Rail transit train-ground communication switching method and system based on heterogeneous network cooperation

Abstract

The application provides a rail transit train-ground communication switching method and system based on heterogeneous network cooperation, and relates to the technical field of traffic communication, which comprises the following steps: collecting real-time running parameters of a train and signal quality parameters of a currently accessed network, extracting signal attenuation trend characteristics through time sequence analysis and mapping a predicted position trajectory to determine an available switching time window; performing parallel channel detection on candidate heterogeneous networks within the time window, calculating coverage duration based on the predicted position trajectory and network coverage information, solving a network switching priority sequence in combination with real-time channel quality parameters, and determining a target access network in combination with load parameter evaluation; establishing a double-layer switching link between the current access network and the target access network, sending a switching request through a control signaling channel and receiving a resource reservation confirmation, generating a transmission permission switching instruction based on a switching preparation completion identifier and executing the instruction, so as to reduce switching delay and improve switching reliability.

Description

Technical Field

[0001] This invention relates to the field of transportation communication technology, and in particular to a method and system for switching vehicle-to-ground communication in rail transit based on heterogeneous network collaboration. Background Technology

[0002] Rail transit vehicle-to-ground communication systems operate in high-speed mobile environments for extended periods. During train operation, trains need to cross multiple ground base station coverage areas. The reliability and real-time performance of vehicle-to-ground communication link switching directly determine the transmission quality of train control signaling, train status monitoring data, and passenger service information. With the evolution of communication technology systems, multiple wireless access networks of various standards are typically deployed along the railway line to ensure coverage continuity. These heterogeneous networks exhibit significant differences in coverage, transmission capacity, and quality of service indicators. When trains move between different network coverage sections, access point changes need to be completed in the network overlap areas to maintain the continuity of communication sessions.

[0003] In existing technologies, the vehicle-to-ground communication handover process generally adopts a signal strength-triggered reactive handover mechanism. The train continuously monitors the reference signal received power or signal-to-interference-plus-noise ratio of the currently accessed network. When the monitored value is lower than a preset handover threshold, the handover process is triggered. The train scans and measures the surrounding available networks, selects the network with the best signal quality from the measurement results as the handover target, and then completes the access point change through a standard handover signaling process. The handover decision depends entirely on the instantaneous signal measurement results.

[0004] In existing technologies, decision-making methods based on instantaneous signal strength are difficult to accurately determine the location of network coverage boundaries. The handover point often lags behind the actual coverage overlap area, which can easily lead to situations where the handover is not completed before the network leaves the original coverage area, resulting in communication interruption. In addition, the selection of the target network is based solely on signal quality parameters without assessing the current load status and available resource reserves of the target network. During peak passenger flow periods or busy sections of the line, the network with better signal strength may already be operating under high load. After handover, it may not be able to obtain sufficient transmission bandwidth, causing data congestion and increased transmission latency, which will affect the normal carrying of vehicle-to-ground communication services. Summary of the Invention

[0005] This invention provides a method and system for switching rail transit vehicle-to-ground communication based on heterogeneous network collaboration, which can at least solve some of the problems existing in the prior art.

[0006] A first aspect of this invention provides a method for handover of rail transit vehicle-to-ground communication based on heterogeneous network collaboration, comprising:

[0007] Collect real-time train operating parameters and signal quality parameters of the currently accessed network, perform time-series analysis on the signal quality parameters to extract signal attenuation trend features, perform mapping calculations on the signal attenuation trend features and real-time train operating parameters to obtain a predicted position trajectory, and determine the available switching time window based on the predicted position trajectory;

[0008] Within the available switching time window, parallel channel probing is performed on candidate heterogeneous networks within the coverage area of ​​the predicted location trajectory to obtain real-time channel quality parameters and network coverage information. Based on the predicted location trajectory and the network coverage information, the network coverage duration of each candidate heterogeneous network is calculated, and the network switching priority sequence is obtained by combining the real-time channel quality parameters. The current load parameters corresponding to each candidate heterogeneous network in the network switching priority sequence are queried to evaluate resource sufficiency, and the target access network is determined by combining the network switching priority sequence.

[0009] A two-layer handover link is established between the current access network and the target access network. A handover request message is sent to the target access network through the control signaling channel in the two-layer handover link. A handover preparation completion identifier is obtained by receiving a resource reservation confirmation message from the current access network. A transmission permission handover instruction is generated and executed based on the handover preparation completion identifier.

[0010] In one alternative implementation,

[0011] The process involves collecting real-time train operating parameters and signal quality parameters of the currently accessed network, performing time-series analysis on the signal quality parameters to extract signal attenuation trend features, mapping the signal attenuation trend features to the real-time train operating parameters to obtain a predicted position trajectory, and determining the available switching time window based on the predicted position trajectory, including:

[0012] The train's real-time operating parameters are obtained by collecting the train's current position coordinates and operating information through an on-board sensor array, and the signal quality parameters are obtained by collecting the received signal strength indication value and signal interference noise ratio of the currently accessed network through a wireless communication interface.

[0013] The signal quality parameters are sampled by a sliding window according to the time series to obtain multiple signal quality subsequences. The difference operation is then performed to obtain the signal change rate sequence. The signal change rate sequence is linearly fitted to obtain the attenuation trend function. Based on the attenuation trend function, the predicted time when the signal strength reaches the switching trigger threshold is calculated to obtain the signal attenuation trend feature.

[0014] Based on the operational information, the train's displacement vector in the future time period is calculated and superimposed with the current position coordinates to obtain the future position coordinate sequence. The signal attenuation trend feature is spatially mapped and associated with the future position coordinate sequence to obtain the predicted position trajectory.

[0015] The signal strength is calculated from the handover trigger threshold to the handover necessary threshold based on the attenuation trend function, and the time interval is obtained by spatiotemporally corresponding with the predicted position trajectory to obtain the handover start position point and the handover end position point. The available handover time window is determined based on the time corresponding to the handover start position point and the time corresponding to the handover end position point.

[0016] In one alternative implementation,

[0017] Within the available switching time window, parallel channel probing is performed on candidate heterogeneous networks within the coverage area of ​​the predicted location trajectory to obtain real-time channel quality parameters and network coverage information. Based on the predicted location trajectory and the network coverage information, the network coverage duration of each candidate heterogeneous network is calculated, and the network switching priority sequence is obtained by combining the real-time channel quality parameters. This sequence includes:

[0018] Based on the predicted location trajectory, candidate heterogeneous networks are queried and a channel probe request message is sent. The channel state information returned by the candidate heterogeneous networks is received and multipath components are separated. The delay spread is calculated based on the multipath components to obtain the channel coherence time. The received signal strength indication value is extracted based on the channel state information and combined with the channel coherence time to obtain the real-time channel quality parameters. The network coverage boundary coordinates are extracted based on the channel state information to obtain the network coverage range information.

[0019] The sequence of trajectory coordinate points is extracted from the predicted location trajectory and the spatial distance between each trajectory coordinate point and the coordinates of the network coverage boundary is calculated. The cutting angle is calculated based on the spatial distance and the pre-acquired operation information, and the network coverage boundary is dynamically expanded according to the cutting angle to obtain the corrected coverage boundary. The network coverage duration is obtained by calculating the time difference between the coverage start time and the coverage end time based on the corrected coverage boundary.

[0020] The train running speed is obtained to calculate the Doppler frequency shift, and the received signal strength indication value is frequency shift compensated to obtain the frequency shift compensated signal strength. The channel coherence time and the network coverage duration are combined with the frequency shift compensated signal strength to calculate the switching cost of each candidate heterogeneous network and arrange them in ascending order to obtain the network switching priority sequence.

[0021] In one alternative implementation,

[0022] The Doppler frequency shift is calculated by obtaining the train speed, and frequency shift compensation is applied to the received signal strength indication value to obtain the frequency shift compensated signal strength. The switching cost of each candidate heterogeneous network is calculated by combining the channel coherence time and the network coverage duration with the frequency shift compensated signal strength, and then sorted in ascending order to obtain the network switching priority sequence, including:

[0023] Train speed and direction of travel are extracted from the real-time train operating parameters. Doppler frequency shift is calculated for each candidate heterogeneous network based on the train speed and the carrier frequency of each candidate heterogeneous network. Subcarrier phase information is extracted from the channel state information and phase rotation compensation is performed on the subcarrier phase information based on the Doppler frequency shift to obtain a phase compensation matrix. Frequency domain coherent compensation is performed on the received signal strength indication value of each candidate heterogeneous network based on the phase compensation matrix to obtain the frequency shift compensated signal strength.

[0024] The channel stability coverage ratio is calculated based on the channel coherence time and the network coverage duration. Angle spread information is extracted based on the channel state information and the beam alignment offset angle is calculated by combining the angle spread information and the train running direction. The channel stability coverage ratio is dynamically corrected based on the beam alignment offset angle to obtain a corrected channel stability coverage ratio. The handover cost of each candidate heterogeneous network is calculated based on the corrected channel stability coverage ratio and the frequency shift compensation signal strength, and the network handover priority sequence is obtained.

[0025] In one alternative implementation,

[0026] Querying the current load parameters of each candidate heterogeneous network in the network handover priority sequence to assess resource sufficiency and determining the target access network in conjunction with the network handover priority sequence includes:

[0027] Extract candidate heterogeneous network identifiers from the network switching priority sequence and send load query request messages to each candidate heterogeneous network. Receive current load parameters returned by each candidate heterogeneous network and extract physical resource block occupancy and scheduling authorization ratio based on the current load parameters.

