A rail transit sensing and communication integrated system resource allocation method, device and equipment based on edge-cloud cooperation

Abstract

The application provides a rail transit communication and sensing integrated system resource allocation method, device and equipment based on end-edge-cloud cooperation, which comprises the following steps: establishing a rail transit communication and sensing integrated system model based on end-edge-cloud cooperation; based on the rail transit communication and sensing integrated system model based on end-edge-cloud cooperation, a target model is established; the target model is used to maximize the communication rate while ensuring the sensing rate; and according to the target model, the communication bandwidth and the sensing bandwidth are determined. The method of the application embodiment realizes the dynamic division of the communication and sensing sub-bandwidth in the total bandwidth of the base station, achieves the dynamic balance of the maximum communication rate and the sensing performance, and effectively solves the problems of efficient utilization of spectrum resources and reliable service guarantee in the high-speed moving scene of rail transit.

Description

Technical Field

[0001] This invention relates to the field of wireless communication technology, and in particular to a resource allocation method, apparatus and equipment for an integrated sensing system for rail transit based on end-edge-cloud collaboration. Background Technology

[0002] Integrated Sensing and Communication (ISAC) technology deeply integrates communication and sensing functions into a single system, enabling coordinated data transmission and real-time monitoring and sensing of the surrounding environment. This has demonstrated significant application value in the rail transit sector. As a high-density, highly dynamic, and complex transportation scenario, the rail system's operations include both communication and sensing services. Communication services encompass the transmission of dispatching instructions between trains and ground control centers, serving as a core support for ensuring safe train operation and improving operational efficiency. Sensing services include intelligent obstacle detection for trains, specifically trackside obstacle identification and trackside equipment status monitoring, directly impacting train safety and the timeliness of track maintenance.

[0003] However, existing ISAC systems still face numerous challenges in practical applications within orbital scenarios. On the one hand, critical services such as scheduling instructions and emergency alarms impose stringent requirements on transmission latency; on the other hand, communication and sensing services share limited wireless spectrum resources, leading to intense resource competition among multiple services. Communication services require sufficient bandwidth to ensure the continuity and reliability of data transmission and prevent scheduling failures due to signal interruptions; sensing services, on the other hand, have specific requirements for bandwidth stability and real-time performance. For example, obstacle recognition relies on continuous spectrum resource support to ensure continuous monitoring. In this context, how to rationally allocate bandwidth proportions within the limited wireless spectrum resources in complex orbital scenarios to meet the differentiated needs of communication and sensing services and balance resource competition has become a key issue restricting the large-scale application of ISAC technology in orbital systems. Summary of the Invention

[0004] This invention provides a resource allocation method, apparatus, and equipment for an integrated sensing system for rail transit based on edge-cloud collaboration. It realizes the dynamic division of communication and sensing sub-bandwidths in the total bandwidth of the base station, achieving a dynamic balance between maximizing communication rate and sensing performance, and effectively solving the problems of efficient utilization of spectrum resources and ensuring service reliability in high-speed rail transit scenarios.

[0005] This invention provides a resource allocation method for an integrated sensing system for rail transit based on edge-cloud collaboration, comprising the following steps.

[0006] Establish a collaborative edge-cloud integrated sensing system model for rail transit; Based on the aforementioned edge-cloud collaborative integrated rail transit sensing system model, a target model is established; the target model is used to maximize the communication rate while ensuring the sensing rate. Based on the target model, determine the communication bandwidth and sensing bandwidth.

[0007] According to the present invention, a resource allocation method for an integrated sensing system for rail transit based on edge-cloud collaboration is provided, wherein the target model includes: in, Indicates communication rate; Indicates time slot The perception rate must be greater than or equal to the perception rate threshold. For bandwidth allocation ratio, Indicates communication bandwidth; This represents the perceived bandwidth.

[0008] According to the present invention, a resource allocation method for an integrated sensing system for rail transit based on edge-cloud collaboration is provided, which determines the communication bandwidth and sensing bandwidth based on the target model, including: Based on the deep reinforcement learning method and the target model, the target bandwidth ratio is determined. Based on the target bandwidth ratio and base station bandwidth, the communication bandwidth and sensing bandwidth are determined; different reward values ​​are set for different bandwidth ratios in the deep reinforcement learning method; the reward values ​​include: in, Indicates the reward value; This represents the sensing rate corresponding to the sensing bandwidth in the current time slot; Indicates the sensing rate threshold; Represents a constant.

