect the delay factor and λ is the weight multiplication coefficient; through this dynamic adjustment mechanism, ensure the stability of the main line and the integrity of service data during the signal switching process are not affected.

[0105] The specific meanings of the above steps are as follows. First, in the case where there may be burst interference or interruption during signal transmission, the system needs to monitor in real time whether the key parameter Sr exceeds the warning threshold Tr. Here, δr% represents a sensitivity setting value of a relative ratio. By determining whether Sr / Tr is equal to or exceeds this sensitivity percentage value, it is decided whether to perform subsequent actions. If there is indeed a risk of interruption, a backup channel is activated to reduce losses. Then, the pre-stored channel configuration information is used to select the optimal alternative channel number Pos_m. To ensure that a relatively small path deviation is still maintained after replacement, the error constraint Delta_Distance(Pos_m old_p) < ψ needs to be followed, that is, the distance between the position of the new channel and the original selected position is less than the predetermined limit ψ, so as to keep the transmission efficiency close to the optimal level. At the same time, the connectivity parameter values between each node are reconfigured, and the calculation formula is Zij = λ × exp[γ * distance(m,n) * time]. Here, a reasonable value is selected for λ to improve the importance of the connection weight, and the best orientation is approximately between 1 and 5, about 2; the delay factor γ is customized according to the actual situation, but the recommended range is from -0.5 to -2, and preferably -1, to balance the trade-off between the response time and resource consumption in different situations; the distance distance(m,n) explicitly indicates the address interval between the current device m and the target site n, and the variable time is introduced to reflect the actual duration consumed by the operation, and then the final corrected value is obtained.

[0106] For example, in a certain practical application scenario, if a 5G router is serving a large number of high-definition video stream users and encounters potential packet loss problems caused by weak ambient noise (the significant increase in Sr breaks through the estimated limit Tr). The system automatically selects a new channel with sufficiently high idle bandwidth and the closest distance as a temporary transmission path according to the algorithm. At the same time, it adjusts the priority order weight settings between all possible points associated with this change point according to the mathematical model to adapt to the existing latency performance characteristics. In one embodiment, when the probability of high packet loss S_rupture in a specific network segment is higher than the safety critical line T_tolerance, that is, S_rupture / T_tolerance reaches a proportional threshold allowing the activation of the emergency strategy, such as 20% or higher, then available alternative communication sequences such as [Channel3A, Channel7B...] are retrieved from the database, and the geographical orientation matching accuracy between these options and the initial deployment plan must be controlled within an acceptable range, such as within ±3 meters; then, according to specific time and space parameters, substitute them into the expression to complete precise mathematical analysis to obtain the ideal result for immediate feedback control. Specifically, in a certain experimental scenario, the lambda optimization value is set to be close to 2 and the negative single exponential function form is used to reflect the gradual smooth transition process with the increase of time and space, which meets the actual needs and does not generate excessive additional computational load. Finally, it achieves the effect of smoothly and efficiently processing network transient changes while ensuring that the customer experience remains continuously excellent, the seamless handover is successfully completed throughout the process, and the core link operation indicators in the normal working state are not affected at all.

[0107] Next, refer to Figure 9 , and describe the process of another embodiment of the present invention, which includes:

[0108] S901: Count the identity identification codes id_s of the original owners to whom the temporarily occupied resources belong, and count their respective remaining amounts to be transmitted L_remaining;

[0109] S902: Construct a step-by-step concession progress plan graph SPM(id_s): Give appropriate return compensation bandwidth Wr = id_s × θ(L_remaining / B_total) to all affected customers one by one;

[0110] S903: Compare the overall throughput T_curr in the current state with the initial expected level T_target to calculate the difference ΔT, and use the formula adjust_factor = log(T_curr + λ) to guide fine-tuning.

[0111] First, count the identity code id_s of the original owner to which the temporarily occupied resources belong, and record its remaining transmission volume L_remaining. This operation aims to clarify which users' traffic is delayed or their priority is reduced, and provide a data basis for subsequent compensation. id_s is a positive integer identification number used to uniquely determine the identity of a certain user; L_remaining is the remaining data transmission volume of this user, and the larger the value, the higher the possibility of additional resource allocation required.