[0028] Based on the real-time train operation parameters, service transmission requirements are extracted. Channel quality indicator values ​​and multi-user interference information are extracted from the pre-acquired channel state information. Interference suppression correction is performed on the channel quality indicator values ​​based on the multi-user interference information to obtain effective channel quality indicator values. The achievable throughput is calculated by combining the physical resource block occupancy rate and the scheduling authorization ratio. The achievable throughput and the service transmission requirements are matched and verified to obtain throughput satisfaction.

[0029] The number of user connections for each candidate heterogeneous network is obtained to calculate the network congestion probability. Based on the network congestion probability, the throughput satisfaction is dynamically corrected to obtain a corrected resource sufficiency evaluation value. Based on the corrected resource sufficiency evaluation value and the handover cost corresponding to each candidate heterogeneous network in the network handover priority sequence, a handover score is calculated, and the candidate heterogeneous network with the highest handover score is selected as the target access network.

[0030] In one alternative implementation,

[0031] Establishing a two-layer handover link between the current access network and the target access network, and sending a handover request message to the target access network through the control signaling channel in the two-layer handover link, including:

[0032] The system extracts the bearer status information from the real-time train operation parameters and constructs a service flow descriptor. It obtains the current session context information from the current access network and encapsulates it with the service flow descriptor to obtain a handover context data packet. Based on the service priority and latency sensitivity in the handover context data packet, it pre-allocates resource blocks to the access channel of the target access network to obtain a temporary resource allocation table and constructs a two-layer handover link based on the temporary resource allocation table.

[0033] Based on the handover context data packet and the temporary resource allocation table, a handover request message is constructed and sent to the target access network through the two-layer handover link. The resource confirmation response returned by the target access network is received and the corresponding allocated resource block identifier is extracted. Based on the allocated resource block identifier, the temporary resource allocation table is updated to obtain the formal resource allocation table and a handover preparation completion identifier is generated.

[0034] The cached data packets in the current access network are pre-forwarded to the target access network through the dual-layer switching link, and a data packet sequence number mapping relationship is established. Integrity verification is performed based on the data packet sequence number mapping relationship, and a transmission permission switching instruction is generated in conjunction with the switching preparation completion identifier. The transmission permission switching instruction is then sent synchronously to the current access network and the target access network to execute the transmission permission switching instruction.

[0035] In one alternative implementation,

[0036] Obtain a handover preparation completion flag from the resource reservation confirmation message received from the current access network, generate a transmission permission handover instruction based on the handover preparation completion flag, and execute it, including:

[0037] The system receives a resource reservation confirmation message from the target access network and extracts handover timing parameters. It then extracts a frame synchronization offset from the handover timing parameters and pre-compensates the transmission timing of the current access network based on the frame synchronization offset to obtain timing alignment parameters. Finally, it generates a handover preparation completion flag based on the timing alignment parameters.

[0038] Extract the timing alignment parameter from the switching preparation completion identifier and construct the transmission permission switching instruction by combining it with the service flow status in the real-time train operation parameters. Based on the timing alignment parameter, perform timestamp calibration on the switching execution time in the transmission permission switching instruction to obtain the synchronous switching time.

[0039] The transmission permission switching instruction is sent synchronously to the current access network and the target access network. The arrival of the synchronization switching time is monitored and the current access network is triggered to stop sending data, while the target access network is triggered to start receiving data.

[0040] A second aspect of this invention provides a rail transit vehicle-to-ground communication handover system based on heterogeneous network collaboration, comprising:

[0041] The trajectory prediction unit is used to collect real-time train operating parameters and signal quality parameters of the currently accessed network, perform time-series analysis on the signal quality parameters to extract signal attenuation trend features, perform mapping calculations on the signal attenuation trend features and real-time train operating parameters to obtain the predicted position trajectory, and determine the available switching time window based on the predicted position trajectory.

[0042] The network selection unit is configured to perform parallel channel probing on candidate heterogeneous networks within the coverage area of ​​the predicted location trajectory within the available switching time window to obtain real-time channel quality parameters and network coverage information; calculate the network coverage duration of each candidate heterogeneous network based on the predicted location trajectory and the network coverage information; solve for the network switching priority sequence by combining the real-time channel quality parameters; query the current load parameters corresponding to each candidate heterogeneous network in the network switching priority sequence to evaluate resource adequacy; and determine the target access network by combining the network switching priority sequence.

[0043] The handover execution unit is used to establish a two-layer handover link between the current access network and the target access network, send a handover request message to the target access network through the control signaling channel in the two-layer handover link, receive a resource reservation confirmation message from the current access network to obtain a handover preparation completion identifier, generate a transmission permission handover instruction based on the handover preparation completion identifier, and execute it.

[0044] A third aspect of the present invention provides an electronic device, comprising:

[0045] A processor and a memory for storing processor-executable instructions, wherein the processor is configured to invoke instructions stored in the memory to perform the aforementioned method.

[0046] A fourth aspect of the present invention provides a computer-readable storage medium having stored thereon computer program instructions that, when executed by a processor, implement the aforementioned method.

[0047] In this invention, time-series analysis is performed using real-time train operating parameters and signal quality parameters to extract signal attenuation trend characteristics. Combined with predicted position trajectories, available handover time windows are determined, enabling forward-looking handover decisions. Handover timing is based on attenuation trends and train position predictions, rather than passively waiting for signal deterioration. This effectively avoids untimely handovers caused by rapid signal fading, significantly reduces the probability of handover interruptions, and ensures continuous and stable train-to-ground communication. The coverage duration of each network based on predicted position trajectories is calculated, and the handover priority sequence is solved using channel quality parameters. Finally, the current load and resource sufficiency of each candidate network are evaluated to determine the target access network. This approach allows for target network selection that balances signal strength, coverage continuity, and load balancing. It effectively avoids networks with strong signals that are about to leave their coverage area or are already heavily loaded, reducing the probability of ping-pong handovers and improving handover success rates and system resource utilization efficiency. Handover requests are sent to the target network in advance via the control signaling channel, and transmission permission handover instructions are generated only after resource reservation confirmation. This ensures that service data is only switched after the target network is ready, significantly reducing downtime during the handover execution phase. Signaling interaction and data forwarding are performed in parallel, enhancing the fault tolerance of the handover process. This effectively improves handover reliability and communication service quality in high-speed mobile scenarios, fully supporting the high-safety communication requirements of rail transit. Attached Figure Description

[0048] Figure 1 This is a flowchart illustrating the method for switching rail transit vehicle-to-ground communication based on heterogeneous network collaboration, according to an embodiment of the present invention.

[0049] Figure 2 This is a diagram showing the signal strength attenuation trend and switching window of the rail transit vehicle-to-ground communication switching method based on heterogeneous network collaboration in an embodiment of the present invention.

[0050] Figure 3 This is a handover signaling timing diagram of a rail transit vehicle-to-ground communication handover method based on heterogeneous network collaboration, according to an embodiment of the present invention.

[0051] Figure 4 This is a flowchart illustrating the two-layer handover link signaling interaction and data synchronization process of the rail transit vehicle-to-ground communication handover method based on heterogeneous network collaboration, as described in an embodiment of the present invention. Detailed Implementation

[0052] To make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, the technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0053] The technical solution of the present invention will be described in detail below with reference to specific embodiments. These specific embodiments can be combined with each other, and the same or similar concepts or processes may not be described again in some embodiments.

[0054] Figure 1 This is a flowchart illustrating the method for handover of rail transit vehicle-to-ground communication based on heterogeneous network collaboration, as described in an embodiment of the present invention. Figure 1 As shown, the method includes:

[0055] Collect real-time train operating parameters and signal quality parameters of the currently accessed network, perform time-series analysis on the signal quality parameters to extract signal attenuation trend features, perform mapping calculations on the signal attenuation trend features and real-time train operating parameters to obtain a predicted position trajectory, and determine the available switching time window based on the predicted position trajectory;

[0056] Within the available switching time window, parallel channel probing is performed on candidate heterogeneous networks within the coverage area of ​​the predicted location trajectory to obtain real-time channel quality parameters and network coverage information. Based on the predicted location trajectory and the network coverage information, the network coverage duration of each candidate heterogeneous network is calculated, and the network switching priority sequence is obtained by combining the real-time channel quality parameters. The current load parameters corresponding to each candidate heterogeneous network in the network switching priority sequence are queried to evaluate resource sufficiency, and the target access network is determined by combining the network switching priority sequence.

[0057] A two-layer handover link is established between the current access network and the target access network. A handover request message is sent to the target access network through the control signaling channel in the two-layer handover link. A handover preparation completion identifier is obtained by receiving a resource reservation confirmation message from the current access network. A transmission permission handover instruction is generated and executed based on the handover preparation completion identifier.

[0058] In one alternative implementation,

[0059] The process involves collecting real-time train operating parameters and signal quality parameters of the currently accessed network, performing time-series analysis on the signal quality parameters to extract signal attenuation trend features, mapping the signal attenuation trend features to the real-time train operating parameters to obtain a predicted position trajectory, and determining the available switching time window based on the predicted position trajectory, including:

[0060] The train's real-time operating parameters are obtained by collecting the train's current position coordinates and operating information through an on-board sensor array, and the signal quality parameters are obtained by collecting the received signal strength indication value and signal interference noise ratio of the currently accessed network through a wireless communication interface.

[0061] The signal quality parameters are sampled by a sliding window according to the time series to obtain multiple signal quality subsequences. The difference operation is then performed to obtain the signal change rate sequence. The signal change rate sequence is linearly fitted to obtain the attenuation trend function. Based on the attenuation trend function, the predicted time when the signal strength reaches the switching trigger threshold is calculated to obtain the signal attenuation trend feature.

