Satcom integrated communication system

Abstract

The application relates to the field of communication technology and discloses a satellite communication scattering fusion communication system, which comprises a fusion control gateway, a satellite modem and a troposphere scattering device. The satellite modem and the troposphere scattering device are respectively connected to the fusion control gateway and feed back main and standby link state data upward. A link state monitoring module, a control plane module and a data scheduling module are arranged in the fusion control gateway. The link state monitoring module collects bottom layer state data to generate a cross-layer link comprehensive matrix; the control plane module extracts the matrix to perform threshold comparison, evaluates main link degradation expectation to activate a double-transmit mode, and outputs a state machine scheduling instruction; and the data scheduling module performs adaptive service replication according to the scheduling instruction in combination with a standby link instantaneous net load boundary, and compensates for the physical time delay difference of the bottom layer heterogeneous network through time offset of a sending side. The application utilizes physical layer feedback to dynamically intervene network layer scheduling, avoids service packet loss and disconnection, and prevents network congestion and message disorder.

Description

Technical Field

[0001] This invention relates to the field of communication technology, specifically to a satellite-to-scatter fusion communication system. Background Technology

[0002] In heterogeneous communication networks, satellite communication and tropospheric scattering communication are integrated as complementary network media to improve the reliability of the communication system. Existing integrated communication systems have shortcomings in link switching and data scheduling mechanisms. Traditional link switching often employs a hard switching mechanism, which is triggered only after the current communication link is interrupted, and cannot cope with the gradual attenuation of signals in the underlying physical links. When the primary link signal gradually deteriorates but is not completely disconnected, the system continues to send data on the damaged link, causing packet loss and network disconnection. To address the switching lag problem, some systems introduce a multi-transmission parallel mechanism, which simultaneously sends duplicate data on multiple links to ensure communication.

[0003] However, this concurrency mechanism failed to adjust to the dynamic carrying capacity of the backup communication equipment. When the data arrival rate exceeded the instantaneous available net load of the underlying backup equipment, it caused overflow of the underlying input queue and network channel congestion. Simultaneously, due to the physical time delay difference between long-distance satellite microwave propagation and short-distance tropospheric scattering, out-of-order message delivery occurred when data transmitted concurrently across different physical media arrived at the receiving end. The receiving node recalculated the asynchronously arriving data streams, causing buffer backlog at the receiving end and affecting the continuity of the overall network transmission. Summary of the Invention

[0004] The technical problem solved by this invention is that, when existing heterogeneous communication networks are integrated for transmission, traditional hard handover mechanisms are difficult to cope with service packet loss and disconnection caused by the gradual degradation of links; simple multi-transmission parallel mechanisms fail to combine the dynamic carrying capacity of backup communication equipment, which can easily lead to overflow of the underlying input queue and network congestion; in addition, there are objective physical delay differences in heterogeneous communication links that cross different physical media, and concurrent data transmission will cause packet out-of-order and buffer backlog at the receiving end, resulting in a decrease in the overall continuity and reliability of network transmission.

[0005] To address the above problems, the present invention provides the following technical solution: This invention provides a satellite scattering fusion communication system, including a fusion control gateway, a satellite modem, and a tropospheric scattering device; The converged control gateway is used to internally build a three-layer tunnel to shield the differences in the underlying heterogeneous network media, and to issue status and obtain instructions through an out-of-band management interface. Satellite modem, access converged control gateway, used to receive status acquisition commands and feed back primary link status data; Tropospheric scattering devices are connected in parallel to the fusion control gateway to synchronously receive status acquisition instructions and feed back backup link status data. The converged control gateway internally includes a link status monitoring module, a control plane module, and a data scheduling module; The link status monitoring module is used to collect primary link status data and backup link status data, and calculate and generate a cross-layer link comprehensive matrix.

[0006] The control plane module is used to extract the cross-layer link synthesis matrix, perform threshold comparison analysis, and output state machine scheduling instructions.

[0007] The data scheduling module is used to process business data entering the tunnel according to state machine scheduling instructions.

[0008] In the above scheme, network layer scheduling is dynamically intervened by utilizing the physical feedback of the underlying network medium to ensure the continuity of system business communication when the heterogeneous network environment fluctuates.

[0009] Furthermore, the data scheduling module internally includes a traffic classification unit, a service cloning unit, a delay buffer unit, and a deduplication unit; The traffic classification unit is used to classify the raw downlink traffic to generate high-quality traffic flows. The service cloning unit is used to copy and encapsulate high-priority service flows, generating main path transmission messages and bypass clone messages. The delay buffer unit is used to determine the dwell time of the bypass clone message according to the staggered dwell time calculation formula, and write the bypass clone message into the buffer queue with the same global identifier as the index. The bypass clone message is then allowed to dwell and wait for the bypass clone message and release it upon expiration according to the dwell time, so as to generate a time offset transmission sequence. The deduplication processing unit is used to perform deduplication processing on the arriving messages at the data receiving end.

[0010] Furthermore, the traffic classification unit is specifically used for: Intercept the raw downlink business execution differential service code point matching as business data, and identify the packets carrying key instructions, voice or preset priority video according to the pre-configured business weight strategy; The message carrying key instructions, voice or preset priority video is separated from the original downlink service to generate a high-priority service flow.

[0011] Furthermore, the service cloning unit is specifically used for: Upon receiving high-priority service flows, perform data packet replication and tunnel header sequence number encapsulation to generate the same global identifier; The same global identifier is used to encapsulate the main path transmission message and the bypass clone message respectively.

[0012] By generating a unified tunnel identifier, a basis is provided for dual-path redundant transmission across physical media.

[0013] Furthermore, the delay buffer unit is specifically used for: Receive bypass clone messages carrying the same global identifier, obtain the average transmission delay of the satellite link, the average transmission delay of the scatter link, and the standard deviation of the satellite link delay, and calculate the dwell time of the bypass clone message according to the staggered dwell time calculation formula. Write the bypass cloned message into an independent buffer queue, record the enqueue timestamp of the bypass cloned message, and determine the release time based on the enqueue timestamp and the dwell time of the bypass cloned message. When the current timeout reaches the release time, the resident lock state of the bypass clone message is released, and a time offset transmission sequence is generated according to the release time; The time-off transmission sequence is sent to the tropospheric scattering device by the physical network card, and the main transmission message is sent to the satellite modem for direct transmission.

[0014] Furthermore, the deduplication unit is specifically used for: When the converged control gateway acts as a data receiver, it parses the same global identifier for both the main path transmission message and the bypass clone message; A fixed-length sliding window is established based on the same global identifier; For multiple arriving messages corresponding to the same global identifier, a fixed-length sliding window is used to retain the first valid message that arrives first and discard the duplicate messages that arrive later.

[0015] This enables the same-source separation and reception of dual asynchronous transmission data streams under spatial and temporal isolation.

[0016] Furthermore, the primary link status data specifically includes the automatic gain control level and carrier-to-noise ratio; the backup link status data includes the symbol rate, modulation order, forward error correction code rate, link connection status flag, and received level.

[0017] Furthermore, the link status monitoring module is specifically used for: The primary link status data and backup link status data extracted from the register parameters of the satellite modem and tropospheric scattering device, as well as the equivalent one-way transmission delay of the satellite link, the equivalent one-way transmission delay of the scattering link, and the delay standard deviation of the two links are integrated and encapsulated to calculate and generate a cross-layer link comprehensive matrix. Based on the cross-layer link synthesis matrix, the primary link status data is extracted and substituted into the link degradation expectation evaluation formula based on multi-parameter dynamic weighting for calculation, so as to obtain the dynamic value of primary link degradation expectation at discrete time step.