[0009] According to the present invention, a resource allocation method for an integrated sensing system for rail transit based on edge-cloud collaboration is provided, wherein the communication rate in the target model is determined based on the following method: in, Indicates the communication rate of the current time slot; Indicates communication bandwidth; Indicates the transmit antenna gain; This indicates the receiving antenna gain at the vehicle-mounted wireless unit. Indicates power spectral density; Indicates the wavelength of the signal; This indicates the distance between the transmitting antenna and the vehicle-mounted wireless unit; This represents the noise power spectral density at the vehicle-mounted wireless unit. The sensing rate in the target model is determined based on the following method: in, Indicates the sensing rate of the current time slot; Indicates the pulse repetition interval of the sensed signal; Represents a constant term; For sensing bandwidth.

[0010] According to the present invention, a resource allocation method for a rail transit sensing integrated system based on edge-cloud collaboration is provided, wherein the rail transit sensing integrated system model based on edge-cloud collaboration includes: Train terminal, onboard wireless unit, base station and central node; The base station and the vehicle-mounted wireless unit and train terminal use a two-hop communication architecture.

[0011] According to the present invention, a resource allocation method for an integrated sensing system for rail transit based on edge-cloud collaboration is provided, wherein the base station is used to send sensing information to a central node; the central node is used to send emergency control commands to a terminal; and the terminal is used to execute the emergency control commands.

[0012] This invention provides a resource allocation device for an integrated sensing system for rail transit based on edge-cloud collaboration, comprising the following modules: The first module is used to establish a model of an integrated sensing system for rail transit that combines edge, cloud, and terminal operations. The second establishment module is used to establish a target model based on the edge-cloud collaborative rail transit sensing integrated system model; the target model is used to maximize the communication rate while ensuring the sensing rate. The allocation module is used to determine the communication bandwidth and sensing bandwidth based on the target model.

[0013] The present invention provides an electronic device, including a memory, a processor, and a computer program stored in the memory and executable on the processor. When the processor executes the program, it implements the resource allocation method for an integrated rail transit sensing system based on edge-cloud collaboration as described above.

[0014] The present invention provides a non-transitory computer-readable storage medium storing a computer program thereon, which, when executed by a processor, implements the resource allocation method for an integrated rail transit sensing system based on edge-cloud collaboration as described above.

[0015] The present invention also provides a computer program product, including a computer program that, when executed by a processor, implements the resource allocation method for an integrated rail transit sensing system based on edge-cloud collaboration as described above.

[0016] The present invention provides a resource allocation method, apparatus, and equipment for an integrated sensing system for rail transit based on edge-cloud collaboration. By establishing an integrated sensing system model and target model for rail transit based on edge-cloud collaboration, it maximizes the communication rate while ensuring the sensing rate. It achieves dynamic allocation of communication and sensing sub-bandwidths in the total bandwidth of the base station, thus achieving a dynamic balance between maximizing the communication rate and sensing performance. This effectively solves the problem of efficient utilization of spectrum resources and ensuring service reliability in high-speed rail transit scenarios. Attached Figure Description

[0017] To more clearly illustrate the technical solutions in this invention or the prior art, the accompanying drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are some embodiments of this invention. For those skilled in the art, other drawings can be obtained based on these drawings without any creative effort.

[0018] Figure 1 This is one of the flowcharts illustrating the resource allocation method for an integrated sensing system for rail transit based on edge-cloud collaboration provided by this invention.

[0019] Figure 2 This is a schematic diagram of the integrated sensing system for rail transit based on edge-cloud collaboration provided by the present invention.

[0020] Figure 3 This is the second flowchart of the resource allocation method for the integrated sensing system of rail transit based on edge-cloud collaboration provided by the present invention.

[0021] Figure 4 This is a schematic diagram of the resource allocation device for an integrated sensing system for rail transit based on edge-cloud collaboration provided by the present invention.

[0022] Figure 5 This is a schematic diagram of the structure of the electronic device provided by the present invention. Detailed Implementation

[0023] To make the objectives, technical solutions, and advantages of this invention clearer, the technical solutions of this invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some, not all, of the embodiments of this invention. All other embodiments obtained by those skilled in the art based on the embodiments of this invention without creative effort are within the scope of protection of this invention.

[0024] The following is combined with Figures 1-5 This invention describes a resource allocation method, apparatus, and device for an integrated sensing system for rail transit based on edge-cloud collaboration.

[0025] To facilitate a clearer understanding of the technical solutions of the various embodiments of this application, some technical content related to the various embodiments of this application will be introduced first.

