Wheel robot communication network dynamic load balancing method, device and equipment

Abstract

The application provides a wheeled robot communication network dynamic load balancing method, device and equipment, relates to the technical field of communication networks, and comprises the following steps: acquiring a first bias value associated with a macro base station and a second bias value associated with a small base station in a target area; selecting an associated base station according to the real-time communication channel condition of a wheeled robot in the target area and the initial bias value, and collecting relevant data to identify overloaded base stations and lightly loaded base stations; selecting a first wheeled robot that meets preset migration conditions from the overloaded base stations, switching the first wheeled robot to the lightly loaded base station, and acquiring a switching success rate, which is used for dynamically adjusting the first bias value and the second bias value. By dynamically adjusting the bias value, the balanced distribution of wheeled robots among base stations is guided, and the network resource utilization rate is improved.

Description

TECHNICAL FIELD

[0001] The embodiment of the present application relates to the technical field of communication network, and particularly relates to a wheel robot communication network dynamic load balancing method, device and equipment. BACKGROUND

[0002] With the development of intelligent transportation system, the reliability and delay requirement of wheel robot communication to network is higher and higher. The traditional network resource sharing mode is difficult to meet the demand of different services, especially in the high-density wheel robot environment, the network congestion and uneven bandwidth allocation problem is particularly prominent, a large number of wheel robot terminals access the network at the same time, which may lead to unbalanced load of base station, some base stations are overloaded while other base stations are idle, thereby affecting the communication quality.

[0003] At the same time, due to the high-speed mobility of wheel robot, the transfer of service traffic in space also has great diversity. This leads to the continuous change of user load of the whole wheel robot communication network, thereby bringing great challenge to the wireless resource management between wheel robot communication network slices. SUMMARY

[0004] The embodiment of the present application provides a wheel robot communication network dynamic load balancing method, which intelligently guides the balanced distribution of wheel robot terminals between base stations by dynamically adjusting the bias values of macro base stations and small base stations, and significantly improves the network resource utilization.

[0005] In the first aspect, the embodiment of the present application provides a wheel robot communication network dynamic load balancing method, comprising:

[0006] obtaining initial bias values of base stations in a target area, the base stations including macro base stations and small base stations, the initial bias values including first bias values and second bias values, the first bias values being associated with the macro base stations, and the second bias values being associated with the small base stations;

[0007] selecting associated base stations according to real-time communication channel conditions of wheel robots in the target area and the initial bias values, collecting load data of the base stations, channel states of the wheel robots and moving performance of the wheel robots, and identifying overloaded base stations and lightly loaded base stations;

[0008] selecting a first wheel robot satisfying preset migration conditions from the overloaded base stations, switching the first wheel robot to the lightly loaded base stations, and obtaining a switching success rate, the switching success rate being used for dynamically adjusting the first bias values and the second bias values.

[0009] It can be understood that the bias value is a configurable parameter of the base station (such as the macro base station MBS and the small base station SBS), which is used to adjust the priority of the wheeled robot to select the associated base station. By differentially setting the bias value, guiding the user to migrate from the high-load base station to the low-load base station can avoid local congestion. By dynamically adjusting the bias value, it is ensured that the high-priority service (such as automatic driving) is always associated with the base station with sufficient resources. In this way, by dynamically adjusting the bias value of the macro base station and the small base station, the wheeled robot terminal is intelligently guided to balance the distribution between the base stations, and the network resource utilization is significantly improved.

[0010] In a possible implementation, the initial bias value includes a first bias value and a second bias value, and the first bias value is smaller than the second bias value.

[0011] In a possible implementation, the handover success rate is used to dynamically adjust the first bias value and the second bias value, including:

[0012] In the case where the handover success rate is higher than a preset value, the first bias value is reduced, and the second bias value is increased.

[0013] It can be seen that the high-load base station access is automatically inhibited by the negative feedback mechanism. When the handover success rate is high (for example, > 90%), it indicates that the network state is stable, at this time, reducing the bias value of the macro base station can reduce the access of new users, and increasing the bias value of the small base station can effectively offload the wheeled robot. The strategy can significantly reduce the load difference between base stations. At the same time, dynamic adjustment keeps the handover success rate in the optimal interval (for example, 85%-95%), avoiding the ping-pong effect caused by sudden changes in the bias value, and controlling the handover interruption time of the high-speed wheeled robot to be within 20ms, while ensuring that the throughput fluctuation of the video stream service does not exceed 15%. This closed-loop control can keep the network always working at the best load balancing point.

[0014] In a possible implementation, the first wheeled robot that meets the preset conditions for migration is selected from the overloaded base station, and the first wheeled robot is associated and switched to the lightly loaded base station, and the preset conditions include at least one of the following:

[0015] The target base station after migration can meet the QoS requirement of the wheeled robot, the wheeled robot with a moving speed lower than a set speed threshold, and the wheeled robot with a channel condition better than a set threshold.

[0016] In a possible implementation, the method further includes:

[0017] predicting the moving path of the wheeled robot and a set of potential target base stations for switching, pre-rehearsing the switching process in the digital twin environment, calculating a set of predicted QoS indicators, and virtually migrating the target wheeled robot to the candidate base station;

[0018] When all predicted QoS indicators meet the corresponding service demand threshold, trigger the actual base station switching operation.

[0019] It can be understood that by pre-screening the eligible migrating wheeled robots and combining digital twin verification, the overall network resource utilization is improved under the premise of ensuring service quality, effectively avoiding the coexistence of idle and congested resources.

[0020] In a possible implementation, the method further includes:

[0021] When the QoS indicator after actual switching deviates from the predicted value by more than a set threshold, roll back to the original base station connection.

