Mine 5G signal coverage optimization method and system

Abstract

The invention provides a mine 5G signal coverage optimization method and system, and the method specifically comprises the steps: calculating a space-time channel prediction factor of each service terminal in a future scheduling period through employing a ray tracing algorithm based on a mine three-dimensional model, the real-time position of equipment and a predetermined trajectory; constructing a multi-dimensional service state vector comprising a service type, a security state and a process priority, allocating hard slice resources to the key service or the terminal with poor channel quality by combining a predictive factor, and allocating soft slice resources to the rest; setting a resource utilization rate high-low water level line for the hard slice terminal, adjusting resources and maintaining a shared resource pool; and calculating the scheduling weight of the soft slice terminal according to the predictive factor, the capacity of the shared pool and the number of the terminals, and deciding to obtain the priority and the share of the temporary resource.

Description

TECHNICAL FIELD

[0001] The application belongs to the field of communication, and particularly relates to a mine 5G signal coverage optimization method and system. BACKGROUND

[0002] 5G communication technology supports intelligent application of mines, however, the special environment in mines, irregular roadway structure, large metal mobile equipment and rock wall material factors lead to variable wireless signal propagation path. The 5G network resource scheduling strategy based on the open environment planning on the ground is difficult to be applied to the underground scene, and often causes key business interruption or service quality decline due to the inability to adapt to channel deterioration in real time.

[0003] The mine 5G network resource scheduling method is mostly reactive, that is, resource allocation is performed according to the history or current information of the channel quality indicator CQI reported by the terminal. For the high-speed mobile equipment in the roadway, the lagging scheduling method cannot predict the drastic change of the channel, and is easy to cause connection interruption and data packet loss. The granularity of business priority differentiation is relatively coarse, and the QoS mechanism is difficult to meet the actual needs of mine production, and cannot associate the business request with the specific production process link and the business context of the change of the safety state of the equipment, so that when the resources are scarce, it is difficult to ensure that the highest priority business such as gas monitoring and emergency braking is absolutely guaranteed. The existing network slice resource allocation method is relatively static, once the resources are divided for a certain business, the resources are difficult to release and share even if the business is in an idle or low load state, and when the burst traffic exceeds the preset capacity, it cannot be supplemented in time, resulting in that the overall utilization rate of network resources is not high, and it is difficult to achieve the global optimal balance. SUMMARY

[0004] The application provides a mine 5G signal coverage optimization method, which is used to solve the problems that the existing scheduling method cannot predict the change of the channel, is easy to cause connection interruption and data packet loss, and the existing mechanism is difficult to meet the actual needs of mine production, and comprises the following steps:

[0005] Obtaining a three-dimensional model of a mine, real-time positions and predetermined trajectories of underground mobile equipment, and business requests of each business terminal; based on the three-dimensional model, real-time positions and predetermined trajectories of the equipment, using a ray tracing algorithm, calculating a space-time channel prediction factor of each business terminal and a base station in a future scheduling period;

[0006] Constructing a multi-dimensional business state vector comprising a business type, a safety state of the equipment and a priority of the production process; comprehensively considering the business state vector and the space-time channel prediction factor, allocating hard slice resources to the business terminal meeting a preset key business condition or the predicted channel quality being lower than a predetermined threshold, and allocating soft slice resources to the remaining business terminals;

[0007] For the service terminal allocated with the hard slice resource, high and low water lines of resource utilization are set according to the service quality requirement of the terminal; when the real-time utilization is lower than the low water line, the difference resource is released to a shared resource pool; when the real-time utilization is higher than the high water line, the terminal is allocated with supplementary resource from the shared resource pool.

[0008] For the service terminal allocated with the soft slice resource, a scheduling weight is calculated, which determines the priority and share of the terminal for obtaining temporary resource from the shared resource pool.

[0009] In addition, the application also relates to a mine 5G signal coverage optimization system, comprising the following modules:

[0010] The first calculation module is used for acquiring a mine three-dimensional model, real-time positions and predetermined trajectories of underground mobile devices and service requests of each service terminal; based on the three-dimensional model, real-time positions and predetermined trajectories of the devices, a ray tracing algorithm is used to calculate space-time channel prediction factors of each service terminal and a base station in a future scheduling period;

[0011] The first allocation module is used for constructing a multi-dimensional service state vector comprising service types, device safety states and production process priorities; the service state vector and the space-time channel prediction factors are comprehensively used to allocate hard slice resources to service terminals meeting preset key service conditions or having predicted channel quality lower than a predetermined threshold, and to allocate soft slice resources to the rest of the service terminals;

[0012] The second allocation module is used for, for the service terminal allocated with the hard slice resource, setting high and low water lines of resource utilization according to the service quality requirement of the terminal; when the real-time utilization is lower than the low water line, the difference resource is released to a shared resource pool; when the real-time utilization is higher than the high water line, the terminal is allocated with supplementary resource from the shared resource pool;

[0013] The second calculation module is used for, for the service terminal allocated with the soft slice resource, calculating a scheduling weight, which determines the priority and share of the terminal for obtaining temporary resource from the shared resource pool.