[0026] In today's highly information-driven society, wireless communication technology, as a key infrastructure supporting various intelligent systems, is of paramount importance in terms of reliability and efficiency. In the rail transit sector, with the continuous increase in train speeds, the dramatic rise in the number of onboard devices, and the ever-increasing demands for communication service quality from various vehicle systems, traditional wireless communication systems face unprecedented challenges. On the one hand, limited wireless spectrum resources are insufficient to meet the demands of simultaneous access and data transmission from a large number of devices, exacerbating the spectrum shortage problem. On the other hand, during high-speed train operation, the limited coverage of base stations necessitates frequent handovers, causing the wireless channel to exhibit significant time-varying characteristics, severely impacting communication reliability and stability. Furthermore, ensuring the safe operation of rail transit is also crucial, requiring real-time detection of potential safety threats along the track, such as foreign object intrusion, and timely response.

[0027] Integrated sensing and communication (ISAC) technology, as an emerging research direction in recent years, offers a new approach to solving the aforementioned problems. ISAC technology integrates communication and sensing functions into a single system, enabling the system to simultaneously transmit data and perform real-time monitoring and sensing of the surrounding environment. In rail transit scenarios, by deploying ISAC-enabled base stations (BSs) along the track, not only can conventional communication tasks such as train operation control, wireless dispatch communication, and user data traffic management be achieved, but the system can also continuously receive ISAC sensing echo signals from multiple potential ground targets, enabling real-time monitoring of the trackside environment. However, existing ISAC systems still have many shortcomings in practical applications. For example, in terms of resource allocation, facing the rapid channel changes and stringent latency requirements caused by high-speed train operation, how to rationally allocate bandwidth between communication and sensing services within limited wireless spectrum resources has become a critical issue that urgently needs to be addressed. Traditional resource allocation methods often only focus on communication performance, neglecting the spectrum requirements of sensing tasks, leading to insufficient sensing accuracy or decreased communication reliability. In rail transit scenarios, base stations need to dynamically adjust the allocation strategy of communication bandwidth and sensing bandwidth to adapt to multiple factors such as train position changes, channel status, and sensing task priorities, so as to achieve coordinated optimization of communication and sensing and improve the overall efficiency and robustness of the system.

[0028] Meanwhile, to ensure stable and reliable communication services for onboard equipment during high-speed train operation, deploying onboard wireless units on the roof of each carriage to construct a two-hop communication architecture has become an effective solution. As data receiving and forwarding nodes, the onboard wireless units undertake the communication tasks between the BSS (Browser Support System) and the onboard equipment, effectively avoiding severe signal penetration loss caused by the metal structure of the train body. However, existing onboard wireless unit technology also faces some challenges in practical applications, such as: how to optimize the resource allocation strategy of onboard wireless units to improve the overall system throughput and communication quality; how to reduce frequent switching of onboard terminals during high-speed movement, reducing connection interruptions and latency. Furthermore, in terms of safety management in rail transit, existing systems mainly rely on manual inspections and single-sensor early warning, resulting in slow response speeds and low accuracy, making it difficult to meet the needs of real-time monitoring and rapid response. By constructing an edge-cloud architecture based on ISAC technology, and organically combining train terminals (ends), roadside nodes such as BSs (edges), and central nodes such as the Train Control Center (TCC) and Radio Block Center (RBC), global collaborative decision-making and scheduling optimization can be achieved, improving rail transit safety and efficiency. However, there is currently no mature technical solution that can effectively solve the resource allocation problem under the edge-cloud collaborative architecture.

[0029] Figure 1 This is one of the flowcharts illustrating the resource allocation method for an integrated sensing system for rail transit based on edge-cloud collaboration provided by the present invention. The method includes the following: Step 101: Establish a model of an integrated sensing system for rail transit with end-edge-cloud collaboration.

[0030] Specifically, this application embodiment first establishes an edge-cloud collaborative integrated sensing system model for rail transit. Optionally, to achieve high reliability of wireless communication, multiple BSs with ISAC (Inter-Signal Communication Control) functionality are deployed along the track. These BSs not only undertake tasks such as train operation control, wireless dispatch communication, and user data traffic management, but also continuously receive ISAC sensing echo signals from multiple potential ground targets. Preliminary target location can rely on various early warning devices deployed in the rail transit system, such as GPS-based mobile alarm devices installed beside the track, or lidar systems used to monitor potentially dangerous road sections. When abnormal situations such as foreign object intrusion occur, these devices will trigger alarm conditions, forming the first line of defense. During periods when no trains are passing, the BSs will actively perform time-division beam scanning to continuously sense and track targets in key areas that may affect train safety (such as landslide-prone areas, areas where people or foreign objects may intrude). Once a risky target or abnormal situation is detected, the BSs immediately upload the information to the RBC (Rail-Based Controller) through the edge network, enabling rapid local response and collaborative processing. Upon receiving risk information, the RBC can directly communicate with the train via dedicated track signal channels or railway communication base stations (BSs). Simultaneously, the TCC can issue emergency control commands to the RBC or relevant signaling equipment (such as active transponders) based on large-scale data analysis results and overall operation scheduling strategies in the cloud. If necessary, it can issue emergency braking commands to trains about to enter hazardous areas via the BSs. Through this architecture, a highly reliable rail transit ISAC system is constructed, centered on "end-edge-cloud" collaboration: the train terminal (end) enables rapid response to operational status and commands; roadside nodes (edges) such as BSs are responsible for real-time communication sensing and event reporting; and central nodes (clouds) such as the TCC and RBC undertake global collaborative decision-making and scheduling optimization, providing strong guarantees for rail transit safety and efficiency.