[0022] It can be understood that when the actual QoS indicator such as time delay / packet loss rate deviates from the predicted value by more than a preset value (for example, 10%) (such as when the time delay of autonomous driving service is deteriorated from the predicted 15 ms to 30 ms), a rollback operation is triggered immediately to ensure safety. At the same time, by quickly rolling back to reduce the number of invalid switching, the signaling overhead can be reduced.

[0023] In a possible implementation, the digital twin model is retrained automatically through a rollback event to continuously improve the prediction accuracy. In combination with historical rollback data, network weak points (such as a specific regional base station coverage blind area) can be identified, so that the system has the ability to learn from anomalies, and the prediction deviation rate can be reduced.

[0024] In a possible implementation, the method further includes:

[0025] When there is a high-priority wheeled robot in the target area, a bias value adjustment exemption flag is set for the high-priority wheeled robot, and the base station associated with the high-priority wheeled robot does not participate in the dynamic bias value reduction, and the high-priority wheeled robot includes a wheeled robot performing an emergency task or an authorized emergency communication wheeled robot.

[0026] It can be seen that in the embodiments of the present application, by establishing a pyramid structure of service levels, it is ensured that the emergency communication wheeled robot always obtains the optimal network resources. By setting the bias value adjustment exemption signaling flag mechanism, the exemption decision delay can be effectively reduced.

[0027] In a second aspect, the embodiments of the present application also provide a wheeled robot communication network dynamic load balancing device, comprising:

[0028] The acquisition module is configured to acquire an initial bias value of a base station in a target area, wherein the base station includes a macro base station and a small base station, and the initial bias value includes a first bias value and a second bias value, the first bias value is associated with the macro base station, and the second bias value is associated with the small base station.

[0029] The identification module is configured to select an associated base station according to real-time communication channel conditions of the wheeled robot in the target area and the initial bias value, and collect load data of the base station, channel state of the wheeled robot and movement performance of the wheeled robot, and identify overloaded base stations and lightly loaded base stations.

[0030] The switching module is configured to select a first wheeled robot that meets preset migration conditions from the overloaded base stations, switch the first wheeled robot to the lightly loaded base station, and obtain a switching success rate, the switching success rate being used to dynamically adjust the first bias value and the second bias value.

[0031] In a possible implementation, the initial bias value includes a first bias value and a second bias value, and the first bias value is smaller than the second bias value.

[0032] In a possible implementation, the switching module is further configured to:

[0033] In a case where the switching success rate is higher than a preset value, the first bias value is reduced, and the second bias value is increased.

[0034] In a possible implementation, the switching module is further configured to select a first wheeled robot that meets preset migration conditions from the overloaded base stations, and switch the first wheeled robot to the lightly loaded base station, and the preset conditions include at least one of the following:

[0035] a wheeled robot whose post-migration target base station can meet QoS requirements, a wheeled robot whose movement speed is lower than a set speed threshold, and a wheeled robot whose channel conditions are better than a set threshold.

[0036] In a possible implementation, the device further includes a prediction module configured to:

[0037] predict a movement path of the wheeled robot and a set of potential target base stations for switching, pre-act the switching process in a digital twin environment, calculate a set of predicted QoS indicators, and virtually migrate the target wheeled robot to a candidate base station.

[0038] When all the predicted QoS indicators meet corresponding service requirement thresholds, trigger an actual base station switching operation.

[0039] In a possible implementation, the device further includes a rollback module configured to:

[0040] When a QoS indicator after actual switching deviates from a predicted value by more than a set threshold, roll back to a connection with the original base station.

[0041] In a possible implementation, the device further includes an adjustment module configured to:

[0042] When a high-priority wheeled robot exists in the target area, a bias value adjustment exemption flag is set for the high-priority wheeled robot, a base station associated with the high-priority wheeled robot does not participate in dynamic bias value reduction, and the high-priority wheeled robot includes a wheeled robot performing an emergency task or an authorized emergency communication wheeled robot.

[0043] With reference to the first aspect, in the second aspect, the related beneficial effects can refer to the first aspect, and the embodiments of the present application will not be described here.

[0044] In a third aspect, the embodiments of the present application further provide an electronic device, which comprises:

[0045] at least one processor;

[0046] and a memory in communication connection with the at least one processor;

[0047] The memory stores instructions executable by the at least one processor, and the instructions are executed by the at least one processor to enable the electronic device to perform the method corresponding to any of the embodiments of the first aspect of the present application.

[0048] In a fourth aspect, the embodiments of the present application further provide a computer readable storage medium, which stores computer execution instructions, and the computer execution instructions are executed by a processor to implement any method of the first aspect of the present application.

[0049] In a fifth aspect, the present disclosure further provides a computer program product, which comprises computer execution instructions, and the computer execution instructions are executed by a processor to implement the method of any embodiment corresponding to the first aspect of the present disclosure. BRIEF DESCRIPTION OF DRAWINGS

[0050] The accompanying drawings, which are incorporated into and form part of the specification, illustrate embodiments consistent with the present disclosure and, together with the specification, serve to explain the principles of the present disclosure.

[0051] In order to more clearly illustrate the technical solutions in the embodiments of the present disclosure or the prior art, the drawings needed in the embodiments or the prior art description will be briefly introduced as follows. Obviously, for those skilled in the art, other drawings can also be obtained without creative labor based on these drawings.

[0052] Figure 1 A wheeled robot communication network dynamic load balancing method flowchart provided by the embodiments of the present application.

[0053] Figure 2 A wheeled robot communication network dynamic load balancing device schematic diagram provided by the embodiments of the present application.

[0054] Figure 3 FIG. 1 shows a schematic diagram of an electronic device according to an embodiment of the present application. DETAILED DESCRIPTION

[0055] In the following description, like reference characters designate like or similar elements in the figures. The embodiments described in the following examples are merely exemplary of devices and methods that are in accordance with aspects of the present application.

[0056] The following specific embodiments can be combined with each other, and some of the same or similar concepts or processes can not be repeated in some embodiments. The embodiments of the present application will be described below with reference to the accompanying drawings.