[0014] The application can predict future channel quality by combining the mine three-dimensional model and the device moving track, can configure resources for the mobile terminal about to enter the weak signal area in advance, thereby enhancing the stability and reliability of the underground key business communication link, and avoiding data interruption caused by channel mutation. At the same time, the business request is associated with the device safety state and the production process priority, a business guarantee system is constructed, and when the resources are limited, the business demand such as safety monitoring, emergency control highest level can be preferentially met, and the essential safety level of mine operation is improved. In addition, by setting up a shared resource pool and managing the threshold of hard slice resource utilization, the flexible flow of guarantee resources and shared resources is realized, which can not only meet the burst traffic demand of key businesses, but also release idle resources to other businesses, thereby improving the spectrum resource utilization efficiency of the whole 5G network and realizing the management of resources. BRIEF DESCRIPTION OF DRAWINGS

[0015] Figure 1 is a flowchart of the first embodiment;

[0016] Figure 2 is a schematic diagram of hard slice and soft slice allocation logic;

[0017] Figure 3 is a schematic diagram of soft slice terminal scheduling weight calculation. DETAILED DESCRIPTION

[0018] To make the purposes, technical solutions, and advantages of the embodiments of the present application clearer, the technical solutions in the embodiments of the present application will be described clearly and completely below with reference to the drawings in the embodiments of the present application. Obviously, the described embodiments are some embodiments of the present application, rather than all the embodiments of the present application. Based on the embodiments in the present application, all other embodiments obtained by those skilled in the art without creative labor fall within the scope of protection of the present application.

[0019] The plurality involved in the present application refers to two or more. In addition, it should be understood that in the description of the present application, the words “first”, “second”, etc. are only used for the purpose of distinguishing the description, and cannot be understood as indicating or implying relative importance, nor can it be understood as indicating or implying order.

[0020] In the first embodiment, the present application proposes a mine 5G signal coverage optimization method, as shown in Figure 1 , comprising:

[0021] S1, obtaining a mine three-dimensional model, real-time positions and predetermined trajectories of underground mobile devices, and business requests of each business terminal; based on the three-dimensional model, real-time positions and predetermined trajectories of the devices, using a ray tracing algorithm, calculating the space-time channel prediction factor of each business terminal and base station in a future scheduling period;

[0022] The mine three-dimensional model is provided by the mine design and exploration department, containing the geometry of roadway, chamber, working face and the material information of rock wall and metal support, and is pre-stored in the network control server. The real-time position of the underground mobile device is obtained by the ultra-wideband (UWB) base station deployed in the roadway or by using the 5G positioning technology, and the device reports the three-dimensional coordinates to the network every second. The predetermined trajectory is obtained from the mine production scheduling system, such as the planned path information of the tunneling machine or the unmanned electric car. The service request of each service terminal is reported to the 5G base station through a signaling message when initiating connection, and the message contains the required bandwidth, maximum delay, and service type identifier such as high-definition video, remote control, or data acquisition.

[0023] The network controller takes the base station antenna position as the transmission point and takes the multiple predicted position points of the mobile device along the predetermined trajectory within a future scheduling period, for example, 500 ms, as the receiving point. Based on the wall and device obstacle information in the three-dimensional model, the transmission, direct, reflection, and diffraction paths of the electromagnetic wave rays are simulated. By calculating the vector sum of the signal strength, phase, and delay of multiple ray paths reaching each predicted position point, the predicted path loss, signal-to-noise ratio, and Rician K-factor channel parameters of the position point are obtained. The parameters jointly constitute the space-time channel prediction factor of the terminal within the future time period.

[0024] In an optional embodiment, the space-time channel prediction factor of each service terminal and the base station within a future scheduling period is calculated by using a ray tracing algorithm, which includes:

[0025] The mine three-dimensional model is discretized into a set of triangular facets; for the future position of each service terminal, multiple rays are transmitted to simulate and calculate the propagation loss of the direct path, first-order reflection path, and first-order diffraction path of the rays; the channel gain of the position is obtained by synthesizing the loss of all paths, and the predicted reference signal received power (RSRP) and predicted signal-to-noise ratio (SINR) are calculated as the space-time channel prediction factor in combination with the base station transmission power.

[0026] Specifically, the mine three-dimensional model is obtained, which contains the geometry of the roadway, the rock wall material, and the position information of large equipment. To perform electromagnetic wave propagation simulation, the continuous three-dimensional model is discretized into a set of triangular facets. For example, a roadway model with a length of 200 meters and a width of 5 meters can be discretized into more than ten thousand triangular facets, and each facet is assigned specific electromagnetic parameters such as dielectric constant and conductivity according to the represented material such as coal, rock, or metal support.