[0031] Step 102: Establish a target model for the integrated sensing system of rail transit based on edge-cloud collaboration; the target model is used to maximize the communication rate while ensuring the sensing rate.

[0032] Specifically, after establishing a collaborative edge-cloud integrated rail transit sensing system model, a target model can be established based on this model. The target model aims to maximize the communication rate while ensuring the sensing rate, thereby achieving a dynamic balance between maximizing the communication rate and sensing performance.

[0033] Step 103: Determine the communication bandwidth and sensing bandwidth based on the target model.

[0034] Specifically, after establishing the target model, by solving the target model, the communication bandwidth and sensing bandwidth that ensure the sensing rate and maximize the communication rate can be determined. This enables the dynamic allocation of the communication and sensing sub-bandwidths in the total bandwidth of the base station, achieving a dynamic balance between maximizing the communication rate and sensing performance. This solves the problem of efficient utilization of spectrum resources and ensuring service reliability in high-speed rail transit scenarios.

[0035] The method described above establishes an edge-cloud collaborative integrated rail transit sensing system model and target model, thereby maximizing the communication rate while ensuring the sensing rate. It achieves dynamic allocation of communication and sensing sub-bandwidths in the total bandwidth of the base station, achieving a dynamic balance between maximizing the communication rate and sensing performance. This effectively solves the problem of efficient utilization of spectrum resources and ensuring service reliability in high-speed rail transit scenarios.

[0036] In some embodiments, the target model includes: in, Indicates time slot Communication rate at that time; Indicates time slot The perception rate is greater than or equal to the perception rate threshold. Indicates the bandwidth allocation ratio; Indicates communication bandwidth; This represents the perceived bandwidth.

[0037] Specifically, in the edge-cloud collaborative integrated sensing system model for rail transit, it is assumed that the total system operation time is... Time slot index The total bandwidth allocated by the base station to the train is set to... Divide it into communication bandwidth and sensing bandwidth Among them, the sub-bandwidth of communication is part of the total bandwidth. The proportion of the middle is ,Right now Then, the perceived bandwidth To ensure reasonable bandwidth allocation, the sum of the communication bandwidth and the sensing bandwidth must not exceed the total bandwidth, i.e., satisfy... Assume the power spectral density is (Unit: W / Hz), based on the relationship between power spectral density and bandwidth, the transmit power of the communication signal is: The transmission power of the sensing signal is: .

[0038] To simultaneously meet the performance requirements of communication and sensing services, it is necessary to ensure that the sensing rate is not lower than a specified sensing threshold. Under the premise of optimizing the bandwidth allocation ratio To maximize the data transmission rate of the communication link. This problem can be formally modeled as the following optimization problem: Among them, constraint C1 ensures that the sensing rate of the system in any time slot is not lower than the minimum sensing performance requirement. To ensure the accuracy and real-time performance of target perception; constraints C2 and C3 are determined by scaling factors. Control total bandwidth The division between communication and sensing components; constraint C4 limits the bandwidth allocation ratio. The value is in the range of (0,1) to ensure that both communication and sensing are allocated effective spectrum resources.

[0039] The method described above establishes a target model based on the edge-cloud collaborative integrated rail transit sensing system model, thereby maximizing the communication rate while ensuring the sensing rate, achieving a dynamic balance between maximizing the communication rate and sensing performance.

[0040] In some embodiments, determining the communication bandwidth and sensing bandwidth based on the target model includes: Based on deep reinforcement learning methods and target models, the target bandwidth ratio is determined. Based on the target bandwidth ratio and base station bandwidth, the communication bandwidth and sensing bandwidth are determined; different reward values ​​are set for different bandwidth ratios in deep reinforcement learning methods; the reward values ​​include: in, Indicates the reward value; This represents the sensing rate corresponding to the sensing bandwidth in the current time slot; Indicates the sensing rate threshold; Represents a constant.

