Mobile cell resource minimization slice resource configuration method and system

Abstract

The invention provides a mobile cell resource minimization slice resource configuration method and system, and belongs to the technical field of train communication, and the method comprises the steps: building a train mobile network model based on a relay system, and constructing a resource slice framework; establishing a communication and service quality assurance model for a passenger on-demand service and a beyond-visual-range service by considering a downlink bearing demand of a train intelligent service; establishing an optimization problem taking bandwidth allocation and terminal preemptive grouping as target variables; simplifying an original problem by using a convex optimization theory, and decomposing the original problem into a resource block allocation sub-problem and a mini time slot preemption sub-problem; and iteratively solving the two sub-problems by using a resource block allocation and preemption efficient algorithm based on block coordinate descent to obtain an optimal resource allocation strategy. Under the condition of non-ideal train-ground link transmission caused by the high-speed movement characteristic of a train, the deterministic transmission requirement of heterogeneous services is ensured, and the joint design of a preemption-based slice deployment strategy and a resource allocation algorithm is realized.

Description

Technical Field

[0001] The present invention belongs to the field of train communication technology, and in particular relates to a method and system for configuring slice resources for minimizing mobile cell resources. Background Art

[0002] High-speed railways are gradually developing towards intelligentization, requiring them to meet not only train safety and reliability but also passenger comfort. The rapid development of high-speed rail has also placed higher demands on railway wireless communication networks. To enhance the level of intelligence in high-speed rail, railway communication networks must carry a variety of train services related to safe operation, such as new train status monitoring and assisted driving services. Furthermore, to meet passengers' demands for a high-quality travel experience, communication networks must provide high-capacity passenger video-on-demand services. These various services have varying Quality of Service (QoS) transmission requirements, such as throughput, latency, and connectivity. To address the challenges of simultaneously meeting diverse service requirements and the potential risks of service isolation, network slicing technology within fifth-generation mobile communication technology (5G) offers a viable solution. Network slicing isolates multiple virtual networks to carry different services, meeting the QoS requirements of various services. However, obtaining accurate channel state information (CSI) is difficult in high-speed mobility scenarios. Under non-ideal CSI conditions, existing mixed service carrying and slicing strategies cannot strictly meet the QoS requirements of different services, particularly on-board services related to train safety. Therefore, it is necessary to fully consider the particularity of high-speed rail scenarios and conduct research on service carrying and slice resource allocation in high-speed rail scenarios to ensure safe train operation and provide passengers with a high-quality travel experience. In summary, while ensuring the QoS transmission requirements of train services, a solution that can carry heterogeneous services in mobile cells with non-ideal links is urgently needed. Summary of the Invention

[0003] In response to the above-mentioned deficiencies in the prior art, the present invention provides a method and system for minimizing slice resource configuration for mobile cell resources for non-ideal links, which solves the problems of heterogeneous service carrying and resource optimization configuration under high-speed mobile cell network deployment.

[0004] In order to achieve the above objectives, the technical solution adopted by the present invention is: a method for configuring slice resources for minimizing mobile cell resources, comprising the following steps:

[0005] Build a train mobile network model based on a relay system and construct a resource slicing framework;

[0006] Based on the constructed resource slicing framework, a communication and service quality assurance model is established for passenger on-demand services and beyond-line-of-sight services, taking into account the downlink carrying requirements of train intelligent services.

[0007] Based on the constructed communication and service quality assurance model, an optimization problem with bandwidth allocation and terminal preemption grouping as target variables is established, where the goal is to minimize bandwidth resources;

[0008] Based on the optimization problem of the target variable, the original problem is simplified using convex optimization theory. The optimal allocation strategy of the resource block allocation subproblem is derived through constraint relaxation. The mini-time slot preemption subproblem is converted into an objective function and auxiliary variables are introduced to simplify it into an easily solvable form.

[0009] By using the efficient BCD-AP algorithm for resource block allocation and preemption based on block coordinate descent, the resource block allocation subproblem and the mini-time slot preemption subproblem are iteratively solved to obtain the optimal resource allocation strategy and complete the mobile cell resource minimization slice resource configuration.

[0010] The present invention has the following advantages: it can carry passenger on-demand and beyond-line-of-sight services with differentiated QoS requirements over non-ideal train-to-ground mobile networks. It also utilizes the proposed resource block allocation and preemption (BCD-AP) algorithm to meet the QoS transmission requirements of these heterogeneous services, minimize system bandwidth resources, and achieve an optimal resource allocation strategy. The proposed resource slicing strategy can minimize bandwidth resources and simultaneously meet the transmission requirements of both heterogeneous services in mobile cell network deployments.

[0011] Furthermore, the train mobile network model based on the relay system is built to construct a resource slicing framework, which is specifically as follows:

[0012] Establish a link between the base station and the rooftop relay, where the rooftop relay communicates with the terminal through the train station;

[0013] Trackside cameras are used to capture real-time video data, which is then transmitted to a ground data processing center for processing. The processing results are then sent to the onboard DAS terminal, which includes a VDS terminal and a DAS terminal.

[0014] The system bandwidth is divided into a set of resource blocks, denoted as k = {1, 2, ..., K}, and the bandwidth of a single resource block RB is denoted as B unit , is defined as the minimum resource unit allocated to the VDS terminal in the time-frequency domain, where k represents the number of resource blocks and K represents the number of resource blocks;

[0015] Divide the time domain into time domains of length T s =1 millisecond time slot, where the time length of a single mini-time slot is represented by T ms =T s / Ns , where N s Indicates the number of mini-slots in a single millisecond time slot, T ms Indicates the duration of a single mini-slot, T s Indicates the length of a single millisecond time slot;

[0016] A single resource block (RB) is allocated to the VDS terminal using a standard time slot. Data packets arriving at the DAS terminal seize the bandwidth resources allocated to the VDS terminal and select available mini-time slots to transmit data, completing the construction of the resource slicing framework.