[0062] Based on the operational information, the train's displacement vector in the future time period is calculated and superimposed with the current position coordinates to obtain the future position coordinate sequence. The signal attenuation trend feature is spatially mapped and associated with the future position coordinate sequence to obtain the predicted position trajectory.

[0063] The signal strength is calculated from the handover trigger threshold to the handover necessary threshold based on the attenuation trend function, and the time interval is obtained by spatiotemporally corresponding with the predicted position trajectory to obtain the handover start position point and the handover end position point. The available handover time window is determined based on the time corresponding to the handover start position point and the time corresponding to the handover end position point.

[0064] The onboard sensor array is deployed at multiple locations on the train's front, body, and rear. By fusing output data from GPS / BeiDou positioning modules, speed sensors, and accelerometers, it continuously collects the train's current position coordinates and operational information. The operational information includes components such as the train's current speed, acceleration, and heading angle, collectively constituting the train's real-time operational parameters. Through the onboard wireless communication interface, it periodically initiates signal quality measurement requests to the currently accessed network, collecting Received Signal Strength Indication (RSSI) and Signal-to-Noise Ratio (SINR), which together constitute the signal quality parameters. The acquisition cycle can be dynamically adjusted according to the train's operating speed, with the acquisition interval appropriately shortened in high-speed sections to ensure the timeliness and accuracy of the signal quality parameters.

[0065] The acquired signal quality parameters are sampled using a sliding window based on a time series. Let the sliding window length be... The sampling step size is The signal quality parameters within each window of the continuous time series are truncated to obtain multiple signal quality subsequences. A first-order difference operation is performed on each signal quality subsequence to calculate the signal quality change between adjacent sampling times, thus obtaining a signal rate of change sequence. The signal rate of change sequence reflects the dynamic trend of signal intensity over time and can effectively filter out interference from short-term random fluctuations. The signal rate of change sequence is then linearly fitted using the least squares method to obtain an attenuation trend function. Let the fitted attenuation trend function be... ,in For time variables, The attenuation slope, The intercept is... This indicates that the signal is in a state of continuous attenuation. Based on this attenuation trend function, the switching trigger threshold is substituted. Solve for the condition that satisfies At the given moment, the predicted moment when the signal strength is expected to reach the switching trigger threshold can be obtained. The predicted moment serves as the core output of the signal attenuation trend characteristics and is used for subsequent spatiotemporal mapping calculations.

[0066] Based on the velocity and acceleration vectors in the operational information, the motion state of the train in the future time period is estimated. Let the current time be... The train's current position coordinates are The current velocity vector is The current acceleration vector is Then in the future Each sampling time At that point, the train's displacement vector It can be calculated using kinematic equations, i.e. ,in The time step is the displacement vector. With current position coordinates By performing vector superposition, we can obtain the future nth Predicted position coordinates at each time point Calculate sequentially for multiple future moments to form a sequence of future location coordinates. This sequence is the discretized representation of the train's spatial trajectory within the predicted time range.

[0067] The core of spatially mapping and associating signal attenuation trend features with future position coordinate sequences lies in establishing a correspondence between the time axis and spatial location. The signal attenuation trend features contain predictive information about signal strength changes over time, and each position point in the future position coordinate sequence corresponds to a specific predicted time. Therefore, the time axis of the signal attenuation trend function can be aligned with the time index of the position coordinate sequence, thereby assigning a corresponding predicted signal strength value to each predicted position point. After this spatial mapping and association process, the predicted position trajectory is obtained, which is a comprehensive trajectory description of each spatial position point carrying a timestamp and predicted signal quality information. The predicted position trajectory not only describes the spatial movement path of the train but also incorporates the time-varying characteristics of signal quality, providing complete spatiotemporal information support for determining the subsequent switching time window.

[0068] When determining the available switching time window, the switching trigger threshold should also be considered. and switching must threshold Both define the time constraint that the handover operation must be completed within a certain period. Handover trigger threshold This indicates that the signal quality has deteriorated to a critical value that necessitates initiating a handover process; the handover must meet this threshold. This indicates that the signal quality has degraded to the limit where communication will be interrupted if a handover is not completed. Based on the decay trend function Solve separately and At the corresponding moment, the switching trigger moment is obtained. and the switching must be completed at the time. The difference between the two is the available time interval for the switching operation. ,in This represents the available switching time length.

[0069] Time interval Perform spatiotemporal correspondence with the predicted location trajectory to find the... The corresponding spatial location point is used as the starting point for switching, and the point is found to be the same as the starting point for switching. The corresponding spatial location point serves as the handover termination point. The handover start and end points together define the geographical scope of the handover operation on the spatial trajectory, while the time corresponding to the handover start point... The time corresponding to the switching termination point The available switching time window is then determined on the timeline. The physical meaning of this time window is clear: within this time period, the train is in a range where signal quality allows for switching but has not yet reached the critical state of communication interruption, making it the optimal time to perform heterogeneous network switching. The available switching time window obtained through the above method has constraints in both time and space dimensions, providing precise spatiotemporal boundaries for subsequent channel detection and target network selection of candidate heterogeneous networks, effectively avoiding communication interruptions or unnecessary switching overhead caused by improper selection of switching timing.

[0070] In real-world rail transit scenarios, trains operate at high speeds, and the signal attenuation rate varies non-linearly due to factors such as tunnels, bridges, and building obstructions. Therefore, when performing linear fitting, outliers in the signal rate of change sequence can be removed, and robust linear regression methods can be used to improve the fitting accuracy of the attenuation trend function. Simultaneously, for sections with significant acceleration variations, a piecewise kinematic model can be introduced to calculate displacement vectors for different acceleration intervals, which are then concatenated to form a complete future position coordinate sequence. This improves the spatial accuracy of the predicted position trajectory and further ensures the reliability of the calculation results within the available switching time window.

[0071] Figure 2 This invention presents a signal strength attenuation trend and switching window diagram for a heterogeneous network-based vehicle-to-ground communication handover method in rail transit, illustrating the correspondence between the predicted position trajectory of a train within the next 28 seconds and the current access network signal strength in a high-speed operating scenario. The predicted position trajectory is generated based on the time-series analysis results of the train's real-time operating parameters (operating speed 286 km / h, acceleration 0.12 m / s²) and signal quality parameters, covering a railway mileage range of approximately 2.3 kilometers.

[0072] exist Figure 2 The signal strength attenuation trend with the predicted location can be observed: at the predicted location 0km (corresponding to the initial time 0s), the signal strength is approximately -76dBm, which is in a good reception state; as the train moves along the predicted trajectory, the signal strength shows an attenuation trend: at 4s, the signal strength is approximately -83dBm; at 8s, it attenuates to the handover trigger threshold of -85dBm; at 12s, it is approximately -88dBm; at 16s, it is approximately -92dBm; at 20s, it is approximately -95dBm; at 24s, it is approximately -96dBm; and by 28s, the signal strength drops to approximately -98dBm, reaching the necessary handover threshold. The dashed line in the figure represents the predicted location trajectory, and the solid line represents the signal strength attenuation trend. The two curves are spatially mapped and correlated through the time axis, forming a complete description of the predicted location trajectory.

[0073] At time 8s, approximately 1.2km from the predicted location, the signal strength attenuates to the handover trigger threshold of -85dBm; this location is marked as the handover start point. At time 28s, approximately 7.5km from the predicted location, the signal strength attenuates to the handover mandatory threshold of -98dBm; this location is marked as the handover end point. The time interval between the handover start and end points is 20s, corresponding to a spatial distance of approximately 1.6km; this interval is determined as the available handover time window.

[0074] The figure also shows that in the first segment of the predicted location trajectory (0-8s), the signal strength decreases relatively slowly, with an attenuation slope of approximately -0.11dB / s; while in the latter segment (8-28s), due to factors such as the curve of the line and obstruction by surrounding buildings, the signal attenuation rate accelerates, with the attenuation slope increasing to approximately -0.15dB / s. This nonlinear attenuation characteristic is effectively extracted using a sliding window sampling and piecewise linear fitting method and mapped to the predicted location trajectory, making the determination of the available switching time window more accurate.

[0075] In one alternative implementation,

[0076] Within the available switching time window, parallel channel probing is performed on candidate heterogeneous networks within the coverage area of ​​the predicted location trajectory to obtain real-time channel quality parameters and network coverage information. Based on the predicted location trajectory and the network coverage information, the network coverage duration of each candidate heterogeneous network is calculated, and the network switching priority sequence is obtained by combining the real-time channel quality parameters. This sequence includes:

[0077] Based on the predicted location trajectory, candidate heterogeneous networks are queried and a channel probe request message is sent. The channel state information returned by the candidate heterogeneous networks is received and multipath components are separated. The delay spread is calculated based on the multipath components to obtain the channel coherence time. The received signal strength indication value is extracted based on the channel state information and combined with the channel coherence time to obtain the real-time channel quality parameters. The network coverage boundary coordinates are extracted based on the channel state information to obtain the network coverage range information.

[0078] The sequence of trajectory coordinate points is extracted from the predicted location trajectory and the spatial distance between each trajectory coordinate point and the coordinates of the network coverage boundary is calculated. The cutting angle is calculated based on the spatial distance and the pre-acquired operation information, and the network coverage boundary is dynamically expanded according to the cutting angle to obtain the corrected coverage boundary. The network coverage duration is obtained by calculating the time difference between the coverage start time and the coverage end time based on the corrected coverage boundary.