[0041] Specifically, in the embodiments of this application, after establishing the target model of the integrated sensing system for rail transit based on edge-cloud collaboration, the problem is difficult to solve using traditional convex optimization methods due to the presence of nonlinear terms and coupling relationships between multiple variables in the objective function. To address this challenge, this application introduces a deep reinforcement learning method, utilizing the interactive learning between the agent and the environment to dynamically adjust the bandwidth allocation ratio. This achieves an efficient trade-off between communication performance and sensing performance.

[0042] (1) State Design: In order for the agent to make effective decisions, the state space needs to fully reflect the key information of the system in the current time slot. Let the first state be... The state vector of each time slot is The definition is as follows: The meanings of each element are as follows: The distance between the BS and the vehicle-mounted radio unit in the current time slot affects path loss and received power. The bandwidth allocation ratio of the previous time slot; The communication and sensing bandwidth value of the previous time slot is used to estimate resource utilization. The transmission rate of the sensing and communication link in the previous time slot is used to measure the effectiveness of the previous decision.

[0043] (2) Motion design: In the discrete motion modeling scheme, the continuous bandwidth allocation ratio is determined. Quantization is performed to construct a finite-dimensional discrete action set. The specific steps are as follows: Will Evenly divided into Construct a discrete action set based on a set of value points: In this setting, the agent selects one action from the set of discrete actions in each time slot. This is the current bandwidth allocation ratio.

[0044] (3) Reward Design: To improve the learning efficiency of the agent, we set different reward functions based on the agent's actions, mainly calculating the reward value based on the difference between the perception rate and the threshold. We use the Sigmoid function to design the reward function and map the reward value to a specific interval through a linear transformation, which can be expressed as: in, and These represent the sensing rate and sensing threshold of the current time slot, respectively. When the sensing rate is less than the threshold, the function output is in the range [0, 1]. If they are equal, the maximum reward value is a constant. If the perception rate exceeds a threshold, the function output gradually decreases as the perception rate increases. During training, we use a neural network to estimate the action value function for each state, learning the optimal policy by maximizing the long-term cumulative reward. In each time slot, the agent constructs a state vector based on current perception bandwidth, perception rate, and communication rate. Based on the current state, the agent uses... A greedy policy selects actions and interacts with the environment to update the perception bandwidth. By obtaining feedback from the environment, the actual communication rate and perception rate are obtained. Then, a reward value is calculated according to a predefined reward function, and the next state is obtained. The current state, action, reward, and next state are stored in an experience replay buffer, and a batch of data is randomly sampled to update the network parameters. Specifically, we use a target Q-network to calculate the target value and update the weights of the Q-network using a mean squared error loss function to make it approximate the optimal value function of the current policy. The above steps are repeated to iteratively train the agent's policy. After training, the trained Q-network is used to predict the policy, selecting the optimal action based on the current state, i.e., an appropriate bandwidth resource allocation ratio, to maximize the communication rate.

[0045] In other words, this application solves the non-convex optimization problem in resource allocation by deeply integrating communication and sensing functions and using deep reinforcement learning algorithms, thereby improving the overall system efficiency and meeting the reliable communication and security requirements of high-speed rail transit scenarios.

[0046] In some embodiments, the communication rate in the target model is determined based on the following: in, Indicates the communication rate of the current time slot; Indicates communication bandwidth; Indicates the transmit antenna gain; This indicates the receiving antenna gain at the vehicle-mounted wireless unit. Indicates power spectral density; Indicates the wavelength of the signal; This indicates the distance between the transmitting antenna and the vehicle-mounted wireless unit; This represents the noise power spectral density at the vehicle-mounted wireless unit. The sensing rate in the target model is determined based on the following method: in, Indicates the sensing rate of the current time slot; Indicates the pulse repetition interval of the sensed signal; Represents a constant term; For sensing bandwidth.

[0047] Specifically, based on the wireless propagation model, the received power at the receiving end of the vehicle-mounted wireless unit in the communication link can be modeled as follows: in, It is the transmit antenna gain. It refers to the receiving antenna gain at the vehicle-mounted wireless unit. It is the wavelength of the signal. This is the distance between the transmitting antenna and the vehicle-mounted wireless unit. For the sensing link, according to the radar equation, the radar echo power received by the base station is: in, This indicates the antenna gain of the base station used to receive radar echoes. Let be the scattering cross-section of the target, used to measure its reflection characteristics. This model considers the two-way path loss during the round-trip propagation of the sensing signal. According to Shannon's theorem, given the signal and noise conditions, then at time... The maximum achievable data rate of the communication link is: in, This represents the noise power spectral density at the vehicle-mounted wireless unit. For sensing links, their performance can be measured by the mutual information that the system can acquire within each pulse repetition period.