[0057] It should be noted that, in this document, the terms "first" and "second" and the like are used merely to distinguish one entity or action from another, and do not necessarily require or imply any actual such relationship or order between such entities or actions. Moreover, the terms "comprises", "comprising", or any other variations thereof are intended to cover a non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements does not include only those elements recited, but can also include other elements not expressly listed or inherent to such process, method, article, or apparatus. Without more limitations, an element defined by the phrase "comprising a..." does not exclude the existence of additional identical elements in the process, method, article, or apparatus that includes the element.

[0058] The embodiments of the present disclosure can be applied to the control of various devices such as multiple wheeled robots, mobile robots, aircraft, ships, intelligent rail rapid transit systems (ART, Autonomous rail Rapid Transit), etc. The wheeled robots can be, but are not limited to, passenger cars, commercial vehicles (e.g., trucks, buses, vans, etc.), special-purpose wheeled robots (e.g., ambulances, fire engines, engineering vehicles, rescue vehicles, etc.), agricultural and industrial wheeled robots (e.g., harvesters, forklifts, etc.), transportation and logistics wheeled robots (e.g., container trucks, refrigerated trucks, etc.), new energy wheeled robots (e.g., electric cars, hybrid cars), special vehicles (e.g., garbage trucks, water trucks, etc.). In other words, the "wheeled robot" in the embodiments of the present disclosure is equivalent to the various devices described above. It can be used in scenarios such as unmanned vehicles, manned vehicles, and mixed traffic flow.

[0059] The embodiments of the present disclosure can be applied to urban traffic, highways, ports, mines, farms, closed parks and the like, and can be applied to many aspects such as car travel, public transportation, logistics distribution, unmanned transportation, end distribution, automated agricultural operation, and automated environmental sanitation. Of course, the embodiments of the present disclosure can also be applied to other intelligent control scenarios involving devices such as wheeled robots, and the present disclosure is not limited to the application scenarios and application fields of the embodiments of the present disclosure.

[0060] Exemplarily, the embodiments of the present disclosure can be applied to the following four aspects:

[0061] 1) Urban roads: In a complex urban traffic environment, the embodiments of the present disclosure can handle variable traffic flow and diverse traffic participants, and provide accurate trajectory prediction.

[0062] 2) Highways: On highways, the embodiments of the present disclosure can efficiently, quickly and accurately realize trajectory prediction of wheeled robots in high-speed driving.

[0063] 3) Closed areas such as ports and parks: In a closed or semi-closed environment, the embodiments of the present disclosure are applied to autonomous wheeled robots, which can realize efficient, fast and accurate navigation of autonomous wheeled robots, improve efficiency and reduce labor costs.

[0064] 4) Shared travel service: Integrating the embodiments of the present disclosure into autonomous taxis or shared wheeled robots helps to provide safe and reliable travel services.

[0065] The network used by the wheeled robot data transmission resource of the embodiments of the present application can be, but is not limited to, a Long Range Radio (LoRa) module, a Narrow Band Internet of Things (NB-IoT) module, an Enhanced Machine-Type Communication (eMTC) module or other similar communication modules. It can also be, but is not limited to, a module that supports one or more of the following wireless communication methods: mobile communication, Long Term Evolution Vehicle-to-Everything (LTE-V), Dedicated Short-Range Communication (DSRC), Cellular Vehicle-to-Everything (C-V2X), and Vehicle-to-Everything (V2X).

[0066] In the embodiments of the present application, the following terms are exemplarily explained.

[0067] Macro Base Station (MBS): usually refers to a large power base station with a coverage radius ≥ 500m, working frequency band is Sub-3GHz, wide coverage, suitable for serving high-speed mobile wheeled robots.

[0068] Small Base Station (SBS): usually refers to a low-power node with a coverage radius ≤ 200m, working frequency band is 3.5-28GHz. Small coverage, high capacity, suitable for serving low-speed mobile wheeled robots or hot spot area wheeled robots.

[0069] Macro Base Station Bias (MBS): that is, the first bias value, lower, encourages wheeled robots to preferentially associate with small base stations.

[0070] Small Base Station Bias (SBS): that is, the second bias value, higher, to attract wheeled robots to associate with small base stations.

[0071] Overloaded base station: reduce the bias value, reduce the association of new wheeled robots.

[0072] Light load base station: increase the bias value, attract more wheeled robots to associate.

[0073] Cell Individual Offset (CIO): a parameter used to adjust the switching behavior of wheeled robots, by artificially increasing or decreasing the signal measurement value (such as RSRP / RSRQ) of the base station, affecting the terminal's tendency to access the macro base station or the small base station. Its core functions include:

[0074] (1) Load balancing: by dynamically adjusting CIO, guide the terminal to preferentially access the lightly loaded base station (such as increasing the CIO of the small base station to offload the pressure of the macro station).

[0075] (2) Handover optimization: in high-speed mobile scenarios, adjusting CIO can reduce ping-pong handover (such as increasing the CIO of the macro station to make the wheeled robot switch earlier).

[0076] (3) Service classification: set a fixed CIO exemption for high-priority services (such as emergency wheeled robots) to guarantee the quality of service (QoS).

[0077] Figure 1 A wheeled robot communication network dynamic load balancing method provided by the embodiment of the application, by dynamically adjusting the bias value of the macro base station and the small base station, intelligently guiding the balanced distribution of the wheeled robot terminal between the base stations, significantly improving the network resource utilization.

[0078] In Figure 1 , it includes steps S101, S102 and S103. Specifically:

[0079] Step S101, obtaining an initial bias value of a base station in a target area, the base station including a macro base station and a small base station, the initial bias value including a first bias value and a second bias value, the first bias value being associated with the macro base station, and the second bias value being associated with the small base station.