[0079] The train running speed is obtained to calculate the Doppler frequency shift, and the received signal strength indication value is frequency shift compensated to obtain the frequency shift compensated signal strength. The channel coherence time and the network coverage duration are combined with the frequency shift compensated signal strength to calculate the switching cost of each candidate heterogeneous network and arrange them in ascending order to obtain the network switching priority sequence.

[0080] After the available handover time window is determined, candidate heterogeneous networks within the coverage area of ​​the current train's travel path are queried based on the predicted location trajectory, and channel probe request messages are simultaneously sent to these candidate networks. Parallel channel probe refers to simultaneously probing multiple candidate networks within the same time window, rather than probing them sequentially one by one. This effectively reduces the total probe time and avoids missed handover opportunities due to probe delays. The channel probe request message carries the train's current position coordinates and speed information. After receiving the request, the candidate heterogeneous networks return Channel State Information (CSI). The CSI contains raw measurement data such as amplitude, phase, and delay of each path in multipath propagation, which is the basis for subsequent channel quality assessment.

[0081] After receiving the channel state information (CSI), multipath component separation is performed. Multipath component separation refers to identifying and extracting parameters of each independent propagation path from the frequency or time domain response of the CSI, including the arrival delay of each path. (in (path number) and the corresponding complex gain Calculate the delay spread based on the delay of each path. Delay spread reflects the dispersion of arrival times of different paths in a multipath channel; a larger value indicates stronger channel frequency selectivity and narrower coherence bandwidth. Channel coherence time. The time delay is inversely proportional to the Doppler spread, which is determined by both the train speed and the signal carrier frequency. Given the known time delay spread, the time-varying characteristics of the channel can be further estimated. Channel coherence time. It characterizes the length of time during which channel characteristics remain stable. A larger value indicates that the channel changes more slowly, making it suitable for stable communication over a longer period of time.

[0082] Extract the received signal strength indication value from the channel state information. This value reflects the received power level at the current frequency, and the unit is usually dBm. Channel coherence time The data are combined to form a real-time channel quality parameter vector, which is used for subsequent handover cost calculation. Simultaneously, network coverage boundary coordinates are extracted from the channel state information's auxiliary fields; these coordinates represent the set of vertices of the geographic boundary polygon representing the effective signal coverage area of ​​the candidate network's base station or access point, thus obtaining network coverage information. The accuracy of the coverage boundary coordinates directly affects the accuracy of subsequent coverage duration calculations; therefore, candidate networks are required to synchronously report their latest coverage boundary measurement data when returning CSI.

[0083] Extracting trajectory coordinate point sequences from predicted location trajectories (in (For the trajectory point index), calculate the coordinates of each trajectory point. Spatial distance to each side of the candidate network coverage boundary polygon Spatial distance is calculated using the shortest Euclidean distance method from a point to a line segment. When a trajectory point is located inside the coverage boundary, the distance is taken as a negative value to distinguish the internal and external positional relationship. Based on the sign change of the spatial distance sequence, the location of the trajectory point when the train enters the coverage area can be identified, and then the angle of intersection between the train trajectory and the coverage boundary can be calculated. Angle of entry Defined as the angle between the train's direction of motion vector and the tangent direction of the coverage boundary, with a value ranging from 0° to 90°.

[0084] When the angle of entry When the angle is small, the train moves in a direction nearly parallel to the coverage boundary, and the actual effective coverage area will be smaller than the nominal coverage boundary due to the degradation of signal quality at the edges; when the angle of entry is small... When the angle is large, the train crosses the coverage area at a large angle, and the effective coverage area remains relatively stable. Therefore, based on the angle of entry... The coverage boundary is dynamically expanded and corrected, with the correction amount being... and The relationship is: when When it is small, the correction amount Take a larger shrinkage value, that is, shrink the coverage boundary inward to reflect the actual effective coverage area; when When approaching 90°, the correction amount Approaching zero. Specifically, ,in The preset maximum boundary correction distance is determined by the network type and signal propagation model. The corrected coverage boundary, after correction, better reflects the effective access area of ​​the train under actual motion.

[0085] Based on the corrected coverage boundary, the time corresponding to the train's entry point into the corrected coverage area is re-determined on the predicted location trajectory, i.e., the coverage start time. And the time corresponding to the train leaving the correction coverage area, i.e., the end time of the coverage. Network coverage duration The difference between the two is, The unit is seconds. The larger the value, the longer the train stays within the coverage area of ​​the candidate network, the more time it can maintain stable communication after the handover, and the higher the handover value.

[0086] After obtaining real-time channel quality parameters and network coverage duration, the impact of the Doppler effect caused by the high-speed train movement on signal strength measurements is further considered. The current train speed is then obtained. Combined with the signal carrier frequency of the candidate network and the speed of light Calculate the Doppler frequency shift. The calculation formula is: ,in This is the angle between the train's direction of motion and the base station's direction. Doppler frequency shift causes a frequency shift in the received signal, affecting the received signal strength indication value measured based on the uncompensated signal. There is a systematic bias. Regarding... Frequency shift compensation is performed to convert the power loss caused by frequency offset into an equivalent signal strength correction. The frequency shift compensation signal strength is obtained. ,in According to Doppler frequency shift The ratio to the channel bandwidth can be obtained by looking up a table.

[0087] Synthetic channel coherence time Duration of network coverage and frequency shift compensation signal strength Calculate the switching cost of each candidate heterogeneous network. Switching value The calculation takes into account three dimensions: The lower the value, the worse the channel quality and the higher the handover cost; The smaller the value, the faster the channel changes, the more difficult it is to maintain a stable connection, and the higher the handover cost. The shorter the time, the shorter the train stays within the network coverage area, meaning a quicker handover is needed and the higher the cost. The final handover cost is calculated by the normalized weighted sum of the three dimensions. The weighting coefficients are pre-configured based on the service priorities of the rail transit scenario. The switching cost for all candidate heterogeneous networks... Sort in ascending order and switch the value. The smallest candidate network is placed at the beginning of the sequence, indicating the highest priority, thus obtaining the network handover priority sequence. This sequence provides an ordered set of candidates for subsequent resource adequacy assessment and target access network determination.

[0088] In one alternative implementation,

[0089] The Doppler frequency shift is calculated by obtaining the train speed, and frequency shift compensation is applied to the received signal strength indication value to obtain the frequency shift compensated signal strength. The switching cost of each candidate heterogeneous network is calculated by combining the channel coherence time and the network coverage duration with the frequency shift compensated signal strength, and then sorted in ascending order to obtain the network switching priority sequence, including:

[0090] Train speed and direction of travel are extracted from the real-time train operating parameters. Doppler frequency shift is calculated for each candidate heterogeneous network based on the train speed and the carrier frequency of each candidate heterogeneous network. Subcarrier phase information is extracted from the channel state information and phase rotation compensation is performed on the subcarrier phase information based on the Doppler frequency shift to obtain a phase compensation matrix. Frequency domain coherent compensation is performed on the received signal strength indication value of each candidate heterogeneous network based on the phase compensation matrix to obtain the frequency shift compensated signal strength.

[0091] The channel stability coverage ratio is calculated based on the channel coherence time and the network coverage duration. Angle spread information is extracted based on the channel state information and the beam alignment offset angle is calculated by combining the angle spread information and the train running direction. The channel stability coverage ratio is dynamically corrected based on the beam alignment offset angle to obtain a corrected channel stability coverage ratio. The handover cost of each candidate heterogeneous network is calculated based on the corrected channel stability coverage ratio and the frequency shift compensation signal strength, and the network handover priority sequence is obtained.

[0092] Extracting train speed from real-time train operation parameters and the angle of train running direction Combining the carrier frequencies of each candidate heterogeneous network Calculate the Doppler frequency shift for each candidate network. Since the train moves at high speed, there are different angles between the train's direction of motion and the directions of each base station. Therefore, the Doppler frequency shift of each candidate network The calculation method is as follows ,in The speed is the speed of light. Different candidate heterogeneous networks (such as LTE, 5G NR, WLAN, etc.) operate on different carrier frequencies, and the Doppler frequency shift generated at the same train speed is significantly different. Therefore, it is necessary to calculate the corresponding Doppler frequency shift for each candidate network independently to ensure the accuracy of subsequent compensation.

[0093] After obtaining the Doppler shift of each candidate network, the phase information of each subcarrier is extracted from the channel state information (CSI) to construct a subcarrier phase matrix. The subcarrier phase information reflects the actual phase deflection state of each subcarrier under the current channel conditions. Based on the calculated Doppler shift... Perform phase rotation compensation operation on the subcarrier phase information: for the first Each subcarrier, corresponding phase compensation amount The compensated phase is determined by the center frequency of the subcarrier and the Doppler shift. satisfy ,in For the first The estimated propagation delay values ​​for each subcarrier are then calculated. The compensation results for all subcarriers are then aggregated to form a phase compensation matrix. This matrix records the phase distribution of each subcarrier after Doppler effect elimination, providing basic data for subsequent frequency domain coherent compensation.

[0094] Based on phase compensation matrix Received signal strength indication value for each candidate heterogeneous network Frequency-domain coherent compensation is performed, using a phase compensation matrix to coherently superimpose the received signal in the frequency domain. This eliminates inter-carrier interference (ICI) and power dispersion effects caused by Doppler frequency shift, thereby restoring the true power distribution of the signal. The compensated signal strength is then obtained. Its value compared to the uncompensated value This more accurately reflects the actual reception quality of the link. For 5G candidate networks operating in the millimeter-wave band, due to the carrier frequency... High, Doppler frequency shift More significantly, frequency shift compensation has The correction range is also larger, and this step is crucial to the accuracy of switching decisions in high-speed scenarios.