[0018] This mechanism extracts the deterioration rate of physical layer parameters in the time domain, enabling it to extract and output quantitative predictive indicators before network layer services are interrupted.

[0019] Furthermore, the control plane module is specifically used for: Obtain the expected dynamic value of primary link degradation, and compare the expected dynamic value of primary link degradation with the preset state machine trigger high-risk threshold; When the expected dynamic value of primary link degradation is determined to be greater than or equal to the state machine triggering high-risk threshold, dual-mode is activated. Based on the dual-mode encapsulation, a state machine scheduling instruction containing the logic for activating the dual-mode is generated and output; The state machine triggering high-risk threshold is calibrated based on the actual tolerance of the communication system for service packet loss and the statistical average of the index degradation before the service disconnection in historical communication data.

[0020] Furthermore, the control plane module is also used for: Extract various modulation parameters from the backup link status data, substitute them into the dynamic estimation formula of the available net load of the backup link, and calculate to obtain the instantaneous net load boundary of the backup link. The instantaneous net load boundary of the backup link is encapsulated as a capacity constraint parameter into the state machine scheduling instruction and output to the data scheduling module.

[0021] By calculating the instantaneous theoretical net bandwidth available to the underlying device through adaptive physical modulation parameters, channel congestion caused by unrestrained replication can be prevented.

[0022] Furthermore, the control plane module is also used for: Simultaneously activate the dual-fire mode and start the shortest window timer configured synchronously; After the system maintains dual-mode and the shortest window timer reaches the timeout condition, monitor the expected dynamic value of the primary link degradation. If the expected dynamic value of primary link degradation falls below the preset state machine exit safety threshold and meets the window hold timer period, then exit dual-transmission mode and output the corresponding state machine scheduling instruction for state recovery. Among them, the value of the state machine exit safety threshold is less than the state machine trigger high-risk threshold.

[0023] The aforementioned dual-threshold and timing constraint evaluation logic constructs a control hysteresis interval, avoiding repeated switching of the system's transmission mode caused by fluctuations in the underlying signal.

[0024] Furthermore, the service cloning unit is specifically used for: Extract the instantaneous net load boundary of the backup link from the state machine scheduling instructions, and calculate the current arrival rate by combining it with the received high-priority service flow; When the current arrival rate is assessed to be greater than the instantaneous net load boundary of the backup link, adaptive degradation logic is triggered. Based on the adaptive degradation logic, relatively less important data packets are discarded in sequence according to the sub-priority weights configured inside the packet, and only the subset of packets that meet the capacity constraints are retained for replication.

[0025] Adaptive congestion-prevention replication is performed using instantaneous net load boundaries to ensure dual-redundant concurrent transmission of high-value data while maintaining queue stability of physical devices.

[0026] Furthermore, the delay buffer unit is specifically used for: The average one-way transmission delay of the satellite link, the average one-way transmission delay of the scatter link, and the standard deviation of the satellite link delay are periodically calculated as delay samples. The dwell time of bypass cloned messages is calculated by using the staggered dwell time calculation formula on the delay samples; When the cumulative running time of the independent backoff timer reaches the dwell time of the bypass clone message, the dwell lock state of the bypass clone message is released and a time offset transmission sequence is generated.

[0027] This staggered scheduling mechanism introduces time offset intervention from the transmitting side to compensate for the inherent spatial transmission delay difference between long-distance microwave propagation and short-distance tropospheric scattering, thereby enabling data to arrive at the receiving end in relative synchronization and alleviating the computational pressure of reordering at the receiving node caused by out-of-order delivery.

[0028] Furthermore, the control plane module is also used for: During the continuous operation of the dual-transmission mode, when the received level of the data extracted from the backup link status drops below the preset backup link connection interruption threshold, a bypass no-response determination is triggered. When the bypass non-response determination is triggered, a suspension command is sent to the data scheduling module to indicate that the subsequent business replication process is suspended; The backup link connection interruption threshold is calibrated based on the factory-set lower limit of the sensitivity of the tropospheric scattering device hardware receiver.

[0029] Furthermore, the deduplication unit is specifically used for: The window length of the fixed-length sliding window is set according to the maximum delay difference between the main path and the bypass path, the link jitter range, and the buffer capacity of the receiver. When the same global identifier exceeds the scope of the fixed-length sliding window or exceeds the preset holding time, the corresponding deduplication status record is deleted.

[0030] The satellite communication scattering fusion communication system provided by this invention has the following beneficial effects: This invention constructs a cross-layer tunnel to shield the physical differences between satellite and tropospheric scattering heterogeneous network media through a converged control gateway. It also actively collects link status data from underlying primary and backup devices using a link status monitoring module to generate a cross-layer link synthesis matrix. The control plane module then compares and analyzes this matrix and outputs scheduling commands, directly driving the data scheduling module to process service data. This overall architecture breaks the limitations of traditional isolation between the network and physical layers, constructing a cross-layer closed-loop dynamic feedback mechanism from the perception of the actual physical link status at the bottom layer to the processing of service data at the upper layer. This enables the system to sensitively capture environmental fluctuations in the underlying medium and intervene in network layer scheduling in real time, significantly improving the overall robustness, resource utilization, and continuity of service transmission of the heterogeneous converged communication system in complex channel environments.

[0031] This invention extracts underlying link status data through a link status monitoring module to calculate the expected dynamic value of primary link degradation. When the value reaches a preset high-risk threshold, the control plane module activates a dual-transmission mode. This method can output quantified predictive indicators before the underlying physical signal is gradually attenuating and before network layer services are interrupted. This avoids packet loss and connection interruptions caused by the lag in traditional hard handover mechanisms, ensuring the continuity of service communication during fluctuations in heterogeneous network environments.

[0032] This invention utilizes a control plane module to extract modulation parameters from backup link status data and calculate the instantaneous net load boundary of the backup link. When the service arrival rate exceeds this boundary, the service cloning unit triggers adaptive degradation logic, retaining a subset of packets that meet the capacity constraints for replication based on sub-priority weights. This approach combines the dynamic carrying capacity of backup communication equipment for service transmission control, preventing the underlying input queue overflow and network congestion problems caused by a simple multi-transmission parallel mechanism.

[0033] This invention employs a delay buffer unit to calculate the dwell time of bypass cloned messages based on delay samples, introduces time offset intervention from the transmitting side to generate a time offset transmission sequence, and combines this with a deduplication unit to perform source stripping and deduplication of arriving messages using a fixed-length sliding window. This method compensates for the objective physical time delay difference between long-distance satellite microwave transmission and short-distance tropospheric scattering transmission, ensuring that data arrives relatively synchronously at the receiving end. This alleviates message out-of-order delivery and buffer backlog caused by dual-path asynchronous transmission at the receiving end, and improves the overall network transmission continuity. Attached Figure Description

[0034] Figure 1 This is a diagram showing the overall physical network and internal entity partitioning of the system according to an embodiment of the present invention; Figure 2 This is a macroscopic step-by-step collaborative workflow diagram of an embodiment of the present invention; Figure 3This is a schematic diagram illustrating the principle of the cross-layer parameter acquisition and degradation assessment mechanism in an embodiment of the present invention. Figure 4 This is a flowchart of the anti-vibration state machine triggering and net load boundary calculation according to an embodiment of the present invention; Figure 5 This is a flowchart illustrating the restricted cloning and tunnel identifier encapsulation process according to an embodiment of the present invention; Figure 6 This is a flowchart of the heterogeneous delay difference compensation scheduling process according to an embodiment of the present invention; Figure 7 This is a flowchart illustrating the tunnel space deduplication mechanism and state machine recovery process according to an embodiment of the present invention. Figure 8 This is a comparison chart of the communication packet loss rate over time in an embodiment of the present invention. Detailed Implementation

[0035] The technical solutions in 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.