[0027] For a specific business terminal, such as a controller on a coal mining machine, the three-dimensional spatial position of the controller after the next scheduling period, for example, 20 ms, is predicted. Starting from the base station antenna, a large number of rays are emitted in the direction of the predicted position, for example, one thousand rays. The ray tracing algorithm calculates the propagation path of each ray, mainly considering three path types: a direct path without obstruction; a path that reflects once via a roadway wall or device surface; and a path that diffracts once around a corner or obstacle edge. For each calculated path, the propagation loss is calculated according to the path length and the interaction with the environment, for example, the direct path loss is mainly based on the free space propagation model, the reflection path loss additionally accounts for the energy loss caused by the Fresnel reflection coefficient, and the diffraction path uses a uniform diffraction theory model to calculate the diffraction loss. The signal energy of all paths reaching the predicted position is vector superimposed to obtain the total channel gain at that position. More specifically, for each propagation path obtained by ray tracing, its path loss, including free space loss, reflection loss and diffraction loss, is calculated and converted into path gain in the corresponding linear scale. The linear gains of all effective propagation paths at the same position are superimposed to obtain the comprehensive channel gain at that position.

[0028] Based on the channel gain, combined with the base station transmit power, for example, 20 dBm, the predicted reference signal received power RSRP can be calculated. By estimating the interference and background noise level from other cells, the predicted signal-to-noise ratio SINR can be calculated. For example, if the predicted RSRP is -85 dBm and the total interference and noise power is -95 dBm, the predicted SINR is 10 dB. The two values, predicted RSRP and predicted SINR, together constitute the spatio-temporal channel prediction factor of the terminal at the future time. In one embodiment, based on the comprehensive channel gain and the base station transmit power, the predicted RSRP is calculated in the form of RSRP = transmit power + antenna gain + channel gain, and further combined with the predicted interference power and noise power, the corresponding predicted SINR is calculated by SINR = signal power / (interference power + noise power).

[0029] S2, construct a multi-dimensional business state vector containing business type, device security state and production process priority; integrate the business state vector and the spatio-temporal channel prediction factor to allocate hard slicing resources to business terminals that meet preset key business conditions or have predicted channel quality below a predetermined threshold, and allocate soft slicing resources to the remaining business terminals;

[0030] A state vector is generated for each service terminal, such as [service type ID, security state code, process priority], wherein the service type ID is 1 representing remote driving, the security state code is 0 representing normal, and the process priority is 5 representing the highest level. The device security state is reported in real time by the sensors such as the gas sensor, and when the gas concentration exceeds the standard, the security state code becomes 2 representing emergency. The production process priority is issued by the production scheduling system according to the current task. The allocation rule is set, for example, the terminal with the security state code of 2 or the remote coal mining machine control terminal with the process priority of 5 are all identified as meeting the key service condition, and the hard slice resource is allocated. At the same time, for other terminals, if the time-space channel prediction factor shows that the future signal-to-noise ratio will be lower than 5 decibels, the hard slice resource is also allocated to the terminal to ensure the connection. The terminal that does not meet the above conditions is allocated to the soft slice. The hard slice resource refers to the exclusive network resource reserved for a specific service with strong isolation, for example, in order to guarantee the millisecond-level delay and 99.999% reliability requirement of the business such as remote control of the heading machine or unmanned mine car, a guaranteed bit rate GBR bearer with a specific QoS flow identifier QFI is allocated, and a group of dedicated physical resource blocks PRB are reserved on the wireless side, which are mainly allocated to key services such as remote control in the application. The soft slice resource refers to the network resource logically divided and shared by priority scheduling, for example, the data acquisition of environmental monitoring sensors and the business terminals of ordinary area high-definition video monitoring are divided into the soft slice, they compete and use the shared resource pool released by the hard slice when the utilization rate is low, and no fixed resource is reserved for them, but the priority of each terminal is calculated in real time according to the scheduling weight formula proposed in the application. The terminal with good predicted channel quality and high business importance will obtain higher scheduling weight, thereby obtaining an advantage in resource competition, and providing a flexible on-demand allocation service with quality of service QoS differentiation for these businesses.

[0031] In an optional embodiment, the multi-dimensional service state vector containing service type, device security state and production process priority is constructed, including:

[0032] The service type is quantified, wherein the remote control type is assigned a value of 10, the high-definition video monitoring type is assigned a value of 8, and the data acquisition type is assigned a value of 5; the device security state is quantified, wherein the alarm state is assigned a value of 1 and the normal state is assigned a value of 0; and the production process priority is quantified, wherein the terminal related to the coal mining working face is assigned a value of 3, the terminal related to the heading working face is assigned a value of 2, and the terminal related to the main transportation roadway is assigned a value of 1.

[0033] In order to manage and schedule resources for various business terminals in the mine, a standardized multi-dimensional business state vector is created for each terminal. The vector contains three dimensions, representing business type, equipment safety state and production process priority. The vector can be expressed as V=[T, S, P], where T represents the quantified value of business type, S represents the quantified value of equipment safety state, and P represents the quantified value of production process priority. The structured data model allows the business needs and importance of different terminals to be uniformly compared and calculated.