[0095] After completing frequency shift compensation, based on channel coherence time and network coverage duration Calculate the channel stability coverage ratio Defined as This ratio reflects the percentage of time the channel remains coherent and stable throughout the entire journey of a train traversing the coverage area of ​​a candidate network. When the value is small, it indicates that the channel coherence time is much shorter than the coverage duration, the channel changes frequently within the coverage area, and the link stability after handover is poor; when A value close to or greater than 1 indicates that the channel maintains good coherence throughout the coverage period, ensuring communication quality after handover. Channel stability coverage ratio. The introduction of this feature enables handover decisions to consider not only the current signal strength but also the temporal stability of the link after the handover.

[0096] The angle spread information is further extracted from the channel state information to obtain the angle spread amount. This characterizes the dispersion of the arrival angle of the multipath signal at the receiving end. Combined with the train's direction of travel angle... Azimuth angle relative to each candidate network base station Calculate the beam alignment offset angle Its definition Beam alignment offset angle This reflects the degree of deviation between the train's direction of motion and the direction of the candidate network beam's main lobe. When When the beam angle is large, the train's direction of motion deviates significantly from the beam direction, and the angle widens. The negative impact on channel quality is more pronounced, necessitating a downward correction to the channel stability coverage ratio; when When the beam is small, the beam alignment is good, the effect of angle spread is relatively small, and the correction range is relatively limited.

[0097] Based on beam alignment offset angle and angular expansion Coverage ratio for channel stability Dynamic correction is performed to obtain the corrected channel stability coverage ratio. The correction method involves introducing an angle correction factor. Defined as ,in Extend the normalized reference value for the preset angle. The range of values ​​is limited to To avoid a negative correction result, the corrected channel stability coverage ratio is adjusted accordingly. It comprehensively reflects the combined impact of multipath angle spread and beam alignment deviation on channel stability. The dynamic correction mechanism enables the handover priority assessment to adaptively respond to the differences in beam characteristics of different candidate networks, which is especially suitable for narrow beam scenarios of 5G millimeter wave networks.

[0098] Based on the corrected channel stability coverage ratio With frequency shift compensation signal strength Calculate the switching cost of each candidate heterogeneous network. Switching value The expected link quality and stability costs after switching to a candidate network are comprehensively quantified, and the calculation formula is as follows: The larger the product of the frequency shift compensation signal strength and the coverage ratio of the corrected channel stability, the smaller the handover cost, and the higher the handover priority of the corresponding candidate network. The handover cost for all candidate heterogeneous networks... Arranged in ascending order, a network handover priority sequence is obtained. Candidate networks ranked higher not only possess strong real-time signal quality but also exhibit good channel temporal stability and beam alignment characteristics, making them priority target candidate networks for handover. Subsequent processes will combine this priority sequence with the current load parameters of each candidate network to further determine the final target access network, thereby achieving load balancing while ensuring communication quality.

[0099] In one alternative implementation,

[0100] Querying the current load parameters of each candidate heterogeneous network in the network handover priority sequence to assess resource sufficiency and determining the target access network in conjunction with the network handover priority sequence includes:

[0101] Extract candidate heterogeneous network identifiers from the network switching priority sequence and send load query request messages to each candidate heterogeneous network. Receive current load parameters returned by each candidate heterogeneous network and extract physical resource block occupancy and scheduling authorization ratio based on the current load parameters.

[0102] Based on the real-time train operation parameters, service transmission requirements are extracted. Channel quality indicator values ​​and multi-user interference information are extracted from the pre-acquired channel state information. Interference suppression correction is performed on the channel quality indicator values ​​based on the multi-user interference information to obtain effective channel quality indicator values. The achievable throughput is calculated by combining the physical resource block occupancy rate and the scheduling authorization ratio. The achievable throughput and the service transmission requirements are matched and verified to obtain throughput satisfaction.

[0103] The number of user connections for each candidate heterogeneous network is obtained to calculate the network congestion probability. Based on the network congestion probability, the throughput satisfaction is dynamically corrected to obtain a corrected resource sufficiency evaluation value. Based on the corrected resource sufficiency evaluation value and the handover cost corresponding to each candidate heterogeneous network in the network handover priority sequence, a handover score is calculated, and the candidate heterogeneous network with the highest handover score is selected as the target access network.

[0104] Identifiers for each candidate heterogeneous network are extracted from the network switching priority sequence, and load query request messages are sent to the network management interface of each candidate heterogeneous network. The load query request message carries a query timestamp and a request identifier. Upon receiving the request, each candidate heterogeneous network packages its current resource usage status into a load parameter response message and returns it. After receiving the current load parameters returned by each candidate heterogeneous network, two key indicators are extracted: physical resource block occupancy rate and scheduling authorization ratio. The physical resource block occupancy rate reflects the actual utilization of current frequency domain resources; a higher occupancy rate indicates a tighter supply of available resources. The scheduling authorization ratio reflects the proportion of user requests successfully scheduled by the network per unit time; a lower ratio indicates that the network scheduling capacity is approaching saturation. These two indicators together form the basis for subsequent resource adequacy assessment.

[0105] Extracting service transmission requirements based on real-time train operating parameters requires comprehensive consideration of the train's current speed, the track section it is on, and the type of onboard services. At high speeds, the bandwidth requirements for real-time control commands and video monitoring data are significantly higher than at low speeds or when the train is stationary. Therefore, the service transmission requirements are not fixed values ​​but rather dynamic, quantified indicators that change with operating parameters, measured in Mbps. Channel quality indicators are extracted from pre-acquired channel state information. The multi-user interference information includes both co-channel interference intensity and neighboring cell interference distribution. Based on this information, the channel quality indicator (CMI) is corrected for interference suppression. During this correction, the co-channel interference power is converted into an equivalent signal-to-noise ratio (SNR) loss, and the interference effect is subtracted from the original CMI to obtain the effective CMI value. Their computational relationship satisfies: ,in The channel quality loss caused by multi-user interference is determined jointly by the intensity of co-channel interference and the distribution of interference in neighboring cells. It more realistically reflects the channel service capabilities that candidate heterogeneous networks can actually provide under the current interference environment.

[0106] Combined with effective channel quality indicator value Physical resource block occupancy rate and scheduling authorization ratio Calculate the achievable throughput of candidate heterogeneous networks for train services. The achievable throughput is calculated based on the Shannon capacity formula, while also incorporating physical resource block occupancy to constrain available frequency domain resources. Furthermore, the availability of time-domain scheduling is weighted by the scheduling grant ratio, resulting in a comprehensive estimate of the actual achievable data transmission rate. This achievable throughput... With business transmission requirements Perform matching verification and define throughput satisfaction. for ,when When the value equals 1, it indicates that the candidate network can fully meet the service transmission requirements; when... A value less than 1 indicates a throughput gap, meaning the candidate network's resource supply capacity is insufficient to cover current business demands. Throughput satisfaction, as a core intermediate measure in resource adequacy assessment, directly impacts the calculation of subsequent revised assessment values.

[0107] Get the current number of user connections for each candidate heterogeneous network. Combined with the maximum number of connections the network can carry Calculate the probability of network congestion The network congestion probability is modeled using a connection saturation-based approach. When the number of user connections approaches the maximum capacity limit, the congestion probability approaches 1, indicating an extremely high risk of new access requests being rejected or a severe deterioration in service quality. Specifically, satisfy: ,in This is the congestion sensitivity coefficient, used to adjust the non-linearity of congestion probability as the number of connections increases. A value greater than 1 indicates greater sensitivity to high load conditions. (Based on network congestion probability) Satisfaction with throughput Dynamic adjustments are made to obtain the adjusted resource sufficiency assessment value. : This correction mechanism embodies the following physical meaning: even if a candidate network performs well in terms of channel quality and frequency domain resources, if its current number of user connections is already saturated, the actual effective service capacity that can be provided to newly connected trains will be greatly reduced. The corrected resource sufficiency assessment value can accurately reflect this constraint.

[0108] Based on the revised resource adequacy assessment value The handover cost corresponding to each candidate heterogeneous network in the network handover priority sequence Calculate the switching score for each candidate heterogeneous network. The handover score comprehensively considers two dimensions: resource sufficiency and handover cost. Higher resource sufficiency and lower handover cost result in a higher handover score, indicating that the candidate network possesses both sufficient resource supply capacity and can complete network access with lower handover overhead. The handover score is calculated using a weighted linear combination method: ,in For the weighting coefficient of resource sufficiency, For the switching cost weighting coefficient, The switching cost is a dimensionless quantity after normalization, with the normalization benchmark being the maximum value of the switching costs across all candidate networks. Weight coefficients. and satisfy The resource availability and switching efficiency can be configured according to the actual business scenario. In scenarios where real-time control is prioritized, the efficiency can be appropriately increased. The proportion should be increased to ensure that the target access network has sufficient resource margin; in dense heterogeneous network scenarios with frequent switching, the proportion can be appropriately increased. The proportion of [something] is reduced to decrease the signaling overhead and latency loss caused by frequent switching.

[0109] Iterate through the switching scores of all candidate heterogeneous networks and select... The highest-ranking candidate heterogeneous network is selected as the target access network. If multiple candidate networks have the same handover score, the network with the highest handover priority is selected first, ensuring the uniqueness and certainty of the handover decision. After the target access network is determined, its identifier and corresponding resource reservation parameters are passed to the subsequent handover link establishment process, providing accurate target network information for the establishment of the two-layer handover link and the sending of handover request messages. The entire resource adequacy assessment and target network determination process is completed within the available handover time window, ensuring the timeliness and accuracy of the handover decision and avoiding missed handover opportunities due to assessment delays.