[0112] Next, construct a step-by-step concession progress planning diagram SPM(id_s). This step ensures fair compensation for affected users through a quantization refund mechanism. The formula \(Wr = id_s×θ(L_remaining / B_total)\) determines the compensation bandwidth ratio Wr, where the parameter B_total represents the total bandwidth of the current system, and the θ adjustment factor is usually set in the range of 0.1 to 0.3 to ensure reasonable allocation without excessive resource waste. The best choice depends on the balance of actual usage scenario requirements.

[0113] Subsequently, compare the deviation ΔT (i.e., T_target - T_curr) between the current system throughput status and the preset expected value. The calculated adjust_factor = \(\log(T_{curr}+λ)\), where λ is a non-negative small real number (recommended value such as 0.01), is used to smooth the change of the function input value, and then more sensitive and fine adjustment of the policy optimization target direction is realized. Using the logarithmic form can make the numerical adjustment in a large range tend to be stable and controllable.

[0114] For example, in an embodiment, assume that when N = 50 devices are connected to a high-performance 5G router, a sudden peak load causes some low-priority services to be interrupted. Then, the above method is started for dynamic regulation and the stable operation order is restored. Finally, while maximizing the efficiency of the entire communication network environment, the deterioration of the individual user experience is significantly reduced.

[0115] An intelligent signal optimization method for a 5G router of the present invention includes: First, count the number of real-time users connected to the 5G router, and obtain the current channel load status through data analysis. This process collects real-time data in the network, combines specific algorithms to calculate the access situation of user devices and their bandwidth usage ratios in each time period, so as to dynamically evaluate whether there is congestion in the current channel or whether its load tends to be saturated. Secondly, after clarifying the channel load status, further classify and grade the network status according to the preset congestion level evaluation criteria, which can be specifically divided into multiple levels such as slight congestion, moderate congestion, and severe congestion. Each level corresponds to different degrees of bandwidth bottleneck problems.

[0116] Based on the above evaluation results of network congestion, the system will intelligently adjust the channel bandwidth allocation strategy. For example, in the case of slight congestion, different types of traffic allocation are reasonably divided by optimizing the priority rules; in the case of severe congestion, more stringent data flow control or reduction of non-critical data transmission is adopted. This method effectively solves problems such as slow network speed and poor signal quality caused by multiple devices simultaneously occupying the limited channel. Finally, the newly obtained bandwidth allocation strategy will be immediately deployed to the 5G router to complete the actual application of signal optimization. This technical solution ensures that the entire process can be automatically executed without manual operation, significantly improving the adaptability and processing ability of the router in the face of high-concurrency connections. Therefore, this method provides an effective technical guarantee for improving the network experience of users in the 5G communication environment.

[0117] The specific embodiments described above further elaborate on the purpose, technical solutions, and beneficial effects of the embodiments of the present disclosure. It should be understood that the above are only the specific embodiments of the embodiments of the present disclosure and are not used to limit the protection scope of the embodiments of the present disclosure. Any modifications, equivalent replacements, improvements, etc. made within the spirit and principle of the embodiments of the present disclosure shall be included in the protection scope of the embodiments of the present disclosure.

Claims

1. An intelligent signal optimization method for a 5G router, characterized in that, Including: Determine the current channel load status according to the statistical result of the real-time user connection number; Evaluate the network congestion level according to the channel load status; Based on the network congestion level, adjust the channel bandwidth allocation strategy; Apply the adjusted bandwidth allocation strategy to the router to optimize signal transmission.

2. The intelligent signal optimization method of a 5G router according to claim 1, characterized in that, Adjusting the channel bandwidth allocation strategy based on the network congestion level includes the following steps: Determine that the real-time user connection number is N; Calculate the current network bandwidth B, and set the maximum bandwidth as Bmax; According to the formula W = Bmax * exp(N / λ), dynamically adjust the bandwidth W per user, where λ is a predefined load parameter; When W is less than the minimum guaranteed bandwidth Wmin, set W = Wmin.

3. The intelligent signal optimization method of a 5G router according to claim 2, characterized in that, Further include the following steps: Monitor the user request data traffic F, and record the average rate R = the data volume within the time window of F / T; Evaluate the current channel load status and classify the level according to the following rules: If R > μB, the level is severe congestion, otherwise it is mild congestion, where μ is the system bandwidth utilization efficiency coefficient; Input the classification result into the adjustment algorithm, and dynamically update the channel number K based on the channel congestion level g, K = K_base + ΔK * (1 - g), where ΔK is the increment value; Set an upper limit for the total bandwidth of all users as max(K × W) and ensure that it does not exceed B.