[0036] See Figure 1 The present invention provides a satellite scattering fusion communication system, which may include: a fusion control gateway, a satellite modem, and a tropospheric scattering device.

[0037] The converged control gateway internally constructs a general routing encapsulation or user datagram protocol cross-layer three tunnel to shield the differences in the underlying heterogeneous network media. The converged control gateway issues status acquisition instructions to subordinate devices through the simple network management protocol interface or through the monitoring and control (M&C) out-of-band management interface.

[0038] The satellite modem connects to the converged control gateway via Ethernet or serial port, receives status acquisition instructions, and feeds back the primary link status data.

[0039] The tropospheric scattering device is connected to the converged control gateway in parallel via an Ethernet interface. It synchronously receives status acquisition instructions and feeds back backup link status data.

[0040] The converged control gateway internally includes a link status monitoring module, a control plane module, and a data scheduling module.

[0041] The link status monitoring module is used to collect primary link status data and backup link status data, and it calculates and generates a cross-layer link comprehensive matrix.

[0042] The control plane module is used to extract the cross-layer link synthesis matrix for threshold comparison analysis, and its output state machine scheduling instructions.

[0043] The data scheduling module is used to process the service data entering the tunnel according to the state machine scheduling instructions. It includes a traffic classification unit, a service cloning unit, a delay buffer unit, and a deduplication unit.

[0044] The traffic classification unit is used to intercept the original downlink business execution differential service code point matching, which generates high-priority business flows.

[0045] The service cloning unit is used to receive high-priority service flows and perform data packet copying on them. At the same time, it adds three layers of tunnel encapsulation to the original message and assigns independent sequence numbers, and outputs main path transmission messages and bypass clone messages with the same global identifier.

[0046] The delay buffer unit receives bypass clone packets with the same global identifier, obtains the average transmission delay of the satellite link, the average transmission delay of the scatter link, and the standard deviation of the satellite link delay output by the link status monitoring module, and calculates the dwell time of the bypass clone packets according to the staggered dwell time calculation formula. Using the same global identifier as an index, the delay buffer unit writes the bypass clone packets into an independent buffer queue, records their enqueue timestamp, and determines the release time based on the enqueue timestamp and the dwell time of the bypass clone packets. When the current timeout reaches the release time, the delay buffer unit releases the dwell lock state of the corresponding bypass clone packets and outputs a time offset transmission sequence according to the release time, which is then sent to the underlying physical network interface card (NIC).

[0047] The deduplication unit, when acting as a data receiver at its node, parses the global identifier in the tunnel header of both the main route and the bypass cloned messages, and establishes a fixed-length sliding window. For multiple arriving messages corresponding to the same global identifier, the deduplication unit retains the first valid message that arrives and delivers it upwards, discarding subsequent duplicate messages within the window. When the global identifier exceeds the current sliding window range or exceeds the preset retention time, the corresponding deduplication status record is deleted. The window length of the fixed-length sliding window is set based on the maximum delay difference between the main route and the bypass, the link jitter range, and the receiver's buffering capacity.

[0048] See Figure 2 This invention provides a satellite communication scattering fusion method, comprising the following steps: S1, the converged control gateway extracts the underlying register parameters of the communication equipment to calculate the expected dynamic value of link degradation; S2, the converged control gateway determines that the expected dynamic value of link degradation exceeds the preset threshold and activates the dual-transmission mode, outputting the instantaneous net load boundary of the backup link; where, the dual-transmission mode refers to the system's operating state of redundantly replicating and concurrently transmitting the same application layer business data through the primary communication link and the backup communication link at the same time. S3, the converged control gateway performs asymmetric replication and tunnel header identifier encapsulation on network layer services based on the instantaneous net load boundary of the backup link, and generates main route transmission messages and bypass clone messages; S4, the converged control gateway performs direct transmission of the main path original message, and performs buffering, timed waiting and expiration release of the bypass clone message based on the bypass clone message dwell time calculation formula, forming a spatially and temporally isolated dual asynchronous transmission data stream; S5: The peer communication node receives and parses the arriving dual asynchronous transmission data stream, uses a fixed-length sliding window to perform the stripping and discarding of duplicate messages from the same source, and shuts down the dual transmission mode and restores the single-link operation state after determining that the physical environment parameters meet the exit conditions.

[0049] The following section provides a detailed explanation of the specific control mechanisms and quantitative calculation models for each operational step in the satellite scattering fusion communication system.

[0050] See Figure 3 The cross-layer parameter acquisition and degradation assessment step S1 in the satellite scattering fusion communication system provided by the present invention specifically includes the following sub-steps: S101 In this embodiment, in order to obtain the operating status of the underlying network medium, the link status monitoring module inside the fusion control gateway generates a status acquisition instruction and sends the status acquisition instruction to the satellite modem and the tropospheric scattering device, and accordingly receives the primary link status data returned by the satellite modem and the backup link status data returned by the tropospheric scattering device.

[0051] As a preferred approach, the converged control gateway sends status acquisition commands in parallel to the underlying satellite modem and tropospheric scattering device via a Simple Network Management Protocol (SMLP) interface or a monitoring and control interface. When the satellite modem receives the status acquisition command, it reads its internal physical hardware registers and feeds back the primary link status data. In this information exchange, the primary link status data specifically includes the automatic gain control level and carrier-to-noise ratio. The tropospheric scattering device synchronously receives the status acquisition command, reads its currently effective physical layer operating parameters and link connectivity information, and feeds back the backup link status data. This backup link status data includes at least: symbol rate, modulation order, forward error correction code rate, link connection status flag, and receive level. For reading and writing operations of the communication device hardware registers, those skilled in the art can use standard device driver access technology, which is well-known in the field and will not be elaborated here. Based on the acquisition of the above-mentioned underlying hardware parameters, the link status monitoring module centrally collects the primary link status data and the backup link status data.

[0052] S102, to further understand the transmission characteristics at the network layer, the converged control gateway sends probe messages with timestamps and link identifiers to the peer communication node via the primary satellite link and the backup scatter link, respectively. The peer communication node sends back corresponding response messages. The link status monitoring module calculates the round-trip delay of the primary satellite link and the backup scatter link based on the link identifier, and calculates the equivalent one-way transmission delay of the satellite link and the equivalent one-way transmission delay of the scatter link, provided that the sending end and the peer communication node have synchronized their clocks.

[0053] To capture network status fluctuations, the link status monitoring module establishes a recent sliding time window based on the historical observation duration set by the system. It then statistically records satellite link delay samples and scattering link delay samples within this recent sliding time window, calculating their average delay and jitter statistics respectively. The satellite link jitter statistics are expressed using the standard deviation of satellite link delay, and the scattering link jitter statistics are expressed using the standard deviation of scattering link delay. The specific calculation of these delay standard deviations can be achieved by those skilled in the art using conventional mathematical statistical formulas, which are common knowledge in the field and will not be elaborated upon here.