[0110] In one alternative implementation,

[0111] Establishing a two-layer handover link between the current access network and the target access network, and sending a handover request message to the target access network through the control signaling channel in the two-layer handover link, including:

[0112] The system extracts the bearer status information from the real-time train operation parameters and constructs a service flow descriptor. It obtains the current session context information from the current access network and encapsulates it with the service flow descriptor to obtain a handover context data packet. Based on the service priority and latency sensitivity in the handover context data packet, it pre-allocates resource blocks to the access channel of the target access network to obtain a temporary resource allocation table and constructs a two-layer handover link based on the temporary resource allocation table.

[0113] Based on the handover context data packet and the temporary resource allocation table, a handover request message is constructed and sent to the target access network through the two-layer handover link. The resource confirmation response returned by the target access network is received and the corresponding allocated resource block identifier is extracted. Based on the allocated resource block identifier, the temporary resource allocation table is updated to obtain the formal resource allocation table and a handover preparation completion identifier is generated.

[0114] The cached data packets in the current access network are pre-forwarded to the target access network through the dual-layer switching link, and a data packet sequence number mapping relationship is established. Integrity verification is performed based on the data packet sequence number mapping relationship, and a transmission permission switching instruction is generated in conjunction with the switching preparation completion identifier. The transmission permission switching instruction is then sent synchronously to the current access network and the target access network to execute the transmission permission switching instruction.

[0115] When obtaining the current session context information from the current access network, it is necessary to read the currently established radio bearer configuration parameters, IP address binding information, security key context, and sequence number status. When encapsulating the session context information with the service flow descriptor, a layered encapsulation structure is adopted: the outer encapsulation header contains a version identifier, encapsulation length, and verification fields; the inner layer sequentially arranges the session context information block and the service flow descriptor block, with the two blocks delimited by a separator field. After encapsulation, a handover context data packet is obtained. This handover context data packet serves as the basic data carrier for subsequent resource pre-allocation and handover request construction, maintaining its content unchanged throughout the handover process to ensure that the target access network can obtain complete and consistent session state information.

[0116] Based on the service priority and latency sensitivity in the handover context data packets, resource blocks are pre-allocated to the access channels of the target access network. Physical resource blocks are allocated to each service flow in descending order of service priority. For services with high latency sensitivity (such as train control signaling), resource blocks with better channel conditions and less interference are allocated first, and a certain proportion of redundant resources are reserved to cope with channel fluctuations. For services with low latency sensitivity (such as passenger broadband access), the remaining resource blocks are flexibly allocated while meeting the minimum bandwidth requirements. After pre-allocation, the resource block number, starting timeslot position, number of continuous timeslots, and modulation and coding scheme corresponding to each service flow are recorded in a temporary resource allocation table. Each row of the temporary resource allocation table corresponds to one service flow, and the columns are the service flow identifier, resource block starting number, number of resource blocks, modulation and coding scheme, and pre-allocation validity period, respectively. Based on a temporary resource allocation table, a two-layer handover link is established between the current access network and the target access network. The first layer is a control signaling channel, which is dedicated to transmitting control messages such as handover requests, resource confirmations, and command synchronization. The second layer is a data pre-forwarding channel, which is dedicated to pre-forwarding cached data packets from the current access network to the target access network before the handover is executed. The two channels are logically independent and do not interfere with each other.

[0117] A handover request message is constructed based on the handover context data packet and the temporary resource allocation table. The message body of the handover request message consists of three parts: the message header contains a message type identifier, a source network identifier, and a target network identifier; the message payload contains the complete handover context data packet content and a serialized representation of the temporary resource allocation table; and the message tail contains a message integrity check code. After the handover request message is sent to the target access network through the control signaling channel of the two-layer handover link, the target access network verifies the message integrity and performs an actual resource confirmation operation according to the pre-allocation scheme in the temporary resource allocation table. If the target network has sufficient resources, it returns a resource confirmation response, which includes a list of actually allocated resource block identifiers; if some resource blocks cannot be allocated due to conflicts, the response message provides suggestions for alternative resource blocks. Upon receiving the resource confirmation response, the allocated resource block identifiers in the response message are extracted, the resource block numbers of the corresponding service flows in the temporary resource allocation table are updated to the confirmed actual allocation values, and the pre-allocation validity period field is replaced with the formal allocation validity period, completing the conversion from the temporary resource allocation table to the formal resource allocation table. After the formal resource allocation table is generated, a switchover preparation completion identifier is generated. This identifier includes a timestamp, a hash digest of the formal resource allocation table, and a switchover sequence number, which are used for status synchronization and verification during the subsequent switchover execution phase.

[0118] The data pre-forwarding channel of the two-layer handover link pre-forwards buffered data packets from the current access network to the target access network, forwarding them sequentially according to their sequence numbers in the current access network's buffer queue without altering their original content. During pre-forwarding, a sequence number mapping relationship is established for each data packet: recording the mapping relationship between the original sequence number of the data packet in the current access network and its corresponding receive buffer position in the target access network. All mapping entries are summarized into a data packet sequence number mapping table. This mapping table is used by the target access network to reorder received data packets after handover, avoiding out-of-order packet problems caused by cross-network transmission.

[0119] Integrity verification is performed based on the packet sequence number mapping relationship. Each entry in the mapping relationship table is checked one by one: verifying whether the original sequence number exists in the transmission record on the current access network side, and verifying whether the packet content at the corresponding buffer position on the target network side is consistent with the original packet content. Consistency verification uses a packet-by-packet verification method, calculating the cyclic redundancy check value for the packet payload and comparing it with the original check value. If all packets pass the integrity verification, a verification pass flag is written to the extended field of the handover preparation completion identifier, indicating that the integrity of the pre-forwarded data has been confirmed. If any packet fails verification, it is retransmitted, and verification is performed again after retransmission until all packets pass verification.

[0120] After integrity verification is successful, a transmission permission switching instruction is generated based on the handover preparation completion flag. This instruction includes the handover execution time, a reference flag to the formal resource allocation table, a reference flag to the data packet sequence number mapping table, and a dual-network synchronization flag. When simultaneously sending the transmission permission switching instruction to both the current access network and the target access network, it is broadcast simultaneously via the control signaling channel, ensuring that both networks perceive the switching instruction at the same time. Upon receiving the instruction, the current access network stops sending new data to the train and transfers the handling of remaining buffered data packets to the target access network. Upon receiving the instruction, the target access network activates the allocated resource blocks according to the formal resource allocation table and reassembles the pre-forwarded data packets in an orderly manner based on the data packet sequence number mapping table, beginning to provide data transmission services to the train, thus completing a smooth handover of the entire handover process.

[0121] Figure 3 This is a handover signaling timing diagram of the rail transit vehicle-to-ground communication handover method based on heterogeneous network collaboration in an embodiment of the present invention, showing the time distribution of the five key stages in the handover process. Figure 3 The horizontal axis represents the five signaling stages: handover request, resource confirmation, pre-forwarding, handover instruction, and completion. The vertical axis represents the time consumed in each stage, in milliseconds.

[0122] from Figure 3 As can be observed, the handover request phase takes approximately 28 milliseconds, during which the onboard terminal sends a handover request message carrying a handover context data packet and a pre-allocated resource table to the target access network via the control signaling channel; the resource confirmation phase takes approximately 34 milliseconds, during which the target access network returns a resource confirmation response after completing resource reservation and access control; the pre-forwarding phase takes approximately 44 milliseconds, during which the current access network forwards the cached data packet to the target access network via the data pre-forwarding channel and establishes a data packet sequence number mapping relationship to ensure data integrity; the handover instruction phase takes approximately 18 milliseconds, during which the target access network generates a handover preparation completion flag and issues a transmission permission handover instruction via the control signaling channel; the completion phase takes approximately 12 milliseconds, during which the onboard terminal and the networks on both sides synchronize and execute the data path handover. The entire handover process takes approximately 136 milliseconds in total, with the pre-forwarding phase being the longest, demonstrating the crucial role of data cache migration in the handover process; the handover instruction and completion phases combined take approximately 30 milliseconds, corresponding to the actual communication interruption time during handover execution, meeting the stringent requirements of rail transit train control signaling for communication continuity. This timing diagram clearly shows the time consumption distribution of each signaling link after the handover decision, providing a quantitative basis for handover latency optimization.

[0123] Figure 4 This is a flowchart illustrating the two-layer handover link signaling interaction and data synchronization process of the rail transit vehicle-to-ground communication handover method based on heterogeneous network collaboration, as described in this embodiment of the invention.

[0124] In one alternative implementation,

[0125] Obtain a handover preparation completion flag from the resource reservation confirmation message received from the current access network, generate a transmission permission handover instruction based on the handover preparation completion flag, and execute it, including:

[0126] The system receives a resource reservation confirmation message from the target access network and extracts handover timing parameters. It then extracts a frame synchronization offset from the handover timing parameters and pre-compensates the transmission timing of the current access network based on the frame synchronization offset to obtain timing alignment parameters. Finally, it generates a handover preparation completion flag based on the timing alignment parameters.

[0127] Extract the timing alignment parameter from the switching preparation completion identifier and construct the transmission permission switching instruction by combining it with the service flow status in the real-time train operation parameters. Based on the timing alignment parameter, perform timestamp calibration on the switching execution time in the transmission permission switching instruction to obtain the synchronous switching time.