4. The intelligent signal optimization method of a 5G router according to claim 3, characterized in that, Further adjust the bandwidth allocation based on the channel number K, including the following steps: Recount the number of active users Na, excluding offline users whose idle time exceeds Ta; Refine the allocation of the bandwidth Wi for each user by applying a new formula: Wi = B / ((Na - Ni) + δ), where Ni is the base number of users already allocated on the current channel, and δ is a small positive value; Assign a weighting factor ωi > 1 to high-priority users, and ωi = 1 for others; Check whether Wi is evenly divided and there is no overflow or too low phenomenon.

5. The intelligent signal optimization method of a 5G router according to claim 4, wherein, Further stipulate that the user priority determination step includes the following content: Obtain the Pi value sequence according to the QoS request category classification, reflecting the service importance; If Pi >= η * Pavg is satisfied, enter the high-level queue, where Pavg represents the global average value; Apply the bandwidth reduction coefficient α = β^((Fi - Fbar) / γF) to non-priority group users, where β is the attenuation rate and γF is the smoothing parameter; Verify whether there is a fairness problem after reduction, that is, compare whether the ratio difference in each sub-band is less than the small threshold ε preset in advance.

6. The intelligent signal optimization method of a 5G router according to claim 5, characterized in that, Add constraint limitations to the bandwidth reduction conditions, including the following steps: Introduce a congestion prediction index C, and fit C = f(B, F, Na) through machine learning, which is used to estimate the possible peak value in the future period in advance; When the prediction index indicates approaching the critical point pcr, that is, |C - pcr| < σc, trigger an emergency speed reduction measure; at this time, the per-user limit value Ui = C × Vb * (1 - exp(-ηt)), where Vb and ηt are partial parameters of the speed adjustment function; Continuously monitor the data packet loss ratio D% during the execution of this mode. If it is found that the allowable loss ratio Δd% is exceeded continuously twice, switch to the stable mode; Output the finally determined bandwidth amount to adapt to the current demand.

7. The intelligent signal optimization method of a 5G router according to claim 1, characterized in that, Also include: Audit the allocation of each channel Uj according to the actual channel utilization efficiency Ue, and adjust the weight wj of the jth channel through the following formula: wj = Ue_j / max_Ueff^(χj), where max_Ueff is the total optimal efficiency value and χ is the sensitivity factor; Accumulate all the unloaded bandwidth shares ΔBi in the previous period and add them to the new round of basic configuration; When multiple groups of requests compete for the same additional unit, enable the voting sorting method to select the winning group Gi, Gi = SUM(wkj) / (|Lkj|^ρ), where ρ controls the distribution smoothing effect; Verify that the conflict probability index CP < ζ after all new layouts. If the determination is successful, fix the allocation.

8. The intelligent signal optimization method of a 5G router according to claim 1, wherein, Also included: When it is detected that the sudden interruption risk signal Sr is greater than the warning threshold Tr, i.e., Sr / Tr >= δr%, start the fast-switching standby channel group Sc; Quickly locate and match the alternative channel position number Pos_m through the pre-stored mapping table, so that the distance error from the original path Delta_Distance(Pos_m old_p) < ψ; At the same time, adjust all the node routing parameters Zij involved, Zij = λ × exp[γ * distance(m,n) * time], where γ reflects the delay factor and λ refers to the weight multiplication coefficient.

9. The intelligent signal optimization method of a 5G router according to claim 1, characterized in that, Also included: Statistically count the identity identification codes id_s of the original owners to whom the temporarily occupied resources belong, and count their respective remaining amounts to be transmitted L_remaining; Construct a step-by-step concession progress planning chart SPM(id_s): give appropriate refund compensation bandwidth Wr = id_s × θ(L_remaining / B_total) to all affected customers one by one; Compare the overall throughput T_curr in the current state with the initial expected level T_target to calculate the difference ΔT, and use the formula adjust_factor = log(T_curr + λ) to guide fine-tuning.

Images