[0034] The quantified value of each dimension is based on the degree of influence of the dimension on the safety and production of the mine. In the business type dimension, remote control businesses that require extremely high real-time performance and any interruption can cause production accidents, such as remote control of roadheaders, are assigned the highest value of 10. High-definition video businesses that require stable large bandwidth for safety monitoring are assigned the next highest value of 8. Data acquisition businesses that tolerate a certain amount of delay and have relatively low importance, such as regular data reporting by gas sensors, are assigned a value of 5. In the equipment safety state dimension, a binary indication system is used. When the equipment itself or the environment monitored by the equipment enters an alarm state, such as high hydraulic pressure of a roadheader or excessive detection by a gas sensor, the state value is set to 1, indicating that immediate attention is required. In the normal working state, the value is 0.

[0035] In the production process priority dimension, the value is related to the physical location of the terminal and the core degree of the location in the production link. Terminals located on the coal mining face, such as the control terminals of coal mining machines and hydraulic supports, are the core of production, so they are assigned the highest priority of 3. Terminals located on the heading face, such as roadheader and anchor drill vehicle terminals, are less important and are assigned a priority of 2. Terminals located in the auxiliary area of the main transportation roadway have relatively low priority and are assigned a priority of 1. For example, a remote control terminal of a coal mining machine in an alarm state has a multi-dimensional business state vector of [10, 1, 3], while a normal working main transportation roadway monitoring camera has a vector of [8, 0, 1].

[0036] In an optional embodiment, the hard slice resources are allocated to the business terminal that meets the preset critical business condition or the predicted channel quality is lower than the predetermined threshold, including:

[0037] Based on the multi-dimensional business state vector, a comprehensive business priority score is calculated. When the comprehensive business priority score is higher than a first preset threshold, it is determined that the preset critical business condition is met. In addition, it is determined whether the predicted signal-to-noise ratio in the spatio-temporal channel prediction factor is lower than a second preset threshold for a plurality of consecutive prediction time points. If so, it is determined that the predicted channel quality is lower than the predetermined threshold.

[0038] In particular, two types of decision criteria are used to identify terminals that need to be allocated hard slice resources. The first type is based on the criticality of the service itself. A comprehensive service priority score is calculated using a multi-dimensional service state vector V = [T, S, P]. In one embodiment, the quantized values of the dimensions in the multi-dimensional service state vector are used as features, and pre-defined weights are assigned to the service type, safety status and production process priority, respectively. The features are then summed up with the weights to obtain a single scalar form of the comprehensive service priority score, where safety alarm status and high priority production process are given higher weights in the weighting process to ensure the priority protection of critical services and service terminals in high-risk scenarios. The calculation formula can be a weighted sum, for example, Score = 0.5 x T + 0.3 x S + 0.2 x P.

[0039] A first threshold value is pre-set, for example, 5.5. Suppose a remote control terminal is located at the coal face, and the control terminal associated equipment sends an alarm, the state vector is [10, 1, 3], and the comprehensive score calculated is 5.9. Since 5.9 is higher than the first pre-set threshold value 5.5, the terminal is determined to meet the pre-set critical service condition, and therefore is eligible for hard slice resources.

[0040] The second type of decision criterion is based on the predicted severity of the communication environment. The aforementioned spatio-temporal channel prediction factors, in particular the predicted signal-to-noise ratio (SINR), are used to assess the future communication quality of the terminal. A second pre-set threshold value is set, for example, 3 dB, and a continuous time window is defined, for example, 5 consecutive scheduling periods. For a certain terminal, if the predicted SINR values of the terminal in the next 5 consecutive scheduling periods are 2.5 dB, 2.1 dB, 2.8 dB, 1.9 dB, and 2.3 dB, respectively, all of which are lower than the threshold value of 3 dB, then it is determined that the predicted channel quality of the terminal is lower than the pre-defined threshold value. This situation usually occurs when the terminal is about to move to a signal coverage blind area or a strong interference area.

[0041] A terminal is marked as a hard slice allocation object as long as it meets any one of the above two conditions. For example, even if the comprehensive service priority score of a high-definition video terminal does not reach the first threshold value, but if it is predicted to enter a signal-poor tunnel corner, resulting in a continuous predicted signal-to-noise ratio (SINR) below the second threshold value, it will still be allocated to a hard slice to combat the poor channel environment through the isolation and protection characteristics of the hard slice resources, ensuring service continuity. The dual decision mechanism ensures that whether the service itself is extremely important or the communication environment is extremely poor, the terminal can obtain the most reliable network service protection, such as Figure 2 .

[0042] S3, for a service terminal to which a hard slice resource is allocated, setting a high and low water line of resource utilization according to a quality of service requirement of the terminal; when real-time utilization is lower than the low water line, releasing a difference resource to a shared resource pool; when real-time utilization is higher than the high water line, allocating a supplementary resource from the shared resource pool to the terminal;

[0043] Suppose a high-definition video monitoring service with a quality of service requirement of an average rate of 20 Mbps and a peak rate of 30 Mbps, and the network allocates a physical resource block of 25 Mbps for hard slicing of the service. The low water line is set to 20 Mbps and the high water line is set to 24 Mbps. The network controller periodically monitors the real-time data throughput rate of the service. When the monitoring picture is static and the real-time rate drops to 15 Mbps, the rate is lower than the low water line, and the controller calculates a difference resource of 10 Mbps between 25 Mbps and 15 Mbps, and releases the corresponding physical resource block to the shared resource pool. When a moving device appears in the picture, causing the code rate to surge to 28 Mbps, the rate is higher than the high water line, and the controller immediately applies for at least 4 Mbps of supplementary resource from the shared resource pool to be temporarily allocated to the terminal to prevent video stuttering.