[0128] The transmission permission switching instruction is sent synchronously to the current access network and the target access network. The arrival of the synchronization switching time is monitored and the current access network is triggered to stop sending data, while the target access network is triggered to start receiving data.

[0129] After completing the resource reservation process for the target access network, a resource reservation confirmation message needs to be received from the target access network. This message carries timing reference information from the target network. The resource reservation confirmation message structurally contains multiple fields, among which the handover timing parameter field records information such as the target access network's frame timing reference, subframe boundary offset, and reference clock source. When extracting the handover timing parameters, the message body of the confirmation message is parsed. The frame count value, symbol offset, and clock precision level are read sequentially according to the predefined field offset addresses. This information is then combined into a structured handover timing parameter object for subsequent extraction of frame synchronization offsets.

[0130] The frame synchronization offset reflects the difference in frame timing reference between the current access network and the target access network. The frame start time of the target network is read from the handover timing parameters and compared with the frame start time of the current network maintained locally; the difference between the two is the frame synchronization offset. The unit is microseconds. In rail transit scenarios, the high-speed movement of trains causes signal propagation delay to change rapidly over time. Therefore, the frame synchronization offset is not a static value and needs to be corrected based on the current train position and motion state. Specifically, the propagation delay correction term is calculated based on the real-time distance estimate from the train to the target base station. ,Will Superimposed on the original frame synchronization offset The effective frame synchronization offset after motion compensation is obtained. ,satisfy .

[0131] Based on effective frame synchronization offset This involves pre-compensating the transmission timing of the currently accessed network. The purpose of pre-compensation is to ensure that the uplink signal sent by the train to the target access network after handover falls within the correct reception window at the target base station, avoiding uplink data loss or inter-symbol interference due to timing misalignment. The pre-compensation operation is achieved by adjusting the read pointer position of the local transmit buffer, advancing or delaying the read start position. The corresponding number of sampling points is used to align the transmission timing with the frame boundaries of the target network. After pre-compensation, The compensation direction flag and compensation execution status are packaged to generate timing alignment parameters. This parameter, as the core payload of the handover preparation completion indicator, signifies that the current transmission timing has been aligned with the target network.

[0132] When switching the preparation completion flag during generation, the timing alignment parameters will be adjusted. The encoding is a fixed-length binary field with an attached integrity checksum to ensure that timing parameters are not distorted due to transmission errors during subsequent extraction. When extracting timing alignment parameters from the handover preparation completion flag, the integrity checksum is first verified. After successful verification, the parameters are parsed according to the field offset. Each component includes the frame offset compensation amount, compensation direction, and effective compensation time window.

[0133] Transmission permission switching instructions are constructed by combining the service flow status from real-time train operation parameters. The service flow status includes the flow identifier, sequence number, buffer usage, and quality of service level for each currently transmitting service flow. When constructing the transmission permission switching instructions, an independent switching sub-instruction is generated for each service flow. This sub-instruction records the last transmission sequence number of that service flow at the switching moment. The target access network side uses this information to... Initiate reception to achieve seamless service flow transition. For service flows with high real-time requirements, a priority recovery flag is set in the sub-instruction. After the handover is completed, the target network prioritizes allocating radio resources to this type of service flow, shortening service interruption time.

[0134] The timestamp of the switching execution time in the transmission permission switching command is calibrated to obtain the synchronization switching time. The calibration process consists of two steps: First, using the local high-precision clock as a reference, the initial estimate of the switching execution time is... First, convert to an absolute timestamp format with a precision of at least 1 microsecond; second, utilize timing alignment parameters. The effective time window for compensation in Perform constraint checks to ensure If it falls within the valid compensation window, otherwise... Adjust to the nearest valid time to the window boundary, and finally obtain the synchronization switching time. Synchronization switching time The timestamp field written to the transmission permission switching instruction is used by the current access network and the target access network to perform local clock alignment and trigger judgment after receiving the instruction.

[0135] When simultaneously sending transmission permission switching instructions to the current access network and the target access network, the control signaling channel in the two-layer switching link is used to send identical instruction copies to both networks respectively, and the time interval between the two transmission operations does not exceed the preset maximum transmission time difference. This ensures that the time when both networks receive the command is as close as possible, reducing the risk of handover asynchrony caused by command arrival time differences. During transmission, the local time of each transmission is recorded. If the difference between two transmission times exceeds [a certain threshold], [further action is taken]. If this occurs, the retransmission mechanism is triggered, the command is resent and the timestamp field is updated. This ensures the effectiveness of the synchronization switching time.

[0136] Monitoring synchronization switching time The arrival time is achieved using a local clock polling mechanism with a fixed polling period. Detect the current time and The difference is used to determine the outcome when the difference is first less than or equal to half a polling cycle. Upon arrival, the handover execution process is immediately triggered. This process comprises two parallel actions: first, sending a stop data transmission command to the current access network. This command immediately halts dequeueing from the current network's transmission queue; data frames already in the physical layer transmission pipeline continue transmission but no new frames are appended. Simultaneously, a start data reception command is sent to the target access network. The target network activates the radio resources reserved for this handover, opens the uplink reception window, and initializes the reception sequence number to the value recorded in the sub-commands of each service stream. .

[0137] After the two parallel actions are completed, the handover execution result is confirmed. Handover completion acknowledgment messages are collected from both the current access network and the target access network, and the actual handover completion time is extracted from the acknowledgment messages. Synchronized switching time Compare and calculate the switching time deviation. .like Exceeding the preset switching time tolerance Then record the switching timing exception event and... An adaptive update process, fed back to the timing alignment parameters, is used to correct the estimation accuracy of frame synchronization offsets in subsequent handovers, thereby continuously optimizing the accuracy of handover timing control. Throughout the handover process, the continuity of service flow data is ensured by a sequence number concatenation mechanism and a retransmission request mechanism on the target network side, ensuring that the impact of handover on upper-layer applications is minimized.

[0138] A second aspect of this invention provides a rail transit vehicle-to-ground communication handover system based on heterogeneous network collaboration, comprising:

[0139] The trajectory prediction unit is used to collect real-time train operating parameters and signal quality parameters of the currently accessed network, perform time-series analysis on the signal quality parameters to extract signal attenuation trend features, perform mapping calculations on the signal attenuation trend features and real-time train operating parameters to obtain the predicted position trajectory, and determine the available switching time window based on the predicted position trajectory.

[0140] The network selection unit is configured to perform parallel channel probing on candidate heterogeneous networks within the coverage area of ​​the predicted location trajectory within the available switching time window to obtain real-time channel quality parameters and network coverage information; calculate the network coverage duration of each candidate heterogeneous network based on the predicted location trajectory and the network coverage information; solve for the network switching priority sequence by combining the real-time channel quality parameters; query the current load parameters corresponding to each candidate heterogeneous network in the network switching priority sequence to evaluate resource adequacy; and determine the target access network by combining the network switching priority sequence.

[0141] The handover execution unit is used to establish a two-layer handover link between the current access network and the target access network, send a handover request message to the target access network through the control signaling channel in the two-layer handover link, receive a resource reservation confirmation message from the current access network to obtain a handover preparation completion identifier, generate a transmission permission handover instruction based on the handover preparation completion identifier, and execute it.

[0142] A third aspect of the present invention provides an electronic device, comprising:

[0143] A processor and a memory for storing processor-executable instructions, wherein the processor is configured to invoke instructions stored in the memory to perform the aforementioned method.

[0144] A fourth aspect of the present invention provides a computer-readable storage medium having stored thereon computer program instructions that, when executed by a processor, implement the aforementioned method.

[0145] This invention can be a method, apparatus, system, and / or computer program product. The computer program product may include a computer-readable storage medium having computer-readable program instructions loaded thereon for performing various aspects of the invention.

[0146] 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 or all of the technical features; and these modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the scope of the technical solutions of the embodiments of the present invention.

Claims

1. A method for handover of rail transit vehicle-to-ground communication based on heterogeneous network collaboration, characterized in that, include: Collect real-time train operating parameters and signal quality parameters of the currently accessed network, perform time-series analysis on the signal quality parameters to extract signal attenuation trend features, perform mapping calculations on the signal attenuation trend features and real-time train operating parameters to obtain a predicted position trajectory, and determine the available switching time window based on the predicted position trajectory; Within the available switching time window, parallel channel probing is performed on candidate heterogeneous networks within the coverage area of ​​the predicted location trajectory to obtain real-time channel quality parameters and network coverage information. Based on the predicted location trajectory and the network coverage information, the network coverage duration of each candidate heterogeneous network is calculated, and the network switching priority sequence is obtained by combining the real-time channel quality parameters. The current load parameters corresponding to each candidate heterogeneous network in the network switching priority sequence are queried to evaluate resource sufficiency, and the target access network is determined by combining the network switching priority sequence. A two-layer handover link is established between the current access network and the target access network. A handover request message is sent to the target access network through the control signaling channel in the two-layer handover link. A handover preparation completion identifier is obtained by receiving a resource reservation confirmation message from the current access network. A transmission permission handover instruction is generated and executed based on the handover preparation completion identifier.