[0048] Let the mutual information be ,in Represents the transmitted signal. The sensing rate is defined as follows: (This represents additive white Gaussian noise (AWGN) in radar.) in, This represents the pulse repetition interval of the sensed signal. Assuming that the estimation error of the radar echo and the AWGN generated during radar sensing both follow the same Gaussian distribution, given the signal-to-noise ratio (SNR) received by the radar, the sensing rate can be derived as follows: in, This represents the overall signal-to-noise ratio along the sensing link. Indicates the pulse duration. This represents the power spectral density of the AWGN at the base station; Let Variance be the variance of the radar system's AWGN; Represents the root mean square bandwidth of the sensed signal; This represents a scaling factor associated with the waveform, the value of which depends on the shape of the radar waveform's power spectral density. Additionally, the parameter... , and The relationship between them satisfies the following: This formula describes the influence of the power spectral density shape of the radar waveform on information transmission efficiency and reflects the constraint of waveform design on sensing performance.

[0049] Combining the radar receive power model and the sensing rate to derive the upper bound, we can obtain the result in the first... When there are multiple time slots, the data rate that can be achieved by sensing is expressed as follows: To further simplify the above equation, we define the constant term as: This allows for a more concise expression: in, .

[0050] The method described in the above embodiments accurately represents the communication rate and sensing rate in the edge-cloud collaborative rail transit sensing system model. This allows for maximizing the communication rate while ensuring the sensing rate, thus achieving a dynamic balance between maximizing the communication rate and sensing performance.

[0051] In some embodiments, the edge-cloud collaborative integrated rail transit sensing system model includes: Train terminal, onboard wireless unit, base station and central node; The communication between the base station, the vehicle-mounted wireless unit, and the train terminal is a two-hop communication architecture.

[0052] Specifically, in this embodiment of the application, during the high-speed operation of the train, in order to ensure that the onboard equipment obtains stable and reliable communication services, a [device / system] is deployed on the top of the train. A multi-vehicle wireless unit (VWN) architecture was constructed, forming a typical two-hop communication system. The VWN primarily handles data reception and forwarding; all communication between the BSS and vehicle-mounted equipment is conducted through the VWN, effectively avoiding severe signal penetration loss caused by the vehicle's metal structure. Compared to traditional single-VWN schemes, the multi-VWN architecture fully utilizes spatial diversity gain, significantly improving the overall system throughput.

[0053] In the first hop, the onboard radio unit establishes a connection with the base stations (BSs) deployed along the track via a wireless access link. In the second hop, the onboard radio unit provides access services to the onboard equipment terminals through wired connections (such as network cables or fiber optics) with the access points (APs) in each carriage. This two-hop architecture also has the significant advantage of reducing frequent handovers: communication requests and data from onboard equipment first converge to the onboard radio unit located in the carriage, which then communicates with the BSs, avoiding connection interruptions or delays caused by frequent handovers between onboard equipment terminals and ground BSs due to the high-speed movement of the train.

[0054] The method described in the above embodiments uses an on-board wireless unit to construct a two-hop communication architecture and provides an anti-interference mechanism to enhance communication stability, effectively improve communication reliability, avoid frequent handovers and multipath fading that cause unstable communication quality, and can also effectively solve the problem of frequent handovers caused by high-speed train movement.

[0055] In some embodiments, the base station is used to send sensing information to the central node; the central node is used to send emergency control commands to the terminal; and the terminal is used to execute the emergency control commands.

[0056] Specifically, such as Figure 2 and Figure 3 As shown in this embodiment, to achieve high reliability of wireless communication, multiple BSs with ISAC functionality are deployed along the track. These BSs not only undertake tasks such as train operation control, wireless dispatch communication, and user data traffic management, but also continuously receive ISAC sensing echo signals from multiple potential ground targets. Preliminary target location can rely on various early warning devices deployed in the rail transit system, such as GPS-based mobile alarm devices installed beside the track, or lidar systems used to monitor potentially dangerous sections. When abnormal situations such as foreign object intrusion occur, these devices will trigger alarm conditions, forming the first line of defense. During periods when no trains are passing, the BSs will actively perform time-division beam scanning to continuously sense and track targets in key areas that may affect train safety (such as landslide-prone areas, areas with personnel or foreign object intrusion). Once a risky target or abnormal situation is detected, the BSs immediately upload the information to the RBC via the edge network, enabling rapid local response and collaborative processing. After receiving the risk information, the RBC can directly communicate with the train through the dedicated track signal channel or railway communication BSs. Meanwhile, the TCC can issue emergency control commands to the RBC or related signaling equipment (such as active transponders) based on the results of large-scale data analysis and overall operation scheduling strategies in the cloud. When necessary, it can issue emergency braking commands to trains that are about to enter the danger zone through the BSs, thereby achieving efficient collaboration under the edge-cloud architecture and providing strong support for the safe and efficient operation of rail transit.