[0080] In the embodiment of the present application, the initial bias value contains two independent parameters: the first bias value is used for the macro base station, and the second bias value is used for the small base station. For example, the first bias value Bias_MBS∈[-5dB, 0dB] is used for the macro base station, and the second bias value Bias_SBS∈[+3dB, +10dB] is used for the small base station.

[0081] Step S102, selecting an associated base station according to real-time communication channel conditions of a wheeled robot in the target area and the initial bias value, and collecting load data of the base station, channel state of the wheeled robot, and moving performance of the wheeled robot, to identify an overloaded base station and a lightly loaded base station.

[0082] In a possible implementation, the first wheeled robot meeting preset conditions for migration is selected from the overloaded base station, and the first wheeled robot is associated and switched to the lightly loaded base station, and the preset conditions include at least one of the following:

[0083] The post-migration target base station can meet the QoS requirement of the wheeled robot, the wheeled robot with a moving speed lower than a set speed threshold, and the wheeled robot with a channel condition better than a set threshold.

[0084] For example, in a specific implementation, the wheeled robot to be migrated is selected from the overloaded base station, and needs to meet the following multi-dimensional conditions at the same time: first, the target base station must have the ability to guarantee the quality of service of the wheeled robot, that is, the post-migration target base station can continuously meet the QoS indicators such as delay and bandwidth required by the wheeled robot service; second, a wheeled robot terminal with low speed is preferentially selected, and the speed value needs to be lower than a speed threshold (v_th = R_cell / T_handover, where R_cell is the base station coverage radius, and T_handover is the handover delay) dynamically calculated according to the base station coverage range; finally, the wireless channel condition of the candidate wheeled robot in the current base station should be better than a preset threshold value, including key indicators such as reference signal received power (RSRP>-85dBm) and signal-to-interference noise ratio (SINR>15dB). The combination of the three conditions not only ensures the continuity of the service after migration, but also maximally reduces the risk of handover failure caused by high-speed movement of the wheeled robot, and guarantees the communication quality of the new link after migration.

[0085] In the embodiment of the present application, based on the real-time wireless channel measurement data (including reference signal received power, signal-to-noise ratio and other key indicators) reported by the wheeled robot terminal, combined with the pre-configured differentiated bias value parameter, a virtual signal quality evaluation model is constructed; secondly, through the control plane interface between the core network and the base station, the load state information such as resource utilization rate and access user number of each base station is obtained in real time; at the same time, using the mobility parameters (such as position coordinates, motion speed, acceleration, etc.) provided by the vehicle-mounted communication unit, a multi-dimensional base station selection decision matrix is established. On this basis, a dynamic threshold decision algorithm is adopted, and the base stations whose resource utilization rate continuously exceeds the preset upper limit are marked as overload nodes, and the base stations whose resource utilization rate is long-term lower than the preset lower limit are identified as light-load nodes, thereby providing accurate decision basis for subsequent load balancing adjustment.

[0086] For example, the wheeled robot terminal measures the reference signal received power RSRP (range -140dBm to -44dBm) and signal-to-interference noise ratio SINR (accuracy ±0.5dB) of the serving base station and the adjacent area every 200ms, and calculates the virtual signal strength in combination with the pre-configured initial bias value (macro base station [-5dB, 0dB], small base station [+3dB, +10dB]):

[0087] RSRP virtual =RSRP measured +Bias+ΔHys (hysteresis parameter 3dB)

[0088] After screening out the candidate base stations with RSRP virtual >-110dBm, the optimal base station meeting the service QoS requirement is selected in descending order of SINR; at the same time, the base station load data (including PRB utilization rate sampling period 1 second, active user number statistical granularity 5 seconds, backhaul link load rate accuracy ±2%) are collected through the X2 interface, combined with the real-time speed vector (GPS / INS fusion positioning error <0.5m), acceleration data (range ±2g) and motion direction angle (accuracy ±3°) provided by the vehicle-mounted OBU, when the base station PRB utilization rate is continuously >80% for 30 seconds and the user number is >90% of the capacity, it is judged as an overload base station, and when the PRB utilization rate is continuously <40% for 60 seconds and the user number is <50% of the capacity, it is judged as a light-load base station, Kalman filter is used to compensate for missing data and eliminate outliers during the process, and finally the base station selection accuracy and the load identification time delay within 300ms are realized.

[0089] In step S103, a first wheeled robot meeting the preset migration condition is selected from the overload base station, the first wheeled robot is associated with the light-load base station, and the handover success rate is obtained, which is used to dynamically adjust the first bias value and the second bias value.

[0090] In the embodiments of the present application, intelligent user migration strategy is adopted to realize load balancing, select candidate wheeled robot terminals (i.e. first wheeled robot) meeting the conditions from the overloaded base station, and intelligently switch the service connection of the first wheeled robot to the lightly loaded base station, while the success rate of the switching operation is counted in real time. The switching success rate as a key feedback parameter dynamically adjusts the bias weights of the macro base station (corresponding to the first bias value) and the small base station (corresponding to the second bias value), thereby forming a closed-loop optimization control mechanism.

[0091] For example, the target wheeled robot (first wheeled robot) meeting the requirements is selected from the overloaded base station, and the migration preset conditions include but are not limited to: 1) wheeled robot channel quality indicator (RSRP≥-85dBm and SINR≥15dB); 2) movement characteristics (speed≤preset threshold and motion trajectory stability≥0.8); 3) service demand matching degree (QoS level that the target base station can provide≥current service requirement level). Through signaling interaction, the service connection of the selected first wheeled robot is seamlessly switched to the pre-identified lightly loaded base station, and a double confirmation mechanism is adopted in the switching process to ensure service continuity. The system monitors and records the success rate of the switching operation (defined as the number of successful switching / total number of attempted switching×100%) in real time, and when the switching success rate within the statistical window period (default 300 seconds) exceeds the set threshold (such as 90%), the bias value dynamic adjustment algorithm is triggered: according to the load balancing state and network performance indicators, the first bias value of the macro base station and the second bias value of the small base station are optimized and adjusted in coordination with 0.5dB as the step unit, wherein the adjustment direction of the first bias value is negatively correlated with the switching success rate, and the adjustment direction of the second bias value is positively correlated with the switching success rate, thereby realizing adaptive balancing of network load. This mechanism can ensure that the resource utilization rate is improved while effectively reducing the service interruption probability.