[0054] S103 After completing multi-dimensional data acquisition, the link status monitoring module will extract the underlying register parameters of the communication equipment (including primary link status data and backup link status data), the equivalent one-way transmission delay of the satellite link, the equivalent one-way transmission delay of the scattering link, and the delay standard deviation of the two links, integrate and encapsulate them, and calculate and generate a cross-layer link comprehensive matrix.

[0055] Specifically, the mathematical structure of the cross-layer link synthesis matrix is ​​defined as a two-dimensional matrix, where the row dimension corresponds to different communication link objects, and the column dimension corresponds to the characteristics of various cross-layer state parameters. In this embodiment, the matrix is ​​configured with a 2-row × N-column structure, where the first row is associated with the primary satellite link, and the second row is associated with the backup scattering link. The specific data encapsulation format of the matrix and its row and column correspondence are as follows: Column 1 is the physical link connectivity flag; Column 2 is the physical layer receive level (automatic gain control level for the primary link, and hardware receive level for the backup link); Column 3 is the signal quality parameter (carrier-to-noise ratio for the primary link); Column 4 is the network layer equivalent one-way transmission delay; Column 5 is the network layer delay standard deviation; Columns 6 to N are the physical layer modulation and coding parameters for a specific link (for backup scattering links, these are symbol rate, modulation order, and forward error correction code rate, in that order; for the corresponding column positions of the primary satellite link, preset default values ​​or zero values ​​are used for placeholder alignment).

[0056] The specific steps for generating the cross-layer link synthesis matrix are as follows: the link status monitoring module obtains the original parameters of the above input dimensions according to a fixed sampling period; performs numerical type conversion and dimension normalization processing on parameters with different physical units; finally, according to the above preset row and column index mapping rules, the normalized data is sequentially written into the shared memory buffer of the control plane module to complete the construction and encapsulation of the matrix.

[0057] For the basic calculation of network layer latency, those skilled in the art can use the general Internet Control Message Protocol, which is a well-known technology in this field and will not be elaborated here.

[0058] S104. Based on the constructed cross-layer link synthesis matrix, the link status monitoring module extracts the primary link status data to perform link degradation prediction calculations and obtains the expected dynamic value of primary link degradation.

[0059] In real-world communication scenarios, satellite communication link quality exhibits a gradual attenuation characteristic due to weather conditions. Before substantial packet loss or connection failure occurs at the network layer, the automatic gain control level and carrier-to-noise ratio of the underlying receiver will show a continuous downward trend. Based on this principle, the system extracts the degradation rate of physical layer parameters in the time domain and outputs quantified predictive indicators before the primary link service is interrupted.

[0060] In this embodiment, the link degradation prediction calculation is performed using a link degradation expectation evaluation formula based on multi-parameter dynamic weighting, thereby obtaining the first... The system calculates the expected dynamic value of primary link degradation at discrete time steps, and then determines whether the expected dynamic value of link degradation exceeds a preset threshold to activate dual-mode. The link degradation expectation evaluation formula based on multi-parameter dynamic weighting is as follows: ; In the formula, Indicates the first The expected dynamic value of primary link degradation at each discrete time step; This represents the static weighting coefficient indicating the rate of level degradation. The static weighting coefficient represents the rate of degradation of the carrier-to-noise ratio; Indicates the first Automatic gain control level value for each discrete time step; Indicates the first Automatic gain control level value for each discrete time step; Indicates the first The carrier-to-noise ratio at each discrete time step; Indicates the first The carrier-to-noise ratio at each discrete time step; Indicates a fixed sampling period; Indicates the rate of degradation of the automatic gain control level; This indicates the rate of degradation of the carrier-to-noise ratio.

[0061] As one implementation method, to avoid division-by-zero errors during the calculation process, the fixed sampling period is set to a time constant greater than zero. Since the automatic gain control level and the carrier-to-noise ratio (CNR) are measured on different scales, the introduced static weighting coefficients for both the level degradation rate and the CNR degradation rate not only characterize the contribution of different parameters to the physical availability attenuation but also serve a dimension normalization function, ensuring that the accumulated terms are calculated within the same order of magnitude. Under typical configurations, the static weighting coefficients for both the level degradation rate and the CNR degradation rate are both greater than 0 and less than 1.

[0062] As can be seen from the formula structure, the link degradation prediction assessment formula based on multi-parameter dynamic weighting internally employs a difference structure where the parameter of the previous time step is subtracted from the parameter of the current time step. When the physical environment of the link deteriorates, causing a decrease in the automatic gain control level or carrier-to-noise ratio, the automatic gain control level degradation rate or carrier-to-noise ratio degradation rate outputs a positive value, and the final calculated value is... The expected dynamic value of primary link degradation at each discrete time step is represented as a positive real number. The fusion control gateway will calculate the expected dynamic value of link degradation based on this and trigger the subsequent anti-oscillation state machine protection logic.

[0063] See Figure 4 The anti-oscillation state machine triggering and net load boundary calculation step S2 in the satellite scattering fusion communication system of this invention specifically includes the following sub-steps: In this embodiment, to avoid frequent switching between single-shot and dual-shot modes caused by minor fluctuations in the network environment, the control plane module is configured with dual-threshold and timing constraint evaluation logic.

[0064] S201, the control plane module obtains the output from the link status monitoring module. The dynamic value of expected primary link degradation at each discrete time step is used to determine whether it is greater than or equal to a preset state machine trigger high-risk threshold. Specifically, as a preferred approach, to construct an anti-oscillation hysteresis interval, the value of the state machine trigger high-risk threshold is set to be greater than the state machine exit safety threshold. The specific values ​​of these preset thresholds are calibrated based on the actual tolerance of the communication system to service packet loss and the statistical average of indicator degradation before service disconnection in historical communication data.

[0065] When determining the first When the expected dynamic value of the primary link degradation at each discrete time step is greater than or equal to the high-risk threshold triggered by the state machine, the control plane module determines that the primary link faces a risk of service interruption and activates the dual-transmission mode. To further suppress the ping-pong effect near the critical value, the control plane module starts the synchronously configured shortest window timer while activating the dual-transmission mode. During the operation of the shortest window timer, even if the subsequently acquired value is greater than or equal to the expected dynamic value of the primary link degradation at each discrete time step, the control plane module will activate the shortest window timer. The expected dynamic value of the primary link degradation at each discrete time step decreases, and the system still maintains its dual-mode operation. When the determination of the... When the expected dynamic value of primary link degradation at each discrete time step is less than the high-risk threshold triggered by the state machine, the single-shot mode is maintained and the system returns to retrieve the output from the link status monitoring module. The steps for calculating the expected dynamic value of primary link degradation in discrete time steps.

[0066] If the system maintains dual-mode and the shortest window timer reaches the timeout condition, and if the first... When the expected dynamic value of the primary link degradation at each discrete time step falls below the preset state machine exit safety threshold, the control plane module immediately starts a window-closing hold timer. If this indicator remains below the state machine exit safety threshold throughout the entire cycle of the window-closing hold timer, the control plane module determines that the primary link has stably recovered and then exits the dual-transmission mode. For the basic scheduling implementation of the timer, those skilled in the art can use hardware timer interrupts at the operating system level, which is a well-known technology in the field and will not be elaborated upon here.