[0044] In an optional embodiment, the setting of the high and low water line of resource utilization according to the quality of service requirement of the terminal comprises:

[0045] For a remote control service with a guaranteed bit rate GBR of 50 Mbps, the low water line is set to 60% of the rate and the high water line is set to 95% of the rate; and a monitoring period of 100 ms is set to statistically average the real-time utilization in the period.

[0046] In order to realize the elastic scaling of soft slice resources, an adjustment threshold of resource utilization, i.e. a high and low water line, is set for each soft slice according to the quality of service requirement of the slice bearer service. The water line defines a healthy interval of resource utilization. Taking a soft slice carrying a remote control service as an example, the guaranteed bit rate GBR requirement of the service is 50 Mbps. The administrator can configure the low water line to be 60% of GBR, i.e. 30 Mbps, and the high water line to be 95% of GBR, i.e. 47.5 Mbps. The low water line is set to identify resource redundancy, and the high water line is set to warn of resource shortage.

[0047] After setting the water line, the resource utilization of the soft slice will be continuously tracked in a monitoring period. The monitoring period needs to be long enough to smooth out transient service fluctuations and short enough to quickly respond to service changes, for example, 100 ms. In each 100 ms monitoring period, the network monitoring module collects multiple instantaneous rate samples. For example, one sample is collected every 10 ms, and a total of 10 samples are collected in a period.

[0048] At the end of each monitoring period, the average value of the samples is calculated to obtain the average real-time utilization in the period. For example, in a 100ms period, the 10 rate samples collected are 40, 42, 38, 41, 45, 43, 46, 44, 39, 42 Mbps, and the average real-time utilization calculated is 42 Mbps. The calculated average value will be compared with the preset low water line of 30 Mbps and the high water line of 47.5 Mbps to determine whether the current soft slice resource allocation is excessive, appropriate or tight.

[0049] In an optional embodiment, when the real-time utilization is lower than the low water line, the excess resources are released to a shared resource pool, comprising:

[0050] Comparing the average real-time utilization of the service terminal in the monitoring period with the corresponding low water line;

[0051] When the average real-time utilization is less than the low water line, the difference between the low water line and the average real-time utilization is calculated as the releasable resource amount;

[0052] And the releasable resource amount corresponding to the wireless air interface resource unit is released to the shared resource pool.

[0053] Specifically, at the end of each monitoring period, the calculated average real-time utilization of the soft slice is compared with the preset low water line. Taking a soft slice with a guaranteed bit rate of 50 Mbps and a low water line of 30 Mbps as an example, suppose that in the just passed 100ms monitoring period, the average real-time utilization is only 20 Mbps. Since 20 Mbps is lower than the low water line of 30 Mbps, the resource release process is triggered.

[0054] After triggering, the releasable excess resource amount is calculated. The excess is not the difference between the actual utilization and the guaranteed rate, but the difference from the low water line, to avoid too frequent resource adjustment. In this example, the excess resource is 10 Mbps. The 10 Mbps bandwidth is considered as redundant guaranteed resources that the soft slice temporarily does not need under the current business load.

[0055] The 10Mbps logical bandwidth resource is converted into physical resource. In 5G network, the basic unit of physical resource is physical resource block (PRB). The conversion relationship between bandwidth and PRB quantity is based on modulation and coding strategy (MCS) corresponding to current channel quality. Assuming that each PRB can carry a data rate of 0.5Mbps under current channel condition, the 10Mbps excess resource corresponds to 20 PRBs. The scheduler executes the instruction to remove the 20 PRBs from the exclusive resource configuration of the soft slice and returns them to a global shared resource pool. The resources in the shared resource pool can be allocated to other services with burst demand, thereby realizing optimization and flow of network resources.

[0056] S4, for the service terminal allocated with the soft slice resource, a scheduling weight is calculated, which determines the priority and share of the terminal in obtaining temporary resource from the shared resource pool.

[0057] At the beginning of each scheduling period, a weight value is calculated for each terminal in the soft slice. The weight value is calculated by a weighted formula, for example, weight = 0.5 × predicted signal-to-noise ratio normalized value + 0.3 × service priority normalized value - 0.2 × normalized value of the number of active terminals in the current slice. The higher the predicted signal-to-noise ratio, the better the channel condition, the higher the resource utilization efficiency, and the greater the weight. The higher the service priority of the terminal, the greater the weight. At the same time, the more the competing terminals in the slice, the weight is correspondingly reduced to achieve fairness. When there is available resource in the shared resource pool, the network scheduler allocates resources in the order of the weight values of the terminals from high to low. The higher the weight of the terminal, the higher the priority in obtaining resources, and the share of the allocated resources is also proportional to the weight value. In a possible calculation of the scheduling weight, a basic weight B(t) representing the degree of the terminal's own value of allocated resources is calculated according to the service priority score S(t) of the terminal and the space-time channel prediction factor such as predicted signal-to-noise ratio A system load factor M(t) reflecting the current resource supply and demand relationship is calculated, which is proportional to the available capacity of the shared resource pool and inversely proportional to the total number of terminals N(t) in the soft slice, for example Finally, the final scheduling weight W(t) of each terminal at the scheduling moment is the product of the two, i.e. W(t) = B(t) × M(t).