[0017] The beneficial effect of the above further solution is: building a train mobile network resource slicing framework based on the relay system, combined with the time-correlation statistical distribution of the fast time-varying fading channel, it is possible to accurately characterize the small-scale fading characteristics of the train-to-ground link under non-ideal CSI conditions.

[0018] Furthermore, the communication and service quality assurance model for passenger on-demand services and beyond-line-of-sight services is established, which is specifically as follows:

[0019] Considering the downlink carrying demand of train intelligent services, a linear model is used to describe the relationship between the rate loss of passenger on-demand services and the number of mini-slots occupied by assisted driving services. The transmission data rate of VDS terminal m is:

[0020]

[0021]

[0022] Among them, r m represents the transmission data rate of VDS terminal m, B unit Represents the bandwidth of a single resource block RB, x m represents the total number of single resource blocks (RBs) allocated to VDS terminal m, z m represents the number of mini-slots that DAS terminal n preempts VDS terminal m, γ m represents the signal-to-noise ratio of VDS terminal m, M represents the set of multiple VDS terminals in the system, a m represents the large-scale channel gain, P m represents the transmission power of VDS terminal m, h m represents the channel state information, σ 2 represents the additive white Gaussian noise power;

[0023] Based on the described relationship, the time correlation in the fast time-varying fading channel is used to construct the actual CSI|h j | 2 The statistical probability distribution of , where the conditional probability density function is expressed as follows:

[0024]

[0025] ρ[τ]=J0(2πf D τ)

[0026] f D =vf c / v c

[0027] Among them, f() represents the conditional probability density function, h j represents the real small-scale fading of the j-th terminal, ρ[τ]∈[0,1] represents the channel estimation error coefficient, J0() represents the first-kind zero-order Bessel function, f D represents the maximum Doppler frequency shift, ρ[τ] represents the channel estimation error coefficient, τ represents the channel feedback delay, v represents the train speed, and f c Indicates the carrier frequency, v c represents the speed of light, represents the estimated small-scale fading of the j-th terminal, represents the additive white Gaussian noise power, I0() represents the first-kind zero-order modified Bessel function, e j represents the estimation error, M represents the set of multiple VDS terminals in the system, and n represents only one DAS terminal in the system;

[0028] Utilize|h m | 2 The conditional probability density function of is used to obtain the average data rate of the VDS terminal m on all single resource blocks RB, completing the establishment of the passenger on-demand service communication and service quality assurance model:

[0029]

[0030] in, represents the average data rate of VDS terminal m on all single resource blocks RB, f() represents the conditional probability density function, represents the estimated small-scale fading of the mth VDS terminal, and d() represents the integration operation;

[0031] By calculating the data rate of the DAS terminal, the average data rate of DAS terminal n is obtained;

[0032] The data transmission delay of the DAS terminal is obtained according to the average data rate of the DAS terminal n;

[0033] According to the data transmission delay of the DAS terminal, the conditional risk value is used to capture the tail characteristics of the DAS transmission delay to obtain φ a (D n ,δ) function, completes the establishment of the beyond-line-of-sight business communication and service quality assurance model.

[0034] The beneficial effect of the above further solution is: by introducing the small-scale fading statistical distribution of the time-varying fading channel and the channel estimation error model, using |h j | 2 The probability density function is used to derive the average data rate of VDS terminals.

[0035] Furthermore, the data transmission delay of the DAS terminal is expressed as follows:

[0036]

[0037] Among them, D n represents the data transmission delay of the DAS terminal, F n Indicates the length of the data packet of the DAS terminal. represents the average data rate of DAS terminal n, r n represents the data rate of the DAS terminal, f() represents the conditional probability density function, h n represents the real small-scale fading of DAS terminal n, represents the estimated small-scale fading of DAS terminal n, γ n represents the received signal-to-noise ratio of terminal n, a n represents the large-scale channel gain, P n Indicates the transmission power of DAS terminal n.

[0038] The beneficial effect of the above further solution is that the average rate of the DAS terminals can be further calculated based on the average rate of the above VDS terminals.

[0039] Furthermore, the φ a (D n ,δ) function expression is as follows:

[0040] φ a (D n ,δ):=δ+(1-a) -1 E[(D n -δ)+]

[0041] Among them, φ a (D n ,δ) represents the C-VaR function, δ represents the delay threshold variable, a represents the delay threshold, E represents the expected value, and C-VaR represents the average value of the potential delay exceeding the risk value.

[0042] The beneficial effect of the above further solution is: using the conditional risk value method to capture the tail characteristics of DAS transmission delay, quantify the jitter risk level, and provide reliable service quality assurance for DAS with dual QoS requirements of delay and jitter.

[0043] Furthermore, the bandwidth resource minimization problem can be expressed as follows:

[0044]

[0045] φ a (D n ,δ)≤θ max ,

[0046]

[0047] Among them, F represents the objective function, x and z are both optimization variable sets, x={x m} represents the number of allocated resource blocks, z={z m} represents the number of allocated mini-slots, δ represents the delay threshold variable, x m represents the total number of single resource blocks (RBs) allocated to VDS terminal m, M represents the set of multiple VDS terminals in the system, and B unit Represents the bandwidth of a single resource block RB, r m represents the transmission data rate of VDS terminal m, represents the minimum throughput QoS requirement of VDS terminal m, φ a (D n ,δ) represents the C-VaR function, D n represents the data transmission delay of the DAS terminal, θ max Indicates the maximum C-VaR risk threshold of the DAS terminal. C-VaR represents the average value of potential delay exceeding the risk value. represents a set of positive integers, z m Indicates the number of mini-slots that DAS terminal n preempts VDS terminal m, N s Indicates the number of minislots in a single millisecond slot.

[0048] The beneficial effects of the above further scheme are: establishing a mathematical optimization model for the bandwidth resource minimization problem, jointly optimizing resource block allocation variables and mini-time slot preemption variables, and meeting the throughput QoS requirements of VDS terminals and the delay and jitter QoS requirements of DAS terminals.