[0067] S202 During the continuous operation cycle of dual-transmission mode, considering that the backup communication medium corresponding to the tropospheric scattering device is also likely to be degraded by environmental interference, the control plane module synchronously executes the bypass fallback fault tolerance mechanism.

[0068] The control plane module continuously monitors the backup link status data extracted from the tropospheric scattering device. If, within the set observation period, the link connection status flag fed back by the tropospheric scattering device shows as disconnected, or its received signal level drops below the preset backup link connection interruption threshold, the control plane module will determine that the backup link is in an unresponsive state. In this case, the backup link connection interruption threshold can be calibrated according to the factory sensitivity lower limit of the tropospheric scattering device's hardware receiver.

[0069] When the above bypass non-response determination is triggered, the control plane module sends a suspension command to the data scheduling module, instructing it to suspend the subsequent service replication process and clear the underlying reserved queue cache resources until the latest acquired backup link status data confirms that the physical connection has been restored.

[0070] S203. Under normal dual-transmission mode, to prevent congestion caused by exceeding the capacity of the backup link, the system needs to quantitatively assess the current redundancy service boundary it can accommodate. Given that the tropospheric scattering equipment uses adaptive coding and modulation technology, its physical layer transmission bandwidth dynamically adjusts with changes in spatial channel quality. Therefore, the control plane module dynamically calculates the actual available capacity of the underlying layer by extracting various modulation parameters from the backup link status data.

[0071] In this embodiment, the control plane module uses the dynamic estimation formula for available net load of the backup link to process the above parameters, thereby obtaining the instantaneous net load boundary of the backup link. Based on this, the instantaneous net load boundary of the backup link is output as the data volume basis for guiding the data scheduling module to perform service cloning. The dynamic estimation formula for available net load of the backup link is as follows: ; In the formula, Indicates the instantaneous net load boundary of the backup link; Indicates the engineering margin coefficient; Indicates the symbol rate of the backup link; Indicates the modulation order of the backup link; Indicates the forward error correction code rate of the backup link; Indicates the percentage of fixed overhead at the underlying level; Indicates the background traffic throughput of the scattering link; Indicates the percentage of effective payload; This indicates the theoretically available net bandwidth of the backup link.

[0072] In practical engineering applications, the engineering margin coefficient is set between 0.8 and 0.9. This setting is used to absorb unpredictable background signaling traffic in the underlying network and ensure the reliability of the estimation results. The underlying fixed overhead ratio is jointly determined by the physical frame structure header length of the tropospheric scattering device and the Ethernet Media Access Control layer header length, and is used as a fixed empirical value in the calculation. The backup link symbol rate, backup link modulation order, and backup link forward error correction code rate are all derived from the backup link status data obtained in the preceding steps. In addition, the background service throughput of the scattering link is periodically obtained through the network traffic monitoring interface at the bottom layer of the converged control gateway.

[0073] By applying the aforementioned calculation rules, the control plane module can quantify the upper limit of bandwidth redundancy based on the adaptively and dynamically changing physical layer modulation and coding parameters. If, during the calculation, the theoretically available net bandwidth of the backup link is less than the background service throughput of the scattering link, the control plane module sets the instantaneous net load boundary of the backup link to zero through system settings to prevent negative logical anomalies in subsequent service scheduling. If, during the calculation, the theoretically available net bandwidth of the backup link is greater than or equal to the background service throughput of the scattering link, the instantaneous net load boundary of the backup link is calculated and generated. This calculated instantaneous net load boundary of the backup link is ultimately transmitted to the data scheduling module as the basis for the data volume of service cloning, used to fulfill the output requirements defined in the macro steps and constrain the data volume in the next step.

[0074] See Figure 5 The priority-constrained restricted cloning and tunnel identifier encapsulation step S3 in the satellite communication scattering fusion system provided by the present invention specifically includes the following sub-steps: S301 intercepts downlink business execution differential service code point matching, strips and generates high-optimal business flows.

[0075] In this embodiment, the data scheduling module performs subsequent scheduling operations based on the instantaneous net load boundary of the backup link output from the preceding steps. In this stage, the traffic classification unit deployed within the data scheduling module intercepts raw downlink service data entering the communication system and performs deep inspection on network layer packets.

[0076] As a preferred approach, the traffic classification unit extracts the service type field from the Internet Protocol header in the raw downlink service data and performs differential service code point matching. Based on a pre-configured service weight strategy, the system identifies packets carrying critical instructions, voice, or video with preset priorities. Specifically, packets carrying critical instructions refer to high-reliability control packets used to control the operating status of underlying devices, issue system configuration parameters, or perform emergency operations, such as remote control switching commands or keep-alive signaling from network routing protocols. The aforementioned pre-configured service weight strategy is generated by the network administrator through static mapping based on application layer protocol types or IP address 5-tuple information. When a corresponding high-priority code point is matched, the traffic classification unit extracts these packets from the raw downlink service, thereby generating a high-priority service flow. The basic logic implementation of network layer packet parsing and differential service code point matching can be accomplished by those skilled in the art using standard network protocol stack parsing techniques, which are well-known in the field and will not be elaborated upon here.

[0077] S302, the service cloning unit executes a restricted adaptive degradation anti-congestion replication mechanism based on the current arrival rate of the high-priority service flow and the instantaneous net load boundary of the backup link issued by the control plane module.

[0078] Specifically, network-layer packet replication technology improves communication reliability through multipath transmission. However, when dealing with the objective differences in physical bandwidth of heterogeneous underlying links, the system introduces a dynamic control mechanism based on underlying feedback. Based on the core data extracted from the traffic classification process, the service cloning unit receives the high-priority service flows and prepares to perform cross-link dual replication. In an uninterrupted operating environment, if full replication is performed without limiting all high-priority service flows, the excess redundant traffic will increase the overall system load and exceed the current carrying capacity of the backup links, thereby leading to the risk of input queue overflow and congestion packet loss for the tropospheric scattering equipment.

[0079] To prevent the aforementioned situation, the service cloning unit, in conjunction with the instantaneous net load boundary of the backup link issued by the control plane module, executes a restricted adaptive degradation anti-congestion replication mechanism. In specific implementation, the service cloning unit utilizes a built-in statistical sliding time window to calculate the current arrival rate of high-priority service flows and compares it numerically with the instantaneous net load boundary of the backup link. The length of the statistical sliding time window is set according to the hardware forwarding cycle of the underlying equipment.

[0080] When the current arrival rate is assessed to be less than or equal to the instantaneous net load boundary of the backup link, the service cloning unit performs full data packet replication on the incoming high-priority service flow. When the current arrival rate is assessed to be greater than the instantaneous net load boundary of the backup link, the service cloning unit triggers adaptive degradation logic, discarding relatively less important data packets in ascending order according to the sub-priority weights configured within the packets, and retaining only a subset of packets that meet the capacity constraints of the instantaneous net load boundary for replication. To avoid logical dead zones, if there are packets with the same sub-priority that still need to be discarded to meet capacity constraints, the service cloning unit uses a first-in-first-out queue tail-drop strategy. Through this dynamic decision controlled by the net load boundary, the system ensures dual redundancy for high-priority service flows while also maintaining the stable state of the physical layer input queue of the backup link.

[0081] S303, tunnel encapsulation technology achieves transparent transmission across heterogeneous physical media by adding a new network layer header to the original payload. After completing the above-mentioned restricted replication logic evaluation, the system faces the technical problem of how to identify redundant identities by peer communication nodes. To address this, the service cloning unit generates the same global identifier for the same application layer data payload and encapsulates it separately to form the main transmission message and the bypass clone message.