[0058] In an alternative embodiment, the calculation of the scheduling weight comprises:

[0059] The comprehensive service priority score S(t) of each terminal is calculated based on the multi-dimensional service state vector. The scheduling weight W(t) of each terminal at the scheduling moment t is calculated using the following formula: ​

[0060]

[0061] where W(t) is the scheduling weight of the terminal at scheduling time t, S(t) is the integrated service priority score of the terminal, is the average integrated service priority score of all terminals in the soft slice, is the predicted signal-to-noise ratio of the terminal, is the average predicted signal-to-noise ratio of all terminals in the soft slice, and is the weight coefficient and .

[0062] Specifically, at each scheduling time t, for example, every millisecond, the scheduler needs to calculate a scheduling weight for all terminals in the soft slice to determine the priority of resource allocation. The calculation process needs to prepare two key data: service priority and channel quality. The integrated service priority score S(t) is calculated according to the latest multi-dimensional service state vector of each terminal. At the same time, the predicted signal-to-noise ratio .

[0063] Suppose there are three terminals A, B, and C in a soft slice, and their integrated service priority scores S(t) are 6.0, 4.0, and 5.0, respectively, then the average score is 5.0. Their predicted signal-to-noise ratios are 15 dB, 20 dB, and 10 dB, respectively, so the average predicted signal-to-noise ratio is 15 dB. Set the weight coefficients and to balance the importance of service priority and channel quality, for example, set to 0.7, to 0.3, indicating that high priority services are emphasized.

[0064] For terminal A, the weight . For terminal B, the weight . For terminal C, the weight . After calculation, the scheduler sorts according to the weight value, in this example, terminal A has the highest weight and will get the highest resource allocation priority, followed by terminal B, and finally terminal C. The normalized weighted calculation method ensures that the scheduling decision can consider both service importance and channel opportunity, achieving a balance between quality of service and throughput, as Figure 3 .

[0065] In a second embodiment, the present application also provides a mine 5G signal coverage optimization system, comprising the following modules:

[0066] The first calculation module is configured to acquire a mine three-dimensional model, real-time positions of underground mobile devices, predetermined trajectories of the underground mobile devices, and service requests of each service terminal; based on the three-dimensional model, the real-time positions of the devices, and the predetermined trajectories, a ray tracing algorithm is used to calculate space-time channel prediction factors of each service terminal and a base station in a future scheduling period;

[0067] The first allocation module is configured to construct a multi-dimensional service state vector containing a service type, a device safety state, and a production process priority; and based on the service state vector and the space-time channel prediction factors, hard slice resources are allocated to service terminals that meet preset key service conditions or have predicted channel quality lower than a predetermined threshold, and soft slice resources are allocated to the remaining service terminals;

[0068] The second allocation module is configured to, for a service terminal that has been allocated hard slice resources, set high and low water lines of resource utilization according to service quality requirements of the terminal; when real-time utilization is lower than the low water line, the difference in resources is released to a shared resource pool; and when real-time utilization is higher than the high water line, additional resources are allocated to the terminal from the shared resource pool;

[0069] The second calculation module is configured to, for a service terminal that has been allocated soft slice resources, calculate a scheduling weight according to the space-time channel prediction factors, available capacity of the shared resource pool, and a number of terminals in the soft slice, the scheduling weight determining a priority and a share of the terminal in obtaining temporary resources from the shared resource pool.

[0070] In an optional embodiment, the ray tracing algorithm is used to calculate the space-time channel prediction factors of each service terminal and the base station in the future scheduling period, including:

[0071] The mine three-dimensional model is discretized into a set of triangular facets; for a future position of each service terminal, a plurality of rays are emitted to simulate and calculate propagation losses of direct paths, first-order reflection paths, and first-order diffraction paths of the rays; the losses of all paths are integrated to obtain a channel gain of the position, and a predicted reference signal received power (RSRP) and a predicted signal-to-noise ratio (SINR) are calculated as the space-time channel prediction factors in combination with a base station transmission power.

[0072] In an optional embodiment, the multi-dimensional service state vector containing the service type, the device safety state, and the production process priority is constructed, including:

[0073] The service type is quantified, wherein the remote control type is assigned a value of 10, the high-definition video monitoring type is assigned a value of 8, and the data acquisition type is assigned a value of 5; the device security state is quantified, wherein the alarm state is assigned a value of 1 and the normal state is assigned a value of 0; and the production process priority is quantified, wherein the terminal related to the coal mining operation face is assigned a value of 3, the terminal related to the tunneling operation face is assigned a value of 2, and the terminal related to the main haulage roadway is assigned a value of 1.