[0049] Furthermore, the resource block allocation subproblem is expressed as follows:

[0050]

[0051] Where F represents the objective function, x represents the set of optimized variables, and x={x m} represents the number of allocated resource blocks, r m represents the transmission data rate of VDS terminal m, x mIndicates the total number of single resource blocks RB allocated to VDS terminal m, represents the set of positive integers, represents the optimal resource block allocation for terminal m, represents the minimum throughput QoS requirement of VDS terminal m, B unit Indicates the bandwidth of a single resource block RB, R m represents the transmission rate of VDS terminal m, z m Indicates the number of mini-slots that DAS terminal n preempts VDS terminal m, N s Indicates the number of mini-slots in a single millisecond time slot, γ m represents the signal-to-noise ratio of the VDS terminal m, f() represents the conditional probability density function, h m represents the channel state information, ρ[τ] represents the channel estimation error coefficient, represents the estimated small-scale fading of the mth VDS terminal, τ represents the channel feedback delay, and d() represents the integration operation.

[0052] The beneficial effect of the above further solution is: establishing an expression for the resource block allocation subproblem, solving it to obtain a closed-form solution for optimal resource block allocation, and having low computational complexity.

[0053] Furthermore, the expression of the mini-slot preemption sub-problem is as follows:

[0054]

[0055] s.tδ+(1-a) -1 κ≤θ max ,

[0056] κ≥E[D n ]-δ,

[0057] κ≥0,

[0058]

[0059]

[0060] Among them, F1 represents the sub-problem objective function, z is the set of optimization variables, z={z m} represents the number of allocated mini-slots, δ represents the delay threshold variable, κ represents the auxiliary variable, a represents the delay threshold, θ max represents the maximum risk threshold of C-VaR of DAS terminal. C-VaR represents the average value of potential delay exceeding the risk value. E[D n ] indicates D n Find the expectation, E represents the expected value, D n represents the data transmission delay of the DAS terminal, x mN represents the total number of single resource blocks (RBs) allocated to VDS terminal m. s Indicates the number of mini-slots in a single millisecond time slot, M represents the collection of multiple VDS terminals in the system, q m Indicates the number of mini-slots that the DAS terminal can seize without affecting the VDS transmission rate. m Indicates the number of mini-slots that DAS terminal n preempts VDS terminal m. represents the minimum throughput QoS requirement of VDS terminal m, B unit Indicates the bandwidth of a single resource block RB, R m Indicates the transmission rate of VDS terminal m, F n Indicates the length of the data packet at the DAS terminal, γ n represents the signal-to-noise ratio of DAS terminal n, h n represents the actual small-scale fading of DAS terminal n, ρ[τ] represents the channel estimation error coefficient, d() represents the integral operation, represents the estimated small-scale fading of DAS terminal n, and τ represents the channel feedback delay.

[0061] The beneficial effect of the above further solution is: establishing a mini-time slot preemption sub-problem, minimizing the VDS throughput loss caused by DAS preemption while strictly meeting the DAS service quality requirements, and saving system bandwidth resources.

[0062] The present invention also provides a mobile cell resource minimization slice resource configuration system, comprising:

[0063] The first processing module is used to build a train mobile network model based on the relay system and construct a resource slicing framework;

[0064] The second processing module is used to establish a communication and service quality assurance model for passenger on-demand services and beyond-line-of-sight services based on the constructed resource slicing framework, while considering the downlink carrying requirements of train intelligent services;

[0065] The third processing module is used to establish an optimization problem with bandwidth allocation and terminal preemption grouping as target variables based on the constructed communication and service quality assurance model, wherein the goal is to minimize bandwidth resources;

[0066] The fourth processing module is used to solve the optimization problem based on the target variable. It uses convex optimization theory to simplify the original problem, derives the optimal allocation strategy for the resource block allocation subproblem through constraint relaxation, and converts the mini-time slot preemption subproblem into an objective function, introduces auxiliary variables, and simplifies it into an easily solvable form.

[0067] The fifth processing module is used to use the efficient BCD-AP algorithm for resource block allocation and preemption based on block coordinate descent to iteratively solve the resource block allocation subproblem and the mini-time slot preemption subproblem, obtain the optimal resource allocation strategy, and complete the mobile cell resource minimization slice resource configuration.

[0068] The present invention has the following advantages: it can carry passenger on-demand and beyond-line-of-sight services with differentiated QoS requirements over non-ideal train-to-ground mobile networks. It also utilizes the proposed resource block allocation and preemption (BCD-AP) algorithm to meet the QoS transmission requirements of these heterogeneous services, minimize system bandwidth resources, and achieve an optimal resource allocation strategy. The proposed resource slicing strategy can minimize bandwidth resources and simultaneously meet the transmission requirements of both heterogeneous services in mobile cell network deployments. BRIEF DESCRIPTION OF THE DRAWINGS

[0069] Figure 1 Flow chart of the method of the present invention.

[0070] Figure 2 This is a diagram of the train mobile network architecture based on the relay system in the present invention.

[0071] Figure 3 Schematic diagram of the system structure of the present invention. DETAILED DESCRIPTION

[0072] The specific embodiments of the present invention are described below to facilitate understanding of the present invention by those skilled in the art. However, it should be clear that the present invention is not limited to the scope of the specific embodiments. For those skilled in the art, as long as various changes are within the spirit and scope of the present invention as defined and determined by the appended claims, these changes are obvious, and all inventions and creations utilizing the concepts of the present invention are protected.

[0073] Example 1

[0074] like Figure 1 and Figure 2 As shown, the present invention provides a method for minimizing slice resource configuration of mobile cell resources for non-ideal links, and its implementation method is as follows:

[0075] (1) Build a train mobile network model based on the relay system and construct a resource slicing framework, which is specifically:

[0076] A link is established between the base station and the rooftop relay, where the rooftop relay communicates with the terminal through the train station; trackside cameras are used to capture real-time video data, and the video data is transmitted to the ground data processing center for processing, and the processing results are sent to the on-board DAS terminal, where the on-board terminal includes the VDS terminal and the DAS terminal; the system bandwidth is divided into a set of resource blocks, denoted as k = {1, 2, ..., K}, and the bandwidth of a single resource block RB is denoted as B unit , defined as the minimum resource unit allocated to the VDS terminal in the time-frequency domain, where k represents the number of resource blocks and K represents the number of resource blocks; the time domain is divided into time periods of T s =1 millisecond time slot, where the time length of a single mini-time slot is represented by T ms =T s / N s , where N s Indicates the number of mini-slots in a single millisecond time slot, T ms Indicates the duration of a single mini-slot, T s Represents the time length of a single millisecond time slot; a single resource block RB is allocated to the VDS terminal using a standard time slot. Data packets arriving at the DAS terminal seize the bandwidth resources allocated to the VDS terminal and select available mini-time slots to transmit data, completing the construction of the resource slicing framework.