[0092] As can be understood, in the embodiments of the present application, the bias value is a configurable parameter of the base station (such as the macro base station MBS and the small base station SBS) for adjusting the priority of the wheeled robot selecting the associated base station. By differentially setting the bias value, the user is guided to migrate from the high-load base station to the low-load base station, which can avoid local congestion. By dynamically adjusting the bias value, high-priority services (such as automatic driving) are always associated with base stations with sufficient resources. In this way, by dynamically adjusting the bias values of the macro base station and the small base station, the wheeled robot terminals are intelligently guided to balance between base stations, significantly improving the network resource utilization rate.

[0093] In a possible implementation manner, the initial bias value includes a first bias value and a second bias value, and the first bias value is smaller than the second bias value.

[0094] In a possible implementation manner, the switching success rate is used to dynamically adjust the first bias value and the second bias value, including:

[0095] In the case that the handover success rate is higher than the preset value, the first bias value is reduced and the second bias value is increased.

[0096] For example, when the handover success rate is monitored to be higher than a preset threshold (e.g., 95%), the first bias value is reduced by 1-3 dB to make the wheeled robot more inclined to switch to the small base station, and the second bias value is increased by 1-3 dB to enhance the access attraction of the small base station. It can be understood that when the handover success rate is lower than the threshold (e.g., 95%), the parameters are adjusted in the opposite direction, i.e., the first bias value is increased by 1-3 dB and the second bias value is reduced by 1-3 dB. In addition, a gradient adjustment algorithm can also be introduced to adjust the handover success rate, for example:

[0097]

[0098] wherein ΔOffset is the adjustment amount of the bias value, the bias value (the first / second bias value) is dynamically adjusted by quantifying the deviation of the current success rate from the target success rate, so that the handover success rate approaches the ideal threshold; and a is the learning rate (e.g., set to 0.5), which is automatically relaxed by 5% when the success rate threshold during the morning and evening peak periods is changed greatly due to traffic.

[0099] It can be seen that the high-load base station access is automatically inhibited by the negative feedback mechanism. When the handover success rate is high (e.g., > 95%), it indicates that the network state is stable, at this time, reducing the bias value of the macro base station can reduce the access of new users, and increasing the bias value of the small base station can effectively offload the wheeled robot. The strategy can significantly reduce the load difference between base stations. At the same time, the dynamic adjustment keeps the handover success rate in the optimal interval (e.g., 85%-95%), avoiding the ping-pong effect caused by sudden changes in the bias value, and controlling the handover interruption time of the high-speed wheeled robot to be within 20 ms, while ensuring that the throughput fluctuation of video streaming and other services does not exceed 15%. This closed-loop control can keep the network working at the optimal load balancing point.

[0100] In the embodiments of the present application, the bias value is dynamically adjusted according to the base station load rate, and the bias value of the overloaded base station is updated according to the following formula:

[0101]

[0102] wherein a is the adjustment coefficient, Bias new is the new bias value after adjustment, Bias old is the current bias value (before adjustment), Load current is the current load rate of the base station, Load threshold is the preset load threshold.

[0103] In the above formula, if the difference is positive (current load > threshold), the base station is overloaded and the bias value needs to be lowered to reduce the number of new wheeled robots joining. If the difference is negative (current load < threshold), the base station is underloaded and the bias value can be increased appropriately to attract new wheeled robots. A larger value for α results in a more drastic bias response to load changes (rapid adjustment but potentially causing oscillations), while a smaller value for α results in a smoother adjustment (higher stability but delayed response).

[0104] In the above formula, when the load is severely overloaded, an exponential function is used to accelerate the adjustment:

[0105]

[0106] In one possible implementation, a multi-parameter fusion method can be used to dynamically calculate the bias value. The specific formula is as follows:

[0107]

[0108] where Δ load : Load deviation (in the above formula for calculating the bias value of the overloaded base station, ), P i,norm : the normalized QoS parameters (delay / packet loss / jitter, etc.) of the i-th type of service, β: the weight coefficient of the load item (usually set to 0.5~0.7), w i : Parameter weight corresponding to the service type. Service types include but are not limited to autonomous driving, 4K video streaming, and IoT telemetry.

[0109] In the above formula, when an emergency scenario triggers: when any QoS parameter exceeds the threshold, its weight is automatically increased using the following formula:

[0110]

[0111] Among them, w i ': is the adjusted service type weight, which is used for subsequent bias value calculation or resource allocation; P i : The actual QoS parameter value of the current service (such as delay, packet loss rate, jitter, etc.); Pthresh: The threshold of the QoS parameter of the service (the maximum / minimum critical value allowed).

[0112] In a possible implementation, the method further includes:

[0113] Predict the wheeled robot's movement path and the set of potential handover target base stations, rehearse the handover process in a digital twin environment, calculate and predict the QoS indicator set, and virtually migrate the target wheeled robot to the candidate base station;

[0114] When all predicted QoS indicators meet the corresponding service demand threshold, trigger the actual base station switching operation.