[0057] The method described in the above embodiments constructs an edge-cloud collaborative architecture to achieve global real-time perception and rapid decision-making, effectively solving the problems of slow response speed and low accuracy caused by existing systems relying on manual inspection and single sensor early warning.

[0058] It should be noted that, in terms of bandwidth resource allocation, the total bandwidth of the base station must simultaneously support communication services (equipment data) and sensing services (track safety monitoring). Traditional static allocation cannot adapt to dynamic changes in train position, making it difficult to balance communication and sensing speeds. This application aims to address the core issues faced by rail transit communication systems, such as limited spectrum resources, insufficient communication reliability, lagging safety management, and difficulties in end-edge-cloud architecture collaboration. By utilizing deep reinforcement learning artificial intelligence technology, it rationally allocates limited wireless spectrum resources, optimizes the performance of the BSS-vehicle wireless unit link, improves the system's spectrum efficiency and anti-interference capabilities, and simultaneously achieves efficient collaboration under an end-edge-cloud architecture, providing strong support for the safe and efficient operation of rail transit.

[0059] The following describes the resource allocation device for a rail transit sensory integration system based on edge-cloud collaboration provided by the present invention. The resource allocation device for a rail transit sensory integration system based on edge-cloud collaboration described below can be referred to in correspondence with the resource allocation method for a rail transit sensory integration system based on edge-cloud collaboration described above. The resource allocation device for a rail transit sensory integration system based on edge-cloud collaboration in the embodiments of this application is as follows: Figure 4 As shown, it includes: The first module 410 is used to establish a model of an integrated sensing system for rail transit with edge-cloud collaboration. The second module 420 is used to establish a target model for the integrated sensing system model of rail transit based on edge-cloud collaboration. The target model is used to maximize the communication rate while ensuring the sensing rate. The allocation module 430 is used to determine the communication bandwidth and sensing bandwidth based on the target model.

[0060] Figure 5 A schematic diagram of the physical structure of an electronic device is provided. This electronic device may include a processor 510, a communications interface 520, a memory 530, and a communication bus 540. The processor 510, communications interface 520, and memory 530 communicate with each other via the communication bus 540. The processor 510 can call logical instructions from the memory 530 to execute a resource allocation method for a rail transit sensing integrated system based on edge-cloud collaboration. This method includes: establishing a rail transit sensing integrated system model based on edge-cloud collaboration; establishing a target model based on the rail transit sensing integrated system model; the target model is used to maximize the communication rate while ensuring the sensing rate; and determining the communication bandwidth and sensing bandwidth based on the target model.

[0061] Furthermore, the logical instructions in the aforementioned memory 530 can be implemented as software functional units and, when sold or used as independent products, can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present invention can essentially be embodied in the form of a software product, and its improvements or contributions to the prior art can also be presented in software. This computer software product is stored in a storage medium and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute all or part of the steps of the methods described in the various embodiments of the present invention. The aforementioned storage medium includes various media capable of storing program code, such as USB flash drives, portable hard drives, read-only memory (ROM), random access memory (RAM), magnetic disks, or optical disks.

[0062] On the other hand, the present invention also provides a computer program product, which includes a computer program that can be stored on a non-transitory computer-readable storage medium. When the computer program is executed by a processor, the computer can execute the resource allocation method for an integrated rail transit sensing system based on edge-cloud collaboration provided by the above methods. The method includes: establishing an integrated rail transit sensing system model based on edge-cloud collaboration; establishing a target model based on the integrated rail transit sensing system model based on edge-cloud collaboration; the target model is used to maximize the communication rate while ensuring the sensing rate; and determining the communication bandwidth and sensing bandwidth according to the target model.

[0063] Finally, the present invention also provides a non-transitory computer-readable storage medium storing a computer program thereon, which, when executed by a processor, implements the resource allocation method for a rail transit sensing integrated system based on edge-cloud collaboration provided by the above methods. The method includes: establishing a rail transit sensing integrated system model based on edge-cloud collaboration; establishing a target model based on the rail transit sensing integrated system model based on edge-cloud collaboration; the target model is used to maximize the communication rate while ensuring the sensing rate; and determining the communication bandwidth and sensing bandwidth according to the target model.