[0077] In this embodiment, the present invention considers a train mobile network architecture based on a relay system, such as Figure 2 As shown in Figure 1. To reduce vehicle penetration loss and ensure reliable connectivity, the trackside base station communicates with the train's onboard terminals via a two-hop transmission link. The base station first establishes a link with the rooftop relay, which then communicates with the terminal via the train's abutment station. Consider two types of train services: passenger on-demand service (VDS) and driver assistance service (DAS). The VDS provides passengers with customized, high-quality media videos. Its QoS requirement is high throughput, ensuring smooth video playback and enhancing the passenger experience. The driver assistance service transmits trackside surveillance video containing hazard warnings, providing driving guidance information to the driver. Trackside cameras capture real-time video data and transmit it to a ground data processing center for intelligent recognition and data processing. The relevant video and data are then sent to the onboard DAS terminal, providing the driver with critical information beyond their line of sight. DAS has strict latency and jitter requirements. Therefore, the onboard terminals are divided into VDS terminals and DAS terminals, with the set of VDS terminals set to m = {1, 2, ..., M}. Furthermore, assume that one DAS terminal is deployed in the train cab, denoted by n.

[0078] In order to meet the heterogeneous QoS requirements of the above two types of services, the present invention adopts a slice mini-time slot preemption scheme, such as Figure 2As shown. The system bandwidth is divided into a set of resource blocks, denoted as k = {1, 2, ..., K}, and the bandwidth of a single RB (resource block) is B unit , defined as the minimum resource unit allocated to the VDS terminal in the time-frequency domain. The time domain is divided into two parts with a duration of T s = 1 millisecond time slot, each time slot contains N s =7 mini-slots, the time length of a single mini-slot is expressed as T ms =T s / N s , where N s Indicates the number of mini-slots in a single millisecond time slot, T ms Indicates the duration of a single mini-slot, T s Represents the duration of a single millisecond time slot. To effectively meet high throughput requirements, RBs are allocated to VDS terminals using standard time slots. Furthermore, because DAS data arrives randomly and has strict latency and jitter requirements, arriving DAS packets preempt bandwidth resources allocated to VDS terminals and select available mini-time slots for data transmission.

[0079] Considering the significant Doppler shift caused by the high-speed movement of the train, it is assumed that the base station cannot obtain the ideal channel state information (CSI) of the onboard terminal. Therefore, the base station can only obtain the estimated small-scale fading Where τ represents the channel feedback delay. The current CSIh j Feedback with estimates The relationship expression is as follows:

[0080]

[0081] Among them, ρ[τ]∈[0,1] represents the channel estimation error coefficient, that is, ρ[τ]=J0(2πf D τ), J0() is the first-order Bessel function. The maximum Doppler shift is f D =vf c / v c , where v represents the train speed, f c Indicates the carrier frequency, v c Represents the speed of light, estimated error and Independent of each other.

[0082] In order to deal with non-ideal CSI, the time correlation in fast time-varying fading channels is used to construct a real The statistical probability distribution of , its conditional probability density function can be derived as:

[0083]

[0084] Where I0() represents the first kind of zero-order modified Bessel function, f() represents the conditional probability density function, and h j represents the real small-scale fading of the j-th terminal, represents the estimated small-scale fading of the j-th terminal, represents the additive white Gaussian noise power, M represents the set of multiple VDS terminals in the system, and n represents that there is only one DAS terminal in the system.

[0085] (2) Based on the constructed resource slicing framework, when considering the downlink carrying requirements of train intelligent services, a communication and service quality assurance model is established for passenger on-demand services and beyond-line-of-sight services. Specifically, a communication and service quality assurance model is established for passenger on-demand services and beyond-line-of-sight services, which is as follows:

[0086] Considering the downlink carrying demand of train intelligent services, a linear model is used to describe the relationship between the rate loss of passenger on-demand services and the number of mini-slots occupied by assisted driving services. Based on the described relationship, the time correlation in the fast time-varying fading channel is used to construct the actual CSI|h j | 2 The statistical probability distribution of |h m | 2 The conditional probability density function of is used to obtain the average data rate of the VDS terminal m on all single resource blocks (RBs), thus completing the establishment of the passenger on-demand service communication and service quality assurance model.

[0087] By calculating the data rate of the DAS terminal, the average data rate of DAS terminal n is obtained; according to the average data rate of DAS terminal n, the data transmission delay of the DAS terminal is obtained; according to the data transmission delay of the DAS terminal, the conditional risk value is used to capture the tail characteristics of the DAS transmission delay to obtain φ a (D n ,δ) function, completes the establishment of the beyond-line-of-sight business communication and service quality assurance model.

[0088] In this embodiment, the present invention uses a linear model to describe the relationship between the rate loss of the passenger on-demand service terminal and the number of mini-time slots occupied by the assisted driving service. Assuming that the DAS data packet is transmitted in each time slot through the slice preemption scheme, in each transmission cycle, the system allocates a corresponding number of RBs to the VDS terminal, and the integer allocation index Denotes the total number of RBs allocated to VDS terminal m. Let z m It represents the number of mini-slots that DAS terminal n preempts VDS terminal m. Therefore, the transmission data rate of VDS terminal m is:

[0089]

[0090] Among them, γ m Indicates the signal-to-noise ratio of VDS terminal m:

[0091]

[0092] Among them, r m represents the transmission data rate of VDS terminal m, B unit Represents the bandwidth of a single resource block RB, x m represents the total number of single resource blocks (RBs) allocated to VDS terminal m, z m represents the number of mini-slots that DAS terminal n preempts VDS terminal m, γ m represents the signal-to-noise ratio of VDS terminal m, a m represents the large-scale channel gain, P m represents the transmission power of VDS terminal m, h m represents the channel state information, σ 2 represents the additive white Gaussian noise power.