[0082] In this embodiment, the service cloning unit encapsulates and transmits packets based on a three-layer tunnel, appending the global identifier along with the general routing encapsulation tunnel header or the User Datagram Protocol (UDP) tunnel header to the outside of the packet. The tunnel headers of the primary route transmission packet and the bypass cloning packet carry the same global identifier, enabling the peer communication node to perform same-source packet identification and deduplication. Specifically, the primary route transmission packet is used for transmission via the satellite link, while the bypass cloning packet is used for transmission via the scatter link.

[0083] After separation and encapsulation, the generated main transmission packets are delivered via the default route to the underlying satellite modem for direct transmission. Simultaneously, the generated bypass clone packets with a global identifier are directed to the delay buffer unit within the data scheduling module to await heterogeneous delay difference compensation and timeout release in the next step. The basic code encapsulation for the construction and field filling of the three-layer tunnel header can be implemented using standard socket programming by those skilled in the art, which is well-known in the field and will not be elaborated upon here.

[0084] like Figure 6 As shown, the heterogeneous delay difference compensation scheduling step S4 in the satellite-to-scatter fusion communication system of this invention specifically includes the following sub-steps: S401, in this embodiment, the data scheduling module receives the main path original packet and the bypass clone packet generated by the previous step of separation and encapsulation. To avoid internal resource contention between the two underlying network paths during concurrent scheduling, the data scheduling module adopts a dual-queue logical isolation architecture to perform traffic splitting processing on the above heterogeneous data.

[0085] In practice, the data scheduling module, based on the differences in data types, directly imports the primary route's original packets into the underlying direct transmission queue to initiate hardware-level transmission. Simultaneously, the data scheduling module guides the bypass cloned packets to a dedicated memory area configured within the delay buffer unit for temporary storage. This queue separation mechanism achieves memory isolation between primary and bypass data within the system, forming a spatially and temporally isolated dual-path asynchronous data transmission stream, providing the necessary buffer foundation for subsequent time offset control.

[0086] S402, as a key constraint for heterogeneous network converged transmission, determines that network communication latency is composed of the time consumed in propagation, queuing, and processing stages. In satellite-to-ground microwave scattering converged communication systems, the transmission path of the satellite link is much longer than that of the terrestrial microwave scattering link. This inherent physical latency difference can easily cause out-of-order arrival of bypass cloned messages before the original messages on the main path at the receiver, leading to high buffering and sorting pressure or passive packet loss at the receiver.

[0087] To offset transmission differences across spatial media over time, the delay buffer unit constructs a staggered dwell time model based on historical network state information. During this process, the delay buffer unit obtains delay samples from the link state monitoring module for both links and periodically calculates the average one-way transmission delay of the satellite link, the average one-way transmission delay of the scattering link, and the standard deviation of the satellite link delay. As a preferred approach, the delay buffer unit uses a staggered dwell time calculation formula to derive the dwell time of bypass cloned messages. The staggered dwell time calculation formula is as follows: ; In the formula, Indicates the dwell time of bypass clone messages; Indicates the average transmission delay of the satellite link; Indicates the average transmission delay of the scattering link; This represents the extreme value jitter compensation factor; Indicates the standard deviation of satellite link delay; This represents the difference in basic latency between the two links; This indicates the amount of jitter tolerance compensation.

[0088] Based on engineering implementation experience, the extreme jitter compensation factor is set between 1.5 and 2.0. This setting is used to absorb time delay jitter caused by orbital perturbations or meteorological interference in satellite communication networks. The delay buffer unit dynamically determines the reasonable dwell time under the current network operating conditions based on the aforementioned quantization model, ensuring that the same application layer data payloads transmitted via satellite and scattering media arrive relatively synchronously within the receiving window at the receiving end.

[0089] When implementing the delayed dwell strategy, after receiving a bypass clone message, the delay buffer unit parses the same global identifier carried in the tunnel header of the bypass clone message, uses the same global identifier as a cache index, and writes the bypass clone message into an independent cache queue. The delay buffer unit records the enqueue timestamp of the bypass clone message entering the independent cache queue, and adds the enqueue timestamp to the dwell time of the bypass clone message calculated according to the staggered dwell time formula to determine the release time of the bypass clone message.

[0090] When fluctuations in the underlying network state cause the calculated dwell time of the bypass clone message to be negative, the delay buffer unit sets the dwell time of the bypass clone message to zero and directly includes the bypass clone message in the release sequence. When the calculated dwell time of the bypass clone message is non-negative, the delay buffer unit maintains the bypass clone message in a dwell-locked state according to the corresponding release time. Thus, the delay buffer unit can perform feasible delayed dwell control based on determined time parameters, cache index, and release time, rather than just providing an abstract functional description.

[0091] S403, based on the calculated dwell time of the bypass clone message, the system performs physical release scheduling for dual-path asynchronous transmission. The delay buffer unit establishes a cache record for each bypass clone message entering the independent cache queue. The cache record includes at least the same global identifier, enqueue timestamp, bypass clone message dwell time, release time, and dwell lock status flag.

[0092] The release time is determined by the enqueue timestamp and the dwell time of the bypass clone message. The delay buffer unit uses any one of the following timing methods: backoff timer, timed scan thread, or system timer interrupt, to determine the expiration of cached records in the independent cache queue. When the current timing time has not yet reached the corresponding release time, the delay buffer unit keeps the bypass clone message in a dwell-locked state; when the current timing time reaches or exceeds the corresponding release time, the delay buffer unit sets the dwell-locked state flag of the bypass clone message to the released state and moves the bypass clone message into the waiting-to-send queue.

[0093] When multiple bypass clone messages simultaneously meet the release conditions, the delay buffer unit forms a time-offset transmission sequence according to the order of their release times. If multiple bypass clone messages have the same release time, they are released in a first-in-first-out (FIFO) manner according to their enqueue timestamps. The time-offset transmission sequence is sent to the tropospheric scattering device via the underlying physical network interface card (NIC).

[0094] Meanwhile, the main transmission packets do not enter an independent buffer queue, but are instead delivered directly to the satellite modem via the underlying direct transmission queue for direct transmission. Thus, the main transmission packets, transmitted without delay, and the bypass cloned packets, which undergo buffering and expiration, form a spatially and temporally isolated dual-path asynchronous data stream at the communication architecture level. This spatial-temporal isolation design reduces the probability of concurrent packet loss caused by co-channel interference or momentary single-path interruptions in the spatial environment at the architectural level.

[0095] like Figure 7 As shown, step S5 of tunnel space deduplication and state machine recovery in the satellite scattering fusion communication system of this invention specifically includes the following sub-steps: S501, in this embodiment, the deduplication processing unit deployed at the peer communication node receives dual asynchronous data streams transmitted via heterogeneous physical media. In a multi-path heterogeneous network transmission environment, due to the latency differences of the underlying physical media, data packets arrive out of order in the time dimension. If a single monotonically increasing comparison mechanism is directly used, the system will mistakenly discard legitimate packets that arrive late due to latency differences. Based on the above objective engineering factors, in order to remove redundant traffic and ensure the data uniqueness of upper-layer services, the deduplication processing unit performs a fixed-length sliding window discarding action based on the internally configured memory structure.