2. The method according to claim 1, characterized in that, Real-time train operating parameters and signal quality parameters of the currently accessed network are collected, and time-series analysis is performed on the signal quality parameters to extract signal attenuation trend characteristics. The predicted position trajectory is obtained by mapping the signal attenuation trend characteristics and the real-time train operating parameters. The available switching time window is then determined based on the predicted position trajectory, including: The train's real-time operating parameters are obtained by collecting the train's current position coordinates and operating information through an on-board sensor array, and the signal quality parameters are obtained by collecting the received signal strength indication value and signal interference noise ratio of the currently accessed network through a wireless communication interface. The signal quality parameters are sampled by a sliding window according to the time series to obtain multiple signal quality subsequences. The difference operation is then performed to obtain the signal change rate sequence. The signal change rate sequence is linearly fitted to obtain the attenuation trend function. Based on the attenuation trend function, the predicted time when the signal strength reaches the switching trigger threshold is calculated to obtain the signal attenuation trend feature. Based on the operational information, the train's displacement vector in the future time period is calculated and superimposed with the current position coordinates to obtain the future position coordinate sequence. The signal attenuation trend feature is spatially mapped and associated with the future position coordinate sequence to obtain the predicted position trajectory. The signal strength is calculated from the handover trigger threshold to the handover necessary threshold based on the attenuation trend function, and the time interval is obtained by spatiotemporally corresponding with the predicted position trajectory to obtain the handover start position point and the handover end position point. The available handover time window is determined based on the time corresponding to the handover start position point and the time corresponding to the handover end position point.

3. The method according to claim 1, characterized in that, Within the available switching time window, parallel channel probing is performed on candidate heterogeneous networks within the coverage area of ​​the predicted location trajectory to obtain real-time channel quality parameters and network coverage information. Based on the predicted location trajectory and the network coverage information, the network coverage duration of each candidate heterogeneous network is calculated, and the network switching priority sequence is obtained by combining the real-time channel quality parameters. This sequence includes: Based on the predicted location trajectory, candidate heterogeneous networks are queried and a channel probe request message is sent. The channel state information returned by the candidate heterogeneous networks is received and multipath components are separated. The delay spread is calculated based on the multipath components to obtain the channel coherence time. The received signal strength indication value is extracted based on the channel state information and combined with the channel coherence time to obtain the real-time channel quality parameters. The network coverage boundary coordinates are extracted based on the channel state information to obtain the network coverage range information. The sequence of trajectory coordinate points is extracted from the predicted location trajectory and the spatial distance between each trajectory coordinate point and the coordinates of the network coverage boundary is calculated. The cutting angle is calculated based on the spatial distance and the pre-acquired operation information, and the network coverage boundary is dynamically expanded according to the cutting angle to obtain the corrected coverage boundary. The network coverage duration is obtained by calculating the time difference between the coverage start time and the coverage end time based on the corrected coverage boundary. The train running speed is obtained to calculate the Doppler frequency shift, and the received signal strength indication value is frequency shift compensated to obtain the frequency shift compensated signal strength. The channel coherence time and the network coverage duration are combined with the frequency shift compensated signal strength to calculate the switching cost of each candidate heterogeneous network and arrange them in ascending order to obtain the network switching priority sequence.

4. The method according to claim 3, characterized in that, The Doppler frequency shift is calculated by obtaining the train speed, and frequency shift compensation is applied to the received signal strength indication value to obtain the frequency shift compensated signal strength. The switching cost of each candidate heterogeneous network is calculated by combining the channel coherence time and the network coverage duration with the frequency shift compensated signal strength, and then sorted in ascending order to obtain the network switching priority sequence, including: Train speed and direction of travel are extracted from the real-time train operating parameters. Doppler frequency shift is calculated for each candidate heterogeneous network based on the train speed and the carrier frequency of each candidate heterogeneous network. Subcarrier phase information is extracted from the channel state information and phase rotation compensation is performed on the subcarrier phase information based on the Doppler frequency shift to obtain a phase compensation matrix. Frequency domain coherent compensation is performed on the received signal strength indication value of each candidate heterogeneous network based on the phase compensation matrix to obtain the frequency shift compensated signal strength. The channel stability coverage ratio is calculated based on the channel coherence time and the network coverage duration. Angle spread information is extracted based on the channel state information and the beam alignment offset angle is calculated by combining the angle spread information and the train running direction. The channel stability coverage ratio is dynamically corrected based on the beam alignment offset angle to obtain a corrected channel stability coverage ratio. The handover cost of each candidate heterogeneous network is calculated based on the corrected channel stability coverage ratio and the frequency shift compensation signal strength, and the network handover priority sequence is obtained.

5. The method according to claim 1, characterized in that, Querying the current load parameters of each candidate heterogeneous network in the network handover priority sequence to assess resource sufficiency and determining the target access network in conjunction with the network handover priority sequence includes: Extract candidate heterogeneous network identifiers from the network switching priority sequence and send load query request messages to each candidate heterogeneous network. Receive current load parameters returned by each candidate heterogeneous network and extract physical resource block occupancy and scheduling authorization ratio based on the current load parameters. Based on the real-time train operation parameters, service transmission requirements are extracted. Channel quality indicator values ​​and multi-user interference information are extracted from the pre-acquired channel state information. Interference suppression correction is performed on the channel quality indicator values ​​based on the multi-user interference information to obtain effective channel quality indicator values. The achievable throughput is calculated by combining the physical resource block occupancy rate and the scheduling authorization ratio. The achievable throughput and the service transmission requirements are matched and verified to obtain throughput satisfaction. The number of user connections for each candidate heterogeneous network is obtained to calculate the network congestion probability. Based on the network congestion probability, the throughput satisfaction is dynamically corrected to obtain a corrected resource sufficiency evaluation value. Based on the corrected resource sufficiency evaluation value and the handover cost corresponding to each candidate heterogeneous network in the network handover priority sequence, a handover score is calculated, and the candidate heterogeneous network with the highest handover score is selected as the target access network.

6. The method according to claim 1, characterized in that, Establishing a two-layer handover link between the current access network and the target access network, and sending a handover request message to the target access network through the control signaling channel in the two-layer handover link, including: The system extracts the bearer status information from the real-time train operation parameters and constructs a service flow descriptor. It obtains the current session context information from the current access network and encapsulates it with the service flow descriptor to obtain a handover context data packet. Based on the service priority and latency sensitivity in the handover context data packet, it pre-allocates resource blocks to the access channel of the target access network to obtain a temporary resource allocation table and constructs a two-layer handover link based on the temporary resource allocation table. Based on the handover context data packet and the temporary resource allocation table, a handover request message is constructed and sent to the target access network through the two-layer handover link. The resource confirmation response returned by the target access network is received and the corresponding allocated resource block identifier is extracted. Based on the allocated resource block identifier, the temporary resource allocation table is updated to obtain the formal resource allocation table and a handover preparation completion identifier is generated. The cached data packets in the current access network are pre-forwarded to the target access network through the dual-layer switching link, and a data packet sequence number mapping relationship is established. Integrity verification is performed based on the data packet sequence number mapping relationship, and a transmission permission switching instruction is generated in conjunction with the switching preparation completion identifier. The transmission permission switching instruction is then sent synchronously to the current access network and the target access network to execute the transmission permission switching instruction.

7. The method according to claim 1, characterized in that, Obtain a handover preparation completion flag from the resource reservation confirmation message received from the current access network, generate a transmission permission handover instruction based on the handover preparation completion flag, and execute it, including: The system receives a resource reservation confirmation message from the target access network and extracts handover timing parameters. It then extracts a frame synchronization offset from the handover timing parameters and pre-compensates the transmission timing of the current access network based on the frame synchronization offset to obtain timing alignment parameters. Finally, it generates a handover preparation completion flag based on the timing alignment parameters. Extract the timing alignment parameter from the switching preparation completion identifier and construct the transmission permission switching instruction by combining it with the service flow status in the real-time train operation parameters. Based on the timing alignment parameter, perform timestamp calibration on the switching execution time in the transmission permission switching instruction to obtain the synchronous switching time. The transmission permission switching instruction is sent synchronously to the current access network and the target access network. The arrival of the synchronization switching time is monitored and the current access network is triggered to stop sending data, while the target access network is triggered to start receiving data.

8. A rail transit vehicle-to-ground communication handover system based on heterogeneous network collaboration, used to implement the method of any one of claims 1-7, characterized in that, include: The trajectory prediction unit is used to collect real-time train operating parameters and signal quality parameters of the currently accessed network, perform time-series analysis on the signal quality parameters to extract signal attenuation trend features, perform mapping calculations on the signal attenuation trend features and real-time train operating parameters to obtain the predicted position trajectory, and determine the available switching time window based on the predicted position trajectory. The network selection unit is configured to perform parallel channel probing on candidate heterogeneous networks within the coverage area of ​​the predicted location trajectory within the available switching time window to obtain real-time channel quality parameters and network coverage information; calculate the network coverage duration of each candidate heterogeneous network based on the predicted location trajectory and the network coverage information; solve for the network switching priority sequence by combining the real-time channel quality parameters; query the current load parameters corresponding to each candidate heterogeneous network in the network switching priority sequence to evaluate resource adequacy; and determine the target access network by combining the network switching priority sequence. The handover execution unit is used to establish a two-layer handover link between the current access network and the target access network, send a handover request message to the target access network through the control signaling channel in the two-layer handover link, receive a resource reservation confirmation message from the current access network to obtain a handover preparation completion identifier, generate a transmission permission handover instruction based on the handover preparation completion identifier, and execute it.

9. An electronic device, characterized in that, include: processor; Memory used to store processor-executable instructions; The processor is configured to invoke instructions stored in the memory to execute the method according to any one of claims 1 to 7.

10. A computer-readable storage medium having computer program instructions stored thereon, characterized in that, When the computer program instructions are executed by the processor, they implement the method described in any one of claims 1 to 7.

Images