[0115] Exemplarily, in the embodiments of the present application, an intelligent switching rehearsal method based on digital twinning is provided. First, the improved LSTM neural network is used to predict the future 10-second moving path and the potential set of switching base stations by fusing the wheeled robot motion parameters (GPS coordinates, velocity vector, acceleration) and high-precision map data; then, a virtual simulation platform containing real-time network state (base station load, channel condition, backhaul delay) is constructed in the digital twinning environment, the switching process is rehearsed, and key QoS indicators such as delay and interruption time are accurately calculated; when all predicted indicators meet the service demand threshold (such as autonomous driving delay ≤ 20 ms) and 10% safety margin is reserved, the actual switching operation is triggered, and a real-time monitoring feedback mechanism is established. When the actual performance deviates from the predicted value by more than 5%, the model parameters are automatically rolled back and optimized, and finally the switching success rate of 98.5% is realized, the business interruption time is compressed to within 10 ms, and the number of invalid switches is reduced.

[0116] It can be understood that by pre-screening the qualified migrating wheeled robots and combining with digital twinning verification, the overall network resource utilization is improved under the premise of ensuring service quality, effectively avoiding the coexistence of resource idling and congestion.

[0117] In a possible implementation manner, the method further includes:

[0118] When the QoS indicator after the actual switching deviates from the predicted value by more than a set threshold, roll back to the original base station connection.

[0119] It can be understood that when the QoS indicators such as actual delay / packet loss rate deviate from the predicted value by more than a preset value (for example, 10%) (such as the delay of autonomous driving service deteriorates from the predicted 15 ms to 30 ms), the rollback operation is triggered immediately to ensure safety. At the same time, by quickly rolling back, the number of invalid switches is reduced, and the signaling overhead can be reduced.

[0120] In a possible implementation manner, the digital twinning model is automatically triggered for retraining through the rollback event, which can continuously improve the prediction accuracy. Combined with historical rollback data, the network weak points (such as specific regional base station coverage blind area) can be identified, so that the system has the ability to learn from anomalies, and the prediction deviation rate can be reduced.

[0121] In a possible implementation manner, the method further includes:

[0122] When a high-priority wheeled robot exists in the target area, a bias value adjustment exemption flag is set for the high-priority wheeled robot, the base station associated with the high-priority wheeled robot is not involved in dynamic bias value reduction, and the high-priority wheeled robot includes a wheeled robot performing an emergency task or an authorized emergency communication wheeled robot.

[0123] For example, when the system detects that there is a special wheeled robot with high priority (such as an ambulance, a fire truck, a police car, or an emergency wheeled robot performing an emergency task, or an emergency communication support wheeled robot authorized by the competent department) in the target service area, the network control unit will automatically set an exemption flag for the bias value adjustment for these high-priority wheeled robots. The flag will trigger the following protection mechanism: the base station that establishes a communication connection with the high-priority wheeled robot will be excluded from the execution range of the dynamic bias value reduction strategy, and the service bias value will be maintained at the optimal level to ensure communication quality. This design ensures that the communication link of the key wheeled robot always enjoys the highest level of resource guarantee during base station load balancing, including: 1) The base station associated with the wheeled robot with the exemption flag is not involved in the dynamic bias value reduction process in the load balancing algorithm; 2) The system resource scheduler will reserve the necessary wireless resource margin for this type of base station; 3) The network management system will monitor the performance indicators of the exempted base station in real time, and will trigger the resource expansion mechanism in priority when detecting the risk of communication quality decline. The implementation of this priority strategy effectively solves the problem of communication quality fluctuations that emergency wheeled robots may face in traditional load balancing algorithms, and provides a reliable communication guarantee basis for the execution of emergency tasks.

[0124] In a possible implementation, when implemented, first, a digital certificate verification request is sent to the emergency management platform through the vehicle-mounted eSIM card, and after double authentication by the traffic management system and the emergency management database, a dynamic exemption token with a time stamp and a geographic fence is generated by the network slice manager. After the base station receives the registration request carrying the exemption token, it immediately starts a dedicated resource pool reservation program, and locks the MIMO channel parameters of the wheeled robot in the cell-level beamforming configuration, to ensure that the RSRP receiving strength can still be maintained in complex electromagnetic environments such as dense urban areas. When multiple exempted wheeled robots enter the same cell, the system will perform multi-dimensional weighted sorting based on the priority weight of the wheeled robot (defined according to the relevant national standard), the task urgency (parsed through the task code issued by the emergency command center), and the remaining range (obtained through V2X vehicle networking data), and implement periodical bias value fine-tuning for non-highest-priority wheeled robots. At the same time, through distributed base station cooperation technology, the service load of the exempted wheeled robot is intelligently migrated to the dedicated emergency frequency band of the adjacent cell to avoid overloading of the core network element.

[0125] It can be seen that in the embodiment of the application, by establishing the pyramid structure of service levels, it is ensured that the emergency communication wheeled robot always obtains the optimal network resources. By setting the bias value to adjust the exemption signaling marking mechanism, the exemption decision delay can be effectively reduced.

[0126] In a possible implementation, during the night low traffic period, the bias value of the macro base station can be reduced, part of the radio frequency channels can be closed, and the service can be migrated to the dormant wake-up small base station. After reducing the bias value of the macro base station, the wheeled robot is more inclined to access the small base station, and the coverage radius of the macro base station is contracted, which can effectively alleviate the co-frequency interference, especially in dense urban areas.

[0127] In a possible implementation, the setting of the bias value is related to the speed change rate. For example, when the wheeled robot enters a smooth road section from a congested road section, the speed of the wheeled robot can be increased. When the speed of the wheeled robot is increased, the bias value of the macro base station can be dynamically increased, the switching strategy is optimized in advance, and the signal quality is prevented from suddenly decreasing due to sudden acceleration. For another example, when the wheeled robot approaches a traffic light or a congested area, the traffic jam will gradually decrease, and in this process, the bias value of the macro base station can be gradually reduced, the wheeled robot is guided to be smoothly switched to the small base station, and unnecessary switching is avoided due to sudden deceleration.