[0093] Since the base station cannot obtain the ideal channel state information h m , using |h m | 2 The probability density function of VDS terminal m on all RBs is:

[0094]

[0095] in, represents the average data rate of the VDS terminal m on all single resource blocks RB, f() represents the conditional probability density function, ρ[τ] represents the channel estimation error coefficient, represents the estimated small-scale fading of the mth VDS terminal, τ represents the channel feedback delay, d() represents the integral operation, and B unit Indicates the bandwidth of the resource block RB.

[0096] Since the bandwidth resources allocated to DAS terminal n depend on all VDS terminals, the data rate of the DAS terminal can be expressed as:

[0097]

[0098] Among them, γ n represents the received signal-to-noise ratio of terminal n, which can be expressed as:

[0099]

[0100] Among them, P n represents the transmission power of DAS terminal n. Therefore, the average data rate of DAS terminal n can be expressed as:

[0101]

[0102] Since DAS packets transmit data by immediately preempting the RB resources of the VDS terminal, we do not consider the queuing delay and only consider the DAS data transmission delay. The data transmission delay of the DAS terminal is:

[0103]

[0104] Among them, D n represents the data transmission delay of the DAS terminal, F n Indicates the length of the data packet of the DAS terminal. represents the average data rate of DAS terminal n, r n Indicates the data rate of the DAS terminal a n represents the large-scale channel gain, h n represents the real small-scale fading of DAS terminal n, represents the estimated small-scale fading of DAS terminal n.

[0105] To characterize the QoS requirements of DAS, the Conditional Value-at-Risk (CVaR) function is used to capture the tail characteristics of DAS transmission delay. CVaR represents the average value of potential delays exceeding the risk value. The parameter a-VaR is defined, which represents the probability that the average DAS terminal delay does not exceed a minimum value, greater than or equal to a. It can be expressed as:

[0106]

[0107] Among them, P() represents the probability density function, δ represents the delay threshold variable, and the CVaR function represents the delay threshold variable in D n ≥VaR a (D n ) condition D n The expected value of , which represents the average data delay in the case of violation, can be expressed as:

[0108] CVaR a (D n )=Ε[D n |D n ≥VaR a (D n )]

[0109] Therefore, we can get:

[0110] φ a (D n ,δ):=δ+(1-a) -1 E[(D n -δ)+]

[0111] Among them, φ a (D n ,δ) represents the C-VaR function, C-VaR represents the average value of potential delay exceeding the risk value, δ represents the delay threshold variable, a represents the delay threshold, and E represents the expected value.

[0112] (3) Based on the constructed communication and service quality assurance model, an optimization problem is established with bandwidth allocation and terminal preemption grouping as the objective variables, where the objective is to minimize bandwidth resources and the constraints are the latency and jitter requirements of beyond-line-of-sight services, the data throughput requirements of passenger services, and resource constraints;

[0113] In this embodiment, in order to minimize the system bandwidth resource consumption, the utility function is defined Define the parameter vector as x={x m} and z={z m Therefore, the bandwidth resource minimization problem can be expressed as follows:

[0114]

[0115] φ a (D n ,δ)≤θ max ,

[0116]

[0117] Where F represents the objective function, x and z are both optimization variable sets, x={x m} represents the number of allocated resource blocks, z={z m} represents the number of allocated mini-slots, δ represents the delay threshold variable, x m represents the total number of single resource blocks (RBs) allocated to VDS terminal m, M represents the set of multiple VDS terminals in the system, and B unit Represents the bandwidth of a single resource block RB, r m represents the transmission data rate of VDS terminal m, represents the minimum throughput QoS requirement of VDS terminal m, φ a (D n ,δ) represents the C-VaR function, D n represents the data transmission delay of the DAS terminal, θ max Indicates the maximum C-VaR risk threshold of the DAS terminal. C-VaR represents the average value of potential delay exceeding the risk value. represents a set of positive integers, z m Indicates the number of mini-slots that DAS terminal n preempts VDS terminal m, N s Indicates the number of minislots in a single millisecond slot.

[0118] To ensure that the arriving DAS data packets are transmitted within a single time slot, δ max =1ms is set as the upper limit of δ, that is, δ≤δ max Due to the existence of integer optimization variables, this problem is a mixed integer nonlinear programming problem, δ max =1ms indicates the upper limit value.

[0119] (4) Based on the optimization problem of the target variable, the original problem is simplified using convex optimization theory. The optimal allocation strategy of the resource block allocation subproblem is derived through constraint relaxation. The mini-time slot preemption subproblem is converted into an objective function and auxiliary variables are introduced to simplify it into an easily solvable form.

[0120] In this embodiment, to solve the above optimization problem, the present invention proposes a resource block allocation and preemption (BCD-AP) algorithm based on block coordinate descent and obtains a suboptimal feasible solution. The original problem is decomposed into two sub-problems: (1) RB allocation for passenger on-demand services; (2) mini-slot preemption for assisted driving services.

[0121] In the RB allocation subproblem of passenger on-demand services, the DAS preemption variable z is fixed. m , only optimize RB allocation variable x m Therefore, this sub-problem can be expressed as:

[0122]

[0123] It can be seen that the RB allocation problem of VDS is an integer linear programming problem. To simplify the problem, we first relax the integer constraints. Therefore, the closed-form solution for the optimal RB allocation of terminal m is as follows:

[0124]

[0125] Among them, the auxiliary variables

[0126] in, represents the optimal resource block allocation for terminal m, represents the minimum throughput QoS requirement of VDS terminal m, B unit Indicates the bandwidth of a single resource block RB, R m represents the transmission rate of VDS terminal m, z m Indicates the number of mini-slots that DAS terminal n preempts VDS terminal m, N s Indicates the number of mini-slots in a single millisecond time slot, γ m represents the signal-to-noise ratio of the VDS terminal m, f() represents the conditional probability density function, h m represents the channel state information, |ρ[τ] represents the channel estimation error coefficient, represents the estimated small-scale fading of the mth VDS terminal, and τ represents the channel feedback delay.