[0074] In an optional embodiment, the allocating hard slice resources to the service terminal that meets the preset key service condition or the predicted channel quality is lower than the predetermined threshold comprises:

[0075] The comprehensive service priority score is calculated based on the multi-dimensional service state vector, when the comprehensive service priority score is higher than a first preset threshold, it is determined that the preset key service condition is met, and whether the predicted signal-to-noise ratio in the space-time channel prediction factor is lower than a second preset threshold at a plurality of continuous prediction time points is determined, and if yes, it is determined that the predicted channel quality is lower than the predetermined threshold.

[0076] In an optional embodiment, the setting the high and low water lines of the resource utilization according to the terminal service quality requirement comprises:

[0077] For the remote control service with a guaranteed bit rate (GBR) of 50 Mbps, the low water line is set to 60% of the rate and the high water line is set to 95% of the rate, and the average real-time utilization in a monitoring period of 100 ms is counted.

[0078] In an optional embodiment, the releasing the excess resources to a shared resource pool when the real-time utilization is lower than the low water line comprises:

[0079] The average real-time utilization of the service terminal in the monitoring period is compared with the corresponding low water line;

[0080] When the average real-time utilization is less than the low water line, the difference between the low water line and the average real-time utilization is calculated as the releasable resource amount;

[0081] and the corresponding wireless air interface resource unit of the releasable resource amount is released to the shared resource pool.

[0082] In an optional embodiment, the calculating the scheduling weight comprises:

[0083] The comprehensive service priority score S(t) of each terminal is calculated based on the multi-dimensional service state vector, and the scheduling weight W(t) of each terminal at the scheduling time t is calculated using the following formula:

[0084]

[0085] wherein W(t) is the scheduling weight of the terminal at the scheduling time t, S(t) is the integrated service priority score of the terminal, is the average integrated service priority score of all terminals in the soft slice, is the predicted signal-to-noise ratio of the terminal, is the average predicted signal-to-noise ratio of all terminals in the soft slice, and is a weight coefficient and .

[0086] It should be noted that, for the foregoing method embodiments, for the sake of simple description, they are all expressed as a series of action combinations, but those skilled in the art should know that the embodiments of the present specification are not limited by the action sequence described, because according to the embodiments of the present specification, certain steps can be performed in other sequences or simultaneously. Secondly, those skilled in the art should know that the embodiments described in the specification are all preferred embodiments, and the actions and modules involved are not necessarily essential to the embodiments of the present specification.

[0087] In the above embodiments, the description of each embodiment has its own emphasis, and the parts not described in detail in a certain embodiment can be referred to the related description of other embodiments.

[0088] The preferred embodiments of the present specification disclosed above are only used to help explain the present specification. The alternative embodiments do not describe all the details and do not limit the invention to only the specific embodiments described. Obviously, according to the content of the embodiments of the present specification, many modifications and changes can be made. The present specification selects and specifically describes these embodiments in order to better explain the principles and practical applications of the embodiments of the present specification, so that those skilled in the art can well understand and use the present specification. The present specification is limited only by the claims and their entire scope and equivalents.

Claims

1. A method for optimizing 5G signal coverage in mines, characterized in that, Includes the following steps: Acquire the 3D model of the mine, the real-time location and planned trajectory of underground mobile equipment, and the business requests of each business terminal; Based on the three-dimensional model, the real-time location of the device and the predetermined trajectory, the ray tracing algorithm is used to calculate the spatiotemporal channel prediction factor of each service terminal and base station in the future scheduling cycle. Construct a multi-dimensional service state vector that includes service type, equipment security status and production process priority; combine the service state vector with the spatiotemporal channel prediction factor to allocate hard slice resources to service terminals that meet preset key service conditions or whose predicted channel quality is lower than a predetermined threshold, and allocate soft slice resources to other service terminals. For service terminals that have been allocated hard slice resources, high and low watermarks for resource utilization are set according to the terminal service quality requirements; when the real-time utilization rate is lower than the low watermark, the difference in resources is released to a shared resource pool; when the real-time utilization rate is higher than the high watermark, supplementary resources are allocated to the terminal from the shared resource pool. For a service terminal that has been allocated soft slice resources, a scheduling weight is calculated. The scheduling weight determines the priority and share of the terminal in obtaining temporary resources from the shared resource pool.

2. The method according to claim 1, characterized in that, The method employs a ray tracing algorithm to calculate the spatiotemporal channel prediction factors for each service terminal and base station within a future scheduling cycle, including: The three-dimensional model of the mine is discretized into a set of triangular facets. For the future location of each service terminal, multiple rays are emitted, and the propagation loss of the direct path, first-order reflection path, and first-order diffraction path of the rays is simulated and calculated. The channel gain at the location is obtained by combining the loss of all paths, and the predicted reference signal received power RSRP and predicted signal-to-noise ratio SINR are calculated in combination with the base station transmit power as the spatiotemporal channel prediction factors.