[0064] The device embodiments described above are merely illustrative. The units described as separate components may or may not be physically separate. The components shown as units may or may not be physical units; that is, they may be located in one place or distributed across multiple network units. Some or all of the modules can be selected to achieve the purpose of this embodiment according to actual needs. Those skilled in the art can understand and implement this without any creative effort.

[0065] Through the above description of the embodiments, those skilled in the art can clearly understand that each embodiment can be implemented by means of software plus necessary general-purpose hardware platforms, and of course, it can also be implemented by hardware. Based on this understanding, the above technical solutions, in essence or the part that contributes to the prior art, can be embodied in the form of a software product. This computer software product can be stored in a computer-readable storage medium, such as ROM / RAM, magnetic disk, optical disk, etc., and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute the methods described in the various embodiments or some parts of the embodiments.

[0066] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, and not to limit them; although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some of the technical features; and these modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of the present invention.

Claims

1. A resource allocation method for a rail transit sensor-integrated system based on edge-cloud collaboration, characterized in that, include: Establish a collaborative edge-cloud integrated sensing system model for rail transit; Based on the aforementioned edge-cloud collaborative integrated rail transit sensing system model, a target model is established; The target model is used to maximize the communication rate while ensuring the sensing rate; Based on the target model, determine the communication bandwidth and sensing bandwidth; The target model includes: ， ， ， ， ； in, Indicates time slot Communication rate at that time; Indicates time slot The perception rate must be greater than or equal to the perception rate threshold. For bandwidth allocation ratio, Indicates communication bandwidth; Indicates the perceived bandwidth; Based on the target model, the communication bandwidth and sensing bandwidth are determined, including: Based on the deep reinforcement learning method and the target model, the target bandwidth ratio is determined. Based on the target bandwidth ratio and base station bandwidth, the communication bandwidth and sensing bandwidth are determined; different reward values ​​are set for different bandwidth ratios in the deep reinforcement learning method; the reward values ​​include: ； in, Indicates the reward value; This represents the sensing rate corresponding to the sensing bandwidth in the current time slot; Indicates the sensing rate threshold; Represents a constant; The communication rate in the target model is determined based on the following method: ； in, Indicates the communication rate of the current time slot; Indicates communication bandwidth; Indicates the transmit antenna gain; This indicates the receiving antenna gain at the vehicle-mounted wireless unit. Indicates power spectral density; Indicates the wavelength of the signal; This indicates the distance between the transmitting antenna and the vehicle-mounted wireless unit; This represents the noise power spectral density at the vehicle-mounted wireless unit. The sensing rate in the target model is determined based on the following method: ； in, Indicates the sensing rate of the current time slot; Indicates the pulse repetition interval of the sensed signal; Represents a constant term; For sensing bandwidth.

2. The resource allocation method for an integrated sensing system for rail transit based on edge-cloud collaboration as described in claim 1, characterized in that, The edge-cloud collaborative integrated rail transit sensing system model includes: Train terminal, onboard wireless unit, base station and central node; The base station and the vehicle-mounted wireless unit and train terminal use a two-hop communication architecture.

3. The resource allocation method for an integrated rail transit sensing system based on edge-cloud collaboration according to claim 2, characterized in that, The base station is used to send sensing information to the central node; the central node is used to send emergency control commands to the terminal; and the terminal is used to execute the emergency control commands.

4. A resource allocation device for a rail transit sensory integration system based on edge-cloud collaboration, used to implement the resource allocation method for a rail transit sensory integration system based on edge-cloud collaboration as described in any one of claims 1 to 3, characterized in that, include: The first module is used to establish a model of an integrated sensing system for rail transit that combines edge, cloud, and terminal operations. The second establishment module is used to establish a target model based on the edge-cloud collaborative rail transit sensing integrated system model; the target model is used to maximize the communication rate while ensuring the sensing rate. The allocation module is used to determine the communication bandwidth and sensing bandwidth based on the target model.

5. An electronic device, comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that, When the processor executes the computer program, it implements the resource allocation method for the integrated sensing system of rail transit based on edge-cloud collaboration as described in any one of claims 1 to 3.

6. A non-transitory computer-readable storage medium having a computer program stored thereon, characterized in that, When the computer program is executed by the processor, it implements the resource allocation method for the integrated sensing system of rail transit based on edge-cloud collaboration as described in any one of claims 1 to 3.

7. A computer program product, comprising a computer program, characterized in that, When the computer program is executed by the processor, it implements the resource allocation method for the integrated sensing system of rail transit based on edge-cloud collaboration as described in any one of claims 1 to 3.

Images