[0127] In the mini-slot preemption problem of the assisted driving service, for a given VDS RB allocation strategy x, the original problem can be transformed into the following DAS preemption problem:

[0128]

[0129] δ+(1-a) -1 E[(D n -δ) + ]≤θ max ,

[0130]

[0131] Where F represents the objective function, z is the set of optimization variables, z={z m} indicates a set of allocated mini-slot quantities.

[0132] However, in the process of solving a given RB allocation strategy, the objective function of the above problem remains unchanged. In order to obtain the optimal DAS slice preemption strategy, the objective of this sub-problem is redefined and kept consistent with the original problem objective, that is, minimizing system bandwidth resources and reducing network deployment costs. After DAS preempts VDS bandwidth resources, the algorithm needs to supplement additional RB resources for VDS in subsequent iterations to ensure the service quality requirements of VDS. Therefore, it is crucial to minimize the impact of DAS preemption on the RB resource allocation of VDS terminals. Considering that the two services have different resource allocation dimensions, that is, VDS is allocated based on time slots and DAS is allocated based on mini time slots. Therefore, the design allows DAS to give priority to the use of the excess mini time slots of VDS terminals without affecting VDS traffic, thereby minimizing the number of additional RBs allocated to VDS in subsequent iterations. This scheme effectively minimizes the number of system RB occupancy. We define the number of mini time slots remaining for VDS terminal m as q m , which represents the mini-time slot resources that can be used by the DAS terminal without affecting the QoS of the VDS, is recorded as:

[0133]

[0134] Define the utility function At the same time, to remove non-convex terms (·) + , introduce auxiliary variable κ to replace E[(D n -δ) + ]. Therefore, the atomic problem can be reformulated as:

[0135]

[0136] s.tδ+(1-a) -1 κ≤θ max ,

[0137] κ≥E[D n ]-δ,

[0138] κ≥0,

[0139]

[0140] The average delay of DAS data packets is approximately:

[0141]

[0142] Among them, F1 represents the sub-problem objective function, z is the set of optimization variables, z={z m} represents the number of allocated mini-slots, δ represents the delay threshold variable, κ represents the auxiliary variable, a represents the delay threshold, θ max represents the maximum risk threshold of C-VaR of DAS terminal. C-VaR represents the average value of potential delay exceeding the risk value. E[D n ] indicates D n Find the expectation, E represents the expected value, D n represents the data transmission delay of the DAS terminal, x m N represents the total number of single resource blocks (RBs) allocated to VDS terminal m. s Indicates the number of mini-slots in a single millisecond time slot, M represents the collection of multiple VDS terminals in the system, q m Indicates the number of mini-slots that the DAS terminal can seize without affecting the VDS transmission rate. m Indicates the number of mini-slots that DAS terminal n preempts VDS terminal m. represents the minimum throughput QoS requirement of VDS terminal m, B unit Indicates the bandwidth of a single resource block RB, R m Indicates the transmission rate of VDS terminal m, F n Indicates the length of the data packet at the DAS terminal, γ n represents the signal-to-noise ratio of DAS terminal n, h n represents the actual small-scale fading of DAS terminal n, ρ[τ] represents the channel estimation error coefficient, d() represents the integral operation, represents the estimated small-scale fading of DAS terminal n, and τ represents the channel feedback delay.

[0143] (5) Using the efficient BCD-AP algorithm for resource block allocation and preemption based on block coordinate descent, the resource block allocation subproblem and the mini-time slot preemption subproblem are iteratively solved to obtain the optimal resource allocation strategy and complete the mobile cell resource minimization slice resource configuration;

[0144] In this embodiment, due to the objective function F1 and the integer variable z, the above problem remains non-convex. To address the mini-slot preemption subproblem for assisted driving services, this paper proposes a mini-slot preemption algorithm based on a genetic algorithm to achieve an optimal DAS slicing strategy. Based on the above design, an efficient block coordinate descent (BCD-AP) algorithm is proposed to allocate RBs and mini-slot resources to VDS terminals and DAS terminals, respectively, and a feasible slicing preemption strategy is designed. The proposed slicing strategy can minimize bandwidth resources in mobile cell network deployments while simultaneously meeting the transmission requirements of two types of heterogeneous services.

[0145] In summary, the present invention first establishes a train mobile network model based on a relay system to characterize the mobility characteristics of the train-to-ground communication link; fully considers the downlink carrying requirements of train intelligent services, pays attention to the QoS guarantee requirements of passenger services and beyond-line-of-sight services, and establishes a service communication and service quality assurance model; considering that beyond-line-of-sight services are delay-sensitive services, the present invention designs a slicing scheme for beyond-line-of-sight services to preempt passenger service resource blocks in real time; secondly, with the goal of minimizing bandwidth resource consumption, and with the delay and jitter requirements of beyond-line-of-sight services, the data throughput requirements of passenger services, and resource limitations as constraints, the present invention establishes an optimization problem with bandwidth allocation and terminal preemption grouping as objective variables. The present invention proposes a BCD-AP slice resource configuration algorithm based on block coordinate descent and designs a feasible slice preemption strategy. The present invention effectively completes resource optimization configuration. The proposed slicing strategy can minimize bandwidth resources and simultaneously meet the transmission requirements of two types of heterogeneous services in mobile cell network deployment. It solves the problem of heterogeneous service carrying and resource optimization configuration in high-speed mobile cell network deployment.