[0096] As a preferred method, the deduplication unit extracts the globally independent sequence number encapsulated within the received message and compares it with the boundary value of the currently maintained fixed-length sliding window. If the globally independent sequence number falls within the valid range of the window and the corresponding flag is empty, the deduplication unit receives the message, marks the flag as received, and propagates the original service payload upwards. Simultaneously, to prevent window stagnation, if the globally independent sequence number corresponding to the message is exactly equal to the left boundary value of the fixed-length sliding window, the deduplication unit slides the entire fixed-length sliding window to the right until the left boundary of the fixed-length sliding window aligns with the position of the first sequence number currently in an unreceived state. If the globally independent sequence number has already been marked as received, or its value is less than the left boundary of the fixed-length sliding window, indicating an out-of-bounds state, the deduplication unit determines that the message is a redundant copy or a late message and directly performs a discard operation. Through the above sequence number comparison and message stripping actions, the system achieves transparent deduplication between heterogeneous links.

[0097] In S502, for continuously operating communication systems, the globally independent sequence number cycles and flips after reaching its maximum field width. Simultaneously, communication nodes may experience state restarts due to environmental interference or power fluctuations. To address window state stagnation or synchronization anomalies caused by these conditions, the deduplication unit introduces a sequence number lifecycle management and abnormal restart synchronization mechanism at the scheduling level.

[0098] Specifically, the deduplication unit continuously monitors the increment step size of the globally independent sequence number. When a non-continuous jump in the globally independent sequence number of adjacent arriving packets is detected and the difference in the jump value exceeds a preset sequence number reset threshold, the system determines that a sequence number flip has occurred or the peer device has restarted. In this embodiment, the preset sequence number reset threshold is determined based on the field width of the globally independent sequence number, and its value is set to half of the maximum numerical space corresponding to the field width, so as to clearly distinguish between normal network out-of-order jumps and device-level state restarts in the numerical dimension. For this abnormal state, the deduplication unit clears the current window receive flag and reinitializes the left and right boundary positions of the fixed-length sliding window based on the newly arrived globally independent sequence number. For the basic code logic of sliding window state machine maintenance and sequence number flip processing, those skilled in the art can refer to the standard sliding window specification of the transmission control protocol, which is a well-known technology in the field and will not be described in detail here.

[0099] Based on the aforementioned spatial deduplication mechanism, the S503 satellite communication scattering convergence system achieves decoupling and coordination between the underlying deduplication protection and the upper-layer dynamic routing protocol convergence at the network architecture design level.

[0100] Typically, latency fluctuations or outages in the underlying physical links propagate upwards through the protocol stack, triggering neighbor relationship resets and network routing table convergence in the network layer's dynamic routing protocols. To break this propagation chain, a deduplication unit is deployed at the bottom egress of the three-layer tunnel. Before data packets enter the upper-layer routing protocol processing module, the deduplication unit filters and restores the dual asynchronous data streams into a single time-series data stream. This decoupling design prevents the switching of heterogeneous media and latency jitter at the lower layers from triggering the state of the upper-layer routing protocols, avoiding unnecessary route recalculation overhead and ensuring the continuity of the entire network topology.

[0101] S504, as external physical conditions recover, the system faces a technical choice regarding how to exit the redundant concurrency mechanism in a timely manner. In order to allocate resources reasonably and avoid occupying the spare transmission bandwidth of the tropospheric scattering equipment for a long time, the control plane module executes the state machine exit closed-loop and single-path recovery operations according to the multiple cross-validation rules.

[0102] During dual-mode operation, the control plane module continuously extracts primary link status data and performs a comprehensive condition assessment based on the state of the window-off hold timer activated in the preceding anti-oscillation mechanism. When the physical environment parameters are determined to meet the exit conditions, i.e., the assessment yields the... When the expected dynamic value of primary link degradation at each discrete time step remains below the state machine exit safety threshold throughout the complete operating cycle of the window-off hold timer, and the primary medium connectivity flag fed back by the underlying Ethernet interface is in normal connection, the control plane module determines that the system meets the safety exit criteria.

[0103] After triggering the exit logic, the control plane module sends a control command to the data scheduling module to stop the data packet copying operation at the service cloning unit level, thereby shutting down the dual-transmission mode and restoring the single-link operation state. At this point, the system has completed a full adaptive control closed-loop control process, from link degradation assessment, redundancy cloning, asynchronous scheduling, reception deduplication to smooth exit. For unnecessary mathematical logic derivations and general state judgments in this control closed loop, explicit conditional combinations are used to replace explanations to ensure the clarity and readability of the implementation plan.

[0104] To further verify the practical engineering feasibility and technical effectiveness of the satellite-to-scatter fusion communication system provided by this invention, the following provides an embodiment, comparative experimental data, and mechanism analysis in conjunction with a communication test scenario.

[0105] A communication hardware test platform was constructed, comprising a converged control gateway, a Ku-band satellite modem, and a C-band tropospheric scattering device. A network traffic tester was connected to the LAN side of the converged control gateway, continuously injecting mixed application layer data traffic with a bandwidth of 50Mbps. This mixed application layer data traffic included 15Mbps of video streams and command packets marked as high-priority service flows, with the remainder being ordinary background data packets. The satellite communication link in the test platform was the primary medium, while the link formed by the tropospheric scattering device served as the backup medium.

[0106] The test process included two comparative scenarios. Comparative Scenario 1 employed a hard switchover mechanism between primary and backup links, initiating takeover via the scattering link after the satellite link was disconnected. Comparative Scenario 2 employed a multi-transmission mechanism without delay compensation or capacity limitations, replicating and transmitting all services in parallel across both links when the primary link's performance declined. The implementation example utilized the control mechanism provided by this invention.

[0107] The test was initiated by artificially injecting simulated rain attenuation signals into the satellite link channel, causing the automatic gain control level of the primary link receiver to decrease between the 30th and 60th seconds of the test. The control plane module in the embodiment calculated that the expected dynamic value of link degradation exceeded the preset state machine trigger high-risk threshold, activating the dual-transmission mode. During heterogeneous delay difference compensation scheduling, the system measured an average transmission delay of 642.5 ms for the satellite link, an average transmission delay of 38.1 ms for the scattering link, a standard deviation of 16.4 ms for the satellite link delay, and an extreme jitter compensation factor of 1.8.

[0108] The example uses a staggered dwell time calculation formula to determine the dwell time of bypass cloned messages. The staggered dwell time calculation formula is as follows: ; Substituting the above actual test values ​​into the formula for calculation, the calculation process is as follows: Bypass clone message dwell time Based on this, the delay buffer unit performs a resident wait-to-release action on the bypass cloned message.

[0109] The packet loss rate of high-priority service flows and the percentage of receiver reordering packets exceeding the threshold were extracted from the period from 20s to 100s and summarized in Table 1. The percentage of receiver reordering packets exceeding the threshold represents the proportion of packets whose reordering waiting time exceeds the preset threshold due to out-of-order arrival within the statistical period.

[0110] Table 1. Statistics of Network Performance Tests During Link Degradation and Recovery Periods Note: "-" indicates that the receiver's out-of-order reordering mechanism was not triggered in the corresponding test scenario or that there was no valid statistical data due to a physical link disconnection.