[0128] In a possible implementation, in order to realize more accurate dynamic management of the bias value, the bias value change rate is considered. The bias value change rate refers to the maximum amplitude (such as ±3 dB / s) of the bias value allowed to be adjusted per unit time, which is used to control the response speed and stability of the load balancing strategy. The setting needs to comprehensively consider the mobility of the wheeled robot (such as a high change rate in a high-speed scene to adapt to rapid switching, and a low change rate in a congested scene to avoid oscillation), network load fluctuation (increasing the change rate to quickly balance in a burst traffic, and reducing the change rate to reduce signaling overhead in a stable state), service priority (emergency wheeled robots can break through the change rate limit), and wireless environment stability (low change rate is adopted to prevent misadjustment when the signal fluctuation is large). By dynamically adjusting the change rate, the network resource utilization and the switching success rate can be optimized while the QoS of key services is guaranteed, which is suitable for intelligent load balancing scenarios of 5G vehicle networking.

[0129] For example, scenario one: high-speed scenario, wheeled robot moves fast, switching demand is high, bias value change rate needs to be increased to adapt to the rapidly changing signal quality. Scenario two: in the congestion scenario of urban roads, the wheeled robot moves at low speed or is stationary, the change rate can be reduced to avoid unnecessary parameter fluctuations. Scenario three: the wheeled robot suddenly accelerates (such as entering a high-speed road), the change rate is temporarily increased to quickly optimize the switching strategy. Scenario four: the wheeled robot decelerates (such as approaching a traffic light), the change rate is gradually reduced to avoid ping-pong switching. Scenario five: sudden load surge (such as sudden traffic flow), the bias value needs to be quickly adjusted (high change rate) to alleviate the overload of the base station. Scenario six: stable load, low change rate is adopted to maintain network stability. Scenario seven: for high-priority services (such as emergency wheeled robots), bias value adjustment can be exempted from change rate restrictions to ensure immediate optimization of access quality. Scenario eight: if the switching failure rate is high in a certain area, the change rate is reduced, and the adjustment is cautious to avoid exacerbating the problem; if the switching success rate is high, the change rate can be increased to enhance optimization flexibility.

[0130] As shown in the following table, exemplary recommendations are given for the change rate of typical application scenarios:

[0131]

[0132] In one possible implementation, through the PC5 direct link, the wheeled robot shares the base station load and channel state information in real time, generates CIO optimization suggestions and feeds back to the base station, dynamically adjusts the bias value or bias value change rate after network arbitration, and finally forms a low-latency closed-loop load balancing system. As can be seen, by using the advantage of near-end communication, the delay of traditional network-side decision-making can be reduced, the load prediction accuracy can be improved, and the resource allocation efficiency in high-speed mobile scenarios can be significantly optimized.

[0133] In summary, in the embodiments of the present application, the change rate of the bias value is related to different service scenarios, and the adjustment range of the change rate of the bias value is different for different service scenarios. Conversely, the change rate of the load affects the setting of the initial bias value. The wheeled robot communication network dynamic load balancing method provided in the embodiments of the present application dynamically adjusts the bias values of the macro base station and the small base station, intelligently guides the balanced distribution of the wheeled robot terminal between the base stations in different scenarios, improves the load prediction accuracy, and significantly improves the network resource utilization.

[0134] Figure 2 The embodiments of the present application also provide a wheeled robot communication network dynamic load balancing device, which comprises:

[0135] The acquisition module 201 is configured to acquire an initial bias value of a base station in a target area, wherein the base station comprises a macro base station and a small base station, the initial bias value comprises a first bias value and a second bias value, the first bias value is associated with the macro base station, and the second bias value is associated with the small base station.

[0136] an identification module 202, configured to select an associated base station according to real-time communication channel conditions of the wheeled robot in the target area and the initial bias value, and collect load data, wheeled robot channel state and wheeled robot moving performance of the base station, and identify overloaded base stations and lightly loaded base stations;

[0137] a switching module 203, configured to select a first wheeled robot meeting preset migration conditions from the overloaded base stations, switch the first wheeled robot to the lightly loaded base station, and obtain a switching success rate, the switching success rate being used to dynamically adjust the first bias value and the second bias value.

[0138] In a possible implementation, the initial bias value includes a first bias value and a second bias value, and the first bias value is smaller than the second bias value.

[0139] In a possible implementation, the switching module 203 is further configured to:

[0140] in a case where the switching success rate is higher than a preset value, the first bias value is reduced and the second bias value is increased.

[0141] In a possible implementation, the switching module 203 is further configured to: select a first wheeled robot meeting preset migration conditions from the overloaded base stations, and switch the first wheeled robot to the lightly loaded base station, and the preset conditions include at least one of the following:

[0142] a wheeled robot whose post-migration target base station can meet QoS requirements, a wheeled robot whose moving speed is lower than a set speed threshold, and a wheeled robot whose channel conditions are better than a set threshold.

[0143] In a possible implementation, the apparatus further includes a prediction module, configured to:

[0144] predict a moving path of the wheeled robot and a set of potential target base stations for switching, pre-act the switching process in a digital twin environment, calculate a set of predicted QoS indicators, and virtually migrate the target wheeled robot to a candidate base station.

[0145] trigger an actual base station switching operation when all the predicted QoS indicators meet corresponding service requirement thresholds.

[0146] In a possible implementation, the apparatus further includes a rollback module, configured to:

[0147] when a QoS indicator after actual switching deviates from a predicted value by more than a set threshold, roll back to a connection with an original base station.

[0148] In a possible implementation, the apparatus further includes an adjusting module configured to:

[0149] When a high-priority wheeled robot exists in the target area, a bias value adjustment exemption flag is set for the high-priority wheeled robot, a base station associated with the high-priority wheeled robot does not participate in dynamic bias value reduction, and the high-priority wheeled robot includes a wheeled robot performing an emergency task or an authorized emergency communication wheeled robot.