[0146] Example 2

[0147] like Figure 3 As shown, the present invention provides a mobile cell resource minimization slice resource configuration system for executing embodiment 1, including:

[0148] The first processing module is used to build a train mobile network model based on the relay system and construct a resource slicing framework;

[0149] The second processing module is used to establish a communication and service quality assurance model for passenger on-demand services and beyond-line-of-sight services based on the constructed resource slicing framework, while considering the downlink carrying requirements of train intelligent services;

[0150] The third processing module is used to establish an optimization problem with bandwidth allocation and terminal preemption grouping as target variables based on the constructed communication and service quality assurance model, wherein the goal is to minimize bandwidth resources;

[0151] The fourth processing module is used to solve the optimization problem based on the target variable. It uses convex optimization theory to simplify the original problem, derives the optimal allocation strategy for the resource block allocation subproblem through constraint relaxation, and converts the mini-time slot preemption subproblem into an objective function, introduces auxiliary variables, and simplifies it into an easily solvable form.

[0152] The fifth processing module is used to use the efficient BCD-AP algorithm for resource block allocation and preemption based on block coordinate descent to iteratively solve the resource block allocation subproblem and the mini-time slot preemption subproblem, obtain the optimal resource allocation strategy, and complete the mobile cell resource minimization slice resource configuration.

[0153] In this embodiment, the present application can divide the functional units according to the mobile cell resource minimization slice resource configuration method. For example, each function can be divided into each functional unit, or two or more functions can be integrated into one processing unit. The above-mentioned integrated unit can be implemented in the form of hardware or in the form of a software functional unit. It should be noted that the division of units in the present invention is schematic and is only a logical division. There may be other division methods in actual implementation.

[0154] In this embodiment, in order to realize the principles and beneficial effects of the method proposed in Example 1, the mobile cell resource minimization slice resource configuration system includes hardware structures and / or software modules corresponding to the execution of each function. Those skilled in the art should easily realize that, in combination with the various schematic units and algorithm steps described in the embodiments disclosed in the present invention, the present invention can be implemented in the form of hardware and / or a combination of hardware and computer software. Whether a function is executed in a hardware or computer software driven manner depends on the specific application and design constraints of the technical solution. Different methods can be used for each specific application to implement the described function, but such implementation should not be considered to exceed the scope of this application.

Claims

1. A method for configuring slice resources to minimize mobile cell resources, characterized in that: The following steps are involved: Build a train mobile network model based on a relay system and construct a resource slicing framework; Based on the constructed resource slicing framework, a communication and service quality assurance model is established for passenger on-demand services and beyond-line-of-sight services, taking into account the downlink carrying requirements of train intelligent services. Based on the constructed communication and service quality assurance model, an optimization problem with bandwidth allocation and terminal preemption grouping as target variables is established, where the goal is to minimize bandwidth resources; Based on the optimization problem of the target variable, the original problem is simplified using convex optimization theory. The optimal allocation strategy of the resource block allocation subproblem is derived through constraint relaxation. The mini-time slot preemption subproblem is converted into an objective function and auxiliary variables are introduced to simplify it into an easily solvable form. By using the efficient BCD-AP algorithm for resource block allocation and preemption based on block coordinate descent, the resource block allocation subproblem and the mini-time slot preemption subproblem are iteratively solved to obtain the optimal resource allocation strategy and complete the mobile cell resource minimization slice resource configuration.

2. The method for configuring slice resources for minimizing mobile cell resources according to claim 1, wherein: The train mobile network model based on the relay system is built to construct a resource slicing framework, which is specifically as follows: Establish a link between the base station and the rooftop relay, where the rooftop relay communicates with the terminal through the train station; Trackside cameras are used to capture real-time video data, which is then transmitted to a ground data processing center for processing. The processing results are then sent to the onboard DAS terminal, which includes a VDS terminal and a DAS terminal. The system bandwidth is divided into a set of resource blocks, denoted as , the bandwidth of a single resource block RB is expressed as , defined as the minimum resource unit allocated to the VDS terminal in the time-frequency domain, where k Indicates the number of resource blocks, K Indicates the number of resource blocks; Divide the time domain into time periods of The length of a single mini-slot is expressed as ,in, Indicates the number of mini-slots in a single millisecond time slot. Indicates the duration of a single mini-slot. Indicates the length of a single millisecond time slot; A single resource block (RB) is allocated to the VDS terminal using a standard time slot. Data packets arriving at the DAS terminal seize the bandwidth resources allocated to the VDS terminal and select available mini-time slots to transmit data, completing the construction of the resource slicing framework.

3. The method for configuring mobile cell resource minimization slice resources according to claim 2, characterized in that: The communication and service quality assurance model for passenger on-demand services and beyond-line-of-sight services is established as follows: Considering the downlink carrying demand of train intelligent services, a linear model is used to describe the relationship between the rate loss of passenger on-demand services and the number of mini-time slots occupied by assisted driving services. m The transmission data rate is: in, Indicates a VDS terminal m The transmission data rate, Indicates the bandwidth of a single resource block RB, Indicates that it is assigned to the VDS terminal m The total number of single resource blocks RB, Indicates DAS terminal n Seize VDS terminal m The number of mini-slots, Indicates a VDS terminal m The signal-to-noise ratio, Represents a collection of multiple VDS terminals in the system. represents the large-scale channel gain, Indicates a VDS terminal m The transmission power, Indicates channel state information, represents the additive white Gaussian noise power; Based on the described relationship, the time correlation in fast time-varying fading channels is used to construct a practical The statistical probability distribution of , where the conditional probability density function is expressed as follows: in, represents the conditional probability density function, Indicates the j The real small-scale fading of each terminal, represents the channel estimation error coefficient, represents the zero-order Bessel function of the first kind, represents the maximum Doppler shift, represents the channel estimation error coefficient, represents the channel feedback delay, Indicates the train speed, Indicates the carrier frequency, represents the speed of light, Indicates the j The estimated small-scale fading of the terminals, represents the additive white Gaussian noise power, represents the zero-order modified Bessel function of the first kind, represents the estimation error, Represents a collection of multiple VDS terminals in the system. Indicates that there is only one DAS terminal in the system; use The conditional probability density function of VDS terminal is obtained m The average data rate across all single resource blocks (RBs) completes the establishment of a model for passenger on-demand service communication and service quality assurance: in, Indicates a VDS terminal m The average data rate across all individual resource blocks (RBs), represents the conditional probability density function, Indicates the m Estimated small-scale fading of each VDS terminal, Indicates integral operation; By calculating the data rate of the DAS terminal, we can get the DAS terminal n The average data rate of According to DAS terminal n The average data rate is used to obtain the data transmission delay of the DAS terminal; According to the data transmission delay of the DAS terminal, the conditional risk value is used to capture the tail characteristics of the DAS transmission delay to obtain Function to complete the establishment of beyond-line-of-sight business communication and service quality assurance model.