3. The method according to claim 1, characterized in that, The construction of a multi-dimensional business state vector, which includes business type, equipment safety status, and production process priority, includes: The business types are quantified, with remote control assigned a value of 10, high-definition video surveillance assigned a value of 8, and data acquisition assigned a value of 5; the equipment safety status is quantified, with alarm status assigned a value of 1 and normal status assigned a value of 0; the production process priority is quantified, with terminals related to coal mining faces assigned a value of 3, terminals related to tunneling faces assigned a value of 2, and terminals related to main transport roadways assigned a value of 1.

4. The method according to claim 3, characterized in that, The allocation of hard slice resources to service terminals that meet preset key service conditions or whose predicted channel quality is below a predetermined threshold includes: A comprehensive service priority score is calculated based on the multidimensional service state vector. When the comprehensive service priority score is higher than the first preset threshold, it is determined that the preset key service conditions are met. Also, it is determined whether the predicted signal-to-noise ratio in the spatiotemporal channel prediction factor is lower than the second preset threshold at multiple consecutive prediction time points. If so, it is determined that the predicted channel quality is lower than the predetermined threshold.

5. The method according to claim 1, characterized in that, Setting high and low water level thresholds for resource utilization based on the terminal service quality requirements includes: For remote control services with a guaranteed bit rate (GBR) of 50Mbps, a low watermark is set at 60% of the rate and a high watermark at 95% of the rate; and a monitoring period of 100ms is used to calculate the average real-time utilization rate within the period.

6. The method according to claim 1, characterized in that, The step of releasing the difference in resources to a shared resource pool when the real-time utilization rate is lower than the low water level includes: The average real-time utilization rate of the business terminal during the monitoring period is compared with its corresponding low water level line. When the average real-time utilization rate is less than the low water level, the difference between the low water level and the average real-time utilization rate is calculated as the amount of resources that can be released. The radio interface resource units corresponding to the releasable resource quantity are then released to the shared resource pool.

7. The method according to claim 1, characterized in that, The calculation of the scheduling weight includes: The comprehensive service priority score S(t) of each terminal is calculated based on the multi-dimensional service state vector; the scheduling weight W(t) of each terminal at scheduling time t is calculated using the following formula: Where W(t) is the scheduling weight of the terminal at scheduling time t, and S(t) is the comprehensive service priority score of the terminal. The average comprehensive service priority score of all terminals within the soft slice. The predicted signal-to-noise ratio of the terminal is given. This represents the average predicted signal-to-noise ratio for all terminals within the soft slice. and The weighting coefficients and .

8. A 5G signal coverage optimization system for mines, characterized in that, Includes the following modules: The first calculation module is used to obtain the three-dimensional model of the mine, the real-time location and predetermined trajectory of the underground mobile equipment, and the business requests of each business terminal. Based on the three-dimensional model, the real-time location of the device and the predetermined trajectory, the ray tracing algorithm is used to calculate the spatiotemporal channel prediction factor of each service terminal and base station in the future scheduling cycle. The first allocation module is used to construct a multi-dimensional service state vector that includes service type, equipment security status and production process priority; and to allocate hard slice resources to service terminals that meet preset key service conditions or whose predicted channel quality is lower than a predetermined threshold, and to allocate soft slice resources to other service terminals by combining the service state vector with the spatiotemporal channel prediction factor. The second allocation module is used to set high and low watermarks for resource utilization rate for service terminals that have been allocated hard slice resources, based on the terminal service quality requirements; when the real-time utilization rate is lower than the low watermark, the difference in resources is released to a shared resource pool; when the real-time utilization rate is higher than the high watermark, supplementary resources are allocated to the terminal from the shared resource pool. The second calculation module is used to calculate the scheduling weight for a service terminal that has been allocated soft slice resources. The scheduling weight determines the priority and share of the terminal to obtain temporary resources from the shared resource pool.

9. The system according to claim 8, characterized in that, The method employs a ray tracing algorithm to calculate the spatiotemporal channel prediction factors for each service terminal and base station within a future scheduling cycle, including: The three-dimensional model of the mine is discretized into a set of triangular facets. For the future location of each service terminal, multiple rays are emitted, and the propagation loss of the direct path, first-order reflection path, and first-order diffraction path of the rays is simulated and calculated. The channel gain at the location is obtained by combining the loss of all paths, and the predicted reference signal received power RSRP and predicted signal-to-noise ratio SINR are calculated in combination with the base station transmit power as the spatiotemporal channel prediction factors.

10. The system according to claim 8, characterized in that, The construction of a multi-dimensional business state vector, which includes business type, equipment safety status, and production process priority, includes: The business types are quantified, with remote control assigned a value of 10, high-definition video surveillance assigned a value of 8, and data acquisition assigned a value of 5; the equipment safety status is quantified, with alarm status assigned a value of 1 and normal status assigned a value of 0; the production process priority is quantified, with terminals related to coal mining faces assigned a value of 3, terminals related to tunneling faces assigned a value of 2, and terminals related to main transport roadways assigned a value of 1.

Images