[0111] Figure 8 In the table, solid black lines represent examples, dark gray dashed lines represent comparative example 1, and light gray dotted lines represent comparative example 2. According to Table 1 and... Figure 8 The data from the implementation example shows that, throughout the entire cycle of simulated rain attenuation and link degradation, the packet loss rate of high-priority services remained below 0.11%, and the proportion of reordered over-threshold packets remained stable, maintaining the continuity of network transmission. Observing the performance of the comparison group, Comparative Example 1, which relied on physical layer disconnection signals to trigger route recalculation, experienced a packet loss peak of 16.73% at 50 seconds, reflecting that the primary / backup hard switchover mechanism was unable to cope with the gradual deterioration of the link. Combining the test results of Comparative Example 2, although the simple parallel dual-transmission strategy did not experience large-scale disconnection due to network convergence, the physical delay difference of the underlying heterogeneous media caused packet out-of-order delivery at the receiving end, resulting in queuing and backlog in the receiving buffer; its packet loss rate rose to 5.61% in the middle of the degradation, and the proportion of reordered over-threshold packets reached 31.04%.

[0112] At the underlying control mechanism level, the implementation relies on the deterioration rate calculation of the underlying register parameters to activate the anti-oscillation state machine before service interruption occurs at the physical layer, providing advance time for system fault tolerance processing. For high-priority service flows, the system implements a restricted cloning mechanism in conjunction with the instantaneous net load boundary of the backup link to control input-side congestion of backup communication equipment under space environment interference. On this basis, heterogeneous delay difference compensation scheduling dynamically calculates the dwell time and performs message delay release in memory to offset the spatial delay difference caused by the long-distance propagation path of the satellite link, enabling the original primary message and the bypass cloned message to be orderly separated within a fixed-length sliding window at the other end. This cross-layer collaborative mechanism blocks the interference of underlying physical fluctuations on upper-layer data routing, ensuring the service continuity of the overall communication system in complex environments.

[0113] The above are merely specific embodiments of this application, but the scope of protection of this application is not limited thereto. Any person skilled in the art can easily conceive of various equivalent modifications or substitutions within the technical scope disclosed in this application, and such modifications or substitutions should all be covered within the scope of protection of this application. Therefore, the scope of protection of this application should be determined by the scope of the claims.

Claims

1. A satellite-to-scatter fusion communication system, characterized in that, include: The converged control gateway is used to internally build a three-layer tunnel to shield the differences in the underlying heterogeneous network media, and to issue status and obtain instructions through an out-of-band management interface. The satellite modem is connected to the fusion control gateway and is used to receive the status acquisition command and feed back the primary link status data. The tropospheric scattering device is connected in parallel to the fusion control gateway to synchronously receive the status acquisition command and feed back the status data of the backup link. The fusion control gateway includes: The link status monitoring module is used to collect the primary link status data and the backup link status data, and calculate and generate a cross-layer link comprehensive matrix. The control plane module is used to extract the cross-layer link synthesis matrix, perform threshold comparison analysis, and output state machine scheduling instructions; The data scheduling module is used to process the service data entering the tunnel according to the state machine scheduling instructions.

2. The satellite-to-scatter fusion communication system according to claim 1, characterized in that, The data scheduling module includes: The traffic classification unit is used to classify the raw downlink traffic to generate high-quality traffic flows. The service cloning unit is used to perform copying and encapsulation on the high-priority service flow to generate main path transmission messages and bypass clone messages. The delay buffer unit is used to determine the dwell time of the bypass clone message according to the staggered dwell time calculation formula, and write the bypass clone message into the buffer queue with the same global identifier as the index. The bypass clone message is then allowed to dwell and wait for the bypass clone message and release it upon expiration according to the dwell time, so as to generate a time offset transmission sequence. The deduplication processing unit is used to perform deduplication processing on the arriving messages at the data receiving end.

3. The satellite-to-scatter fusion communication system according to claim 2, characterized in that, The traffic classification unit is specifically used for: Intercept the original downlink service execution differential service code point matching as the service data, and identify the packets carrying key instructions, voice or preset priority video according to the pre-configured service weight strategy; The message carrying key instructions, voice, or preset priority video is separated from the original downlink service to generate a high-priority service flow.

4. The satellite-to-scatter fusion communication system according to claim 3, characterized in that, The service cloning unit is specifically used for: Upon receiving the high-priority service flow, perform data packet replication and tunnel header sequence number encapsulation to generate the same global identifier; The same global identifier is used to encapsulate and output the main path transmission message and the bypass clone message respectively.

5. A satellite-to-scatter fusion communication system according to claim 4, characterized in that, The delay buffer unit is specifically used for: Receive bypass clone messages carrying the same global identifier, obtain the average transmission delay of the satellite link, the average transmission delay of the scatter link, and the standard deviation of the satellite link delay, and calculate the dwell time of the bypass clone message according to the staggered dwell time calculation formula. Write the bypass cloned message into an independent buffer queue, record the enqueue timestamp of the bypass cloned message, and determine the release time based on the enqueue timestamp and the dwell time of the bypass cloned message. When the current timeout reaches the release time, the resident lock state of the bypass clone message is released, and a time offset transmission sequence is generated according to the release time; The time-off transmission sequence is sent to the tropospheric scattering device by the physical network card, and the main transmission message is sent to the satellite modem for direct transmission.

6. The satellite-to-scatter fusion communication system according to claim 5, characterized in that, The deduplication processing unit is specifically used for: When the fusion control gateway acts as a data receiver, it parses the same global identifier of the main path transmission message and the bypass clone message; A fixed-length sliding window is established based on the same global identifier; For multiple arriving messages corresponding to the same global identifier, the fixed-length sliding window is used to retain the first valid message that arrives first and discard the subsequent duplicate messages.

7. The satellite-to-scatter fusion communication system according to claim 1, characterized in that, The primary link status data includes automatic gain control level and carrier-to-noise ratio; The backup link status data includes symbol rate, modulation order, forward error correction code rate, link connection status flag, and received signal level.

8. The satellite-to-scatter fusion communication system according to claim 1, characterized in that, The link status monitoring module is specifically used for: The primary link status data and the backup link status data extracted from the underlying register parameters of the satellite modem and the tropospheric scattering device, as well as the equivalent one-way transmission delay of the satellite link, the equivalent one-way transmission delay of the scattering link, and the delay standard deviation of the two links are integrated and encapsulated to calculate and generate the cross-layer link comprehensive matrix. Based on the cross-layer link synthesis matrix, the primary link status data is extracted and substituted into the link degradation expectation evaluation formula based on multi-parameter dynamic weighting for calculation, so as to obtain the dynamic value of primary link degradation expectation at discrete time step.

9. A satellite-to-scatter fusion communication system according to claim 8, characterized in that, The control plane module is specifically used for: Obtain the expected dynamic value of the primary link degradation, and compare the expected dynamic value of the primary link degradation with the preset state machine trigger high-risk threshold; When the expected dynamic value of the primary link degradation is determined to be greater than or equal to the state machine triggering high-risk threshold, the dual-transmission mode is activated. Based on the dual-mode encapsulation, the state machine scheduling instruction containing the logic to activate the dual-mode is generated and output.

10. A satellite-to-scatter fusion communication system according to claim 9, characterized in that, The control plane module is also used for: Extract various modulation parameters from the backup link status data, substitute them into the backup link available net load dynamic estimation formula for calculation, and obtain the instantaneous net load boundary of the backup link. The instantaneous net load boundary of the backup link is encapsulated as a capacity constraint parameter into the state machine scheduling instruction and output to the data scheduling module.

Images