4. The method for configuring slice resources for minimizing mobile cell resources according to claim 3, wherein: The expression of the data transmission delay of the DAS terminal is as follows: in, Indicates the data transmission delay of the DAS terminal, Indicates the length of the data packet of the DAS terminal. Indicates DAS terminal n The average data rate, Indicates the data rate of the DAS terminal, represents the conditional probability density function, Indicates DAS terminal n The real small-scale fading of Indicates DAS terminal n The estimated small-scale fading, Indicates terminal n The received signal-to-noise ratio, represents the large-scale channel gain, Indicates DAS terminal n transmission power.

5. The method for configuring slice resources for minimizing mobile cell resources according to claim 4, wherein: described The function expression is as follows: in, represents the C-VaR function, represents the delay threshold variable, Indicates the delay threshold. represents the expected value, and C-VaR represents the average value of potential delays exceeding the risk value.

6. The method for configuring slice resources for minimizing mobile cell resources according to claim 1, wherein: The bandwidth resource minimization problem is expressed as follows: in, represents the objective function, x and z are all optimization variable sets, x ={ } represents the set of allocated resource blocks, z ={ } represents the set of allocated mini-slot quantities, represents the delay threshold variable, Indicates that it is assigned to the VDS terminal m The total number of single resource blocks RB, Represents a collection of multiple VDS terminals in the system. Indicates the bandwidth of a single resource block RB, Indicates a VDS terminal m The transmission data rate, Indicates a VDS terminal m The minimum throughput QoS requirement, represents the C-VaR function, Indicates the data transmission delay of the DAS terminal, Indicates the maximum C-VaR risk threshold of the DAS terminal. C-VaR represents the average value of potential delay exceeding the risk value. represents the set of positive integers, Indicates DAS terminal n Seize VDS terminal m The number of mini-slots, Indicates the number of minislots in a single millisecond slot.

7. The method for configuring slice resources for minimizing mobile cell resources according to claim 1, wherein: The resource block allocation subproblem is expressed as follows: in, represents the objective function, x represents the set of optimization variables, x ={ } represents the set of allocated resource blocks, Indicates a VDS terminal m The transmission data rate, Indicates that it is assigned to the VDS terminal m The total number of single resource blocks RB, represents the set of positive integers, Indicates terminal m Optimal resource block allocation, Indicates a VDS terminal m The minimum throughput QoS requirement, Indicates the bandwidth of a single resource block RB, Indicates the transmission rate of VDS terminal m, Indicates DAS terminal n Seize VDS terminal m The number of mini-slots, Indicates the number of mini-slots in a single millisecond time slot. Indicates a VDS terminal m The signal-to-noise ratio, represents the conditional probability density function, Indicates channel state information, represents the channel estimation error coefficient, Indicates the m Estimated small-scale fading of each VDS terminal, represents the channel feedback delay, Represents an integral operation.

8. The method for configuring slice resources for minimizing mobile cell resources according to claim 1, wherein: The expression of the mini-slot preemption sub-problem is as follows: in, represents the subproblem objective function, z To optimize the variable set, z ={ } represents the set of allocated mini-slot quantities, represents the delay threshold variable, represents auxiliary variables, Indicates the delay threshold. Indicates the maximum C-VaR risk threshold of the DAS terminal. C-VaR represents the average value of potential delay exceeding the risk value. Express Seek hope, Indicates the expected value, Indicates the data transmission delay of the DAS terminal, Indicates that it is assigned to the VDS terminal m The total number of single resource blocks RB, Indicates the number of mini-slots in a single millisecond time slot. Represents a collection of multiple VDS terminals in the system. Indicates the number of mini-slots that the DAS terminal can preempt without affecting the VDS transmission rate. Indicates DAS terminal n Seize VDS terminal m The number of mini-slots, Indicates a VDS terminal m The minimum throughput QoS requirement, Indicates the bandwidth of a single resource block RB, Indicates a VDS terminal m The transmission rate, Indicates the length of the data packet of the DAS terminal. Indicates DAS terminal n The signal-to-noise ratio, Indicates DAS terminal n The real small-scale fading of represents the channel estimation error coefficient, represents the integral operation, Indicates DAS terminal n The estimated small-scale fading, Indicates the channel feedback delay.

9. A mobile cell resource minimization slice resource configuration system for executing the mobile cell resource minimization slice resource configuration method according to any one of claims 1 to 8, characterized in that: include: The first processing module is used to build a train mobile network model based on the relay system and construct a resource slicing framework; The second processing module is used to establish a communication and service quality assurance model for passenger on-demand services and beyond-line-of-sight services based on the constructed resource slicing framework, while considering the downlink carrying requirements of train intelligent services; The third processing module is used to establish an optimization problem with bandwidth allocation and terminal preemption grouping as target variables based on the constructed communication and service quality assurance model, wherein the goal is to minimize bandwidth resources; The fourth processing module is used to solve the optimization problem based on the target variable. It uses convex optimization theory to simplify the original problem, derives the optimal allocation strategy for the resource block allocation subproblem through constraint relaxation, and converts the mini-time slot preemption subproblem into an objective function, introduces auxiliary variables, and simplifies it into an easily solvable form. The fifth processing module is used to use the efficient BCD-AP algorithm for resource block allocation and preemption based on block coordinate descent to iteratively solve the resource block allocation subproblem and the mini-time slot preemption subproblem, obtain the optimal resource allocation strategy, and complete the mobile cell resource minimization slice resource configuration.

Images