[0150] In the embodiments of the present application, Figure 2 The apparatus shown is used to implement the method provided in any of the embodiments. Figure 1 The method provided in any of the corresponding embodiments.

[0151] Figure 3 A structural schematic diagram of an electronic device provided in the embodiments of the present application is shown in FIG. 3. Figure 3 The electronic device 300 includes a memory 310 and a processor 320.

[0152] The memory 310 stores a computer program executable by the at least one processor 320. The computer program is executed by the at least one processor 320 to enable the electronic device to implement the method provided in any of the above embodiments.

[0153] The memory 310 and the processor 320 can be connected through a bus 330.

[0154] The relevant description can be understood in correspondence with the relevant description and effects of the method embodiments, which will not be repeated here.

[0155] An embodiment of the present application provides a computer readable storage medium, which stores a computer program. The computer program is executed by a processor to implement the method provided in any of the embodiments. Figure 1 The method provided in any of the corresponding embodiments.

[0156] The computer readable storage medium can be a ROM, a random access memory (RAM), a CD-ROM, a magnetic tape, a floppy disk, and an optical data storage device, etc.

[0157] An embodiment of the present application provides a computer program product, which contains computer execution instructions. When the computer execution instructions are executed by a processor, the computer execution instructions are used to implement the method provided in any of the embodiments. Figure 1 The method provided in any of the corresponding embodiments.

[0158] In several embodiments provided in the present application, it should be understood that the disclosed apparatus and method can be implemented in other ways. For example, the apparatus embodiments described above are merely illustrative, and the division of modules is merely a logical function division. In actual implementation, another division manner can be used, for example, a plurality of modules or components can be combined or integrated into another system, or some features can be ignored or not executed. In addition, the coupling or direct coupling or communication connection between the displayed or discussed modules can be indirect coupling or communication connection through some interfaces, apparatuses or modules, and can be electrical, mechanical or other forms.

[0159] Other embodiments of the application will be apparent to those skilled in the art from consideration of the specification and practice of the application disclosed herein. It is intended that the specification and examples be considered as exemplary only, with the true scope of the application being indicated by the following claims.

[0160] It should be understood that the application is not limited to the precise construction that has been described above and shown in the accompanying drawings, and that various modifications and changes can be made by those skilled in the art without departing from the scope of the application.

Claims

1. A method for dynamic load balancing of a communication network of wheeled robots, characterized in that, The method comprises the following steps: acquiring initial bias values of base stations in a target area, the base stations comprising macro base stations and small base stations, the initial bias values comprising a first bias value and a second bias value, the first bias value being associated with the macro base stations, and the second bias value being associated with the small base stations; selecting associated base stations according to real-time communication channel conditions of wheeled robots in the target area and the initial bias values, collecting load data of the base stations, channel states of the wheeled robots, and moving performance of the wheeled robots, and identifying overloaded base stations and lightly loaded base stations; selecting a first wheeled robot meeting preset migration conditions from the overloaded base stations, switching the first wheeled robot to the lightly loaded base stations, and acquiring a switching success rate, the switching success rate being used for dynamically adjusting the first bias value and the second bias value.

2. The method of claim 1, wherein, The first bias value is smaller than the second bias value.

3. The method according to claim 1 or 2, characterized in that The switching success rate is used for dynamically adjusting the first bias value and the second bias value, comprising: in the case that the switching success rate is higher than a preset value, the first bias value is reduced, and the second bias value is increased.

4. The method of claim 1, wherein, The preset conditions comprise at least one of the following: the target base station after migration can meet the QoS requirements of the wheeled robots, the wheeled robots with a moving speed lower than a set speed threshold, and the wheeled robots with channel conditions better than a set threshold.

5. The method of claim 1, wherein, The method further comprises: predicting a moving path of the wheeled robots and a set of potential switching target base stations, pre-rehearsing a switching process in a digital twin environment, calculating a set of predicted QoS indicators, and virtually migrating the first wheeled robot to a candidate base station; when all the predicted QoS indicators meet corresponding business requirement thresholds, triggering an actual base station switching operation.

6. The method of claim 5, wherein, The method further comprises: when a QoS indicator after actual switching deviates from a predicted value by more than a set threshold, rolling back to original base station connection.

7. The method of claim 1, wherein, The method further comprises: when there is a high-priority wheeled robot in the target area, setting a bias value adjustment exemption marker for the high-priority wheeled robot, the base station associated with the high-priority wheeled robot not participating in dynamic bias value reduction, and the high-priority wheeled robot comprising a wheeled robot performing an emergency task or an authorized emergency communication wheeled robot.

8. A dynamic load balancing apparatus for a wheeled robot communication network, comprising: The method comprises the following steps: an acquisition module is configured to acquire initial bias values of base stations in a target area, the base stations comprising macro base stations and small base stations, the initial bias values comprising a first bias value and a second bias value, the first bias value being associated with the macro base stations, and the second bias value being associated with the small base stations; an identification module is configured to select associated base stations according to real-time communication channel conditions of wheeled robots in the target area and the initial bias values, collect load data of the base stations, channel states of the wheeled robots, and moving performance of the wheeled robots, and identify overloaded base stations and lightly loaded base stations; The switching module is configured to select a first wheeled robot meeting preset migration conditions from the overload base station, associate the first wheeled robot to the light-load base station, and obtain a switching success rate, the switching success rate being used to dynamically adjust the first bias value and the second bias value.

9. An electronic device, comprising: The electronic device comprises: at least one processor; and a memory connected to the at least one processor in communication; wherein the memory stores instructions executable by the at least one processor, and the instructions are executed by the at least one processor to cause the electronic device to perform the method according to any one of claims 1 to 7.

10. A computer-readable storage medium having stored thereon a computer program, characterized in that, The computer program is executed by the processor to implement the method according to any one of claims 1 to 7.

Images