A space-ground integrated network transmission control method based on multi-dimensional congestion signals

Abstract

The application discloses a kind of transmission control methods of space-ground integrated network based on multidimensional congestion signal, belong to network communication technical field.In the data transmission process, sampling and statistical estimation bandwidth signal and adjacent acknowledgement message interval and adjacent acknowledgement message confirmation sequence number difference are used as multidimensional congestion signal;According to the comprehensive judgment result of multidimensional congestion signal, the current network congestion state is divided into congestion state, topology change disturbance state and non-congestion state;According to the congestion state distinguished, the data sending rate is dynamically adjusted, and the maximum estimated bandwidth parameter is prohibited to update when judging as topology change disturbance state.The application can accurately judge the real congestion state in space-ground integrated network, avoid the sending rate misjudgment caused by the round trip time change due to topology dynamic, effectively prevent network congestion, and improve the efficiency and stability of data transmission.

Description

Technical Field

[0001] This invention belongs to the field of network communication technology, and in particular relates to a method for integrated space-ground network transmission control based on multidimensional congestion signals. Background Technology

[0002] The space-ground integrated network relies on traditional terrestrial networks and extends them with satellite internet, combining terrestrial and satellite networks to form a unified network system. In recent years, global research on space-ground integrated networks has entered a stage of rapid development. However, in space-ground integrated networks, the high-speed movement of satellites causes dynamic changes in network topology, which will result in frequent network link interruptions and recovery, leading to changes in round-trip latency. This will affect end-to-end data transmission performance.

[0003] In the Linux kernel TCP / IP protocol stack, the sending end can obtain various end-to-end connection statistics. For example, based on acknowledgment messages, the sending end can obtain the timestamp of receiving the acknowledgment message, the acknowledgment sequence number, the amount of packet loss during network transmission, the amount of successfully delivered data, and the round-trip time. Bandwidth-based transmission control (BBR) is a widely used transmission control method. It predicts the size of the network bottleneck bandwidth by continuously sampling the data delivery rate and then adjusts the sending rate accordingly. In integrated space-ground networks, topology dynamics can cause changes in the end-to-end data transmission path, resulting in round-trip time fluctuations, which in turn cause sudden increases in the delivery rate, ultimately leading to a sudden increase in the estimated bandwidth calculated by the BBR. Therefore, in integrated space-ground networks, round-trip time fluctuations and sudden increases in estimated bandwidth are not entirely caused by changes in network congestion; some are due to the dynamics of the space-to-ground and inter-satellite topologies. Therefore, using estimated bandwidth alone as a congestion signal in integrated space-ground networks will cause the BBR to misjudge the network congestion state, resulting in the sending end sending data at a rate exceeding the actual network bottleneck link bandwidth, ultimately causing network congestion and severely impacting data transmission performance.

[0004] To improve data transmission performance in integrated space-ground network scenarios, this invention discloses a transmission control method for integrated space-ground networks based on multi-dimensional congestion signals. This method combines multiple congestion signals to accurately determine the congestion state of the integrated space-ground network from multiple dimensions. It categorizes network congestion states into three types: congested state, topology change disturbance state, and non-congested state, and dynamically adjusts the data transmission rate according to the network congestion state. This avoids the impact of round-trip delay changes caused by topology dynamics on data transmission performance, thereby improving data transmission efficiency and stability. Summary of the Invention

[0005] To address the aforementioned technical problems, this invention provides a method for integrated space-ground network transmission control based on multi-dimensional congestion signals, comprising the following steps: During data transmission in the integrated space-ground network, the sending end samples and counts at least two different types of congestion signals from the received acknowledgment messages based on the characteristics of dynamic changes in network topology. The congestion signals include the estimated bandwidth signal, the interval between adjacent acknowledgment messages, and the difference in acknowledgment sequence numbers between adjacent acknowledgment messages. Based on the comprehensive judgment result of the estimated bandwidth signal, the adjacent acknowledgment message interval, and the difference in the acknowledgment sequence number of the adjacent acknowledgment message, the current network congestion state is divided into congestion state, topology change disturbance state, and non-congestion state. Based on the current network congestion status, the data transmission rate of the sending end is dynamically adjusted. When it is determined to be a topology change disturbance, the sending end is prohibited from updating the maximum estimated bandwidth parameter it maintains.

[0006] Optionally, during data transmission, the sending end continuously receives acknowledgment messages, calculates the interval between adjacent acknowledgment messages based on the timestamps of two adjacent acknowledgment messages, and calculates the difference in acknowledgment sequence numbers between adjacent acknowledgment messages based on the acknowledgment sequence numbers carried by the two adjacent acknowledgment messages.

[0007] Optionally, the adjacent acknowledgment message interval and the difference in acknowledgment sequence number of the adjacent acknowledgment message are compared with their respective preset thresholds, and the comparison results are used to determine whether the current network is under topology change disturbance.

[0008] Optionally, when the interval between adjacent acknowledgment messages is greater than a first threshold and less than a second threshold, it is determined that the current network is in a topology change disturbance state.

[0009] Optionally, when the difference in the confirmation sequence number of adjacent confirmation messages is greater than or equal to a preset third threshold, it is determined that the current network is in a topology change disturbance state.

[0010] Optionally, when the interval between adjacent acknowledgment messages does not trigger the topology change disturbance dynamic judgment condition and the difference in the acknowledgment sequence number of adjacent acknowledgment messages does not trigger the topology change disturbance dynamic judgment condition, the current network is determined to be in a congested state or a non-congested state based on the comparison result between the estimated bandwidth signal and the preset bandwidth threshold.

[0011] Optionally, based on the result of determining whether the current network is in a topology change disturbance state, it is determined whether to allow updating the maximum estimated bandwidth parameter maintained by the sending end; when it is determined that the current network is in a topology change disturbance state, updating the maximum estimated bandwidth parameter is prohibited.

[0012] Optionally, the transmission rate gain coefficient of the transmitting end changes at a preset period; after determining that the current network is under topology change disturbance, the state of prohibiting the updating of the maximum estimated bandwidth parameter is maintained for a preset duration, and the duration is greater than the preset period. After the duration ends, the updating of the maximum estimated bandwidth parameter is resumed.

[0013] On the other hand, the present invention also provides an electronic device including a memory, a processor, and a computing program stored in the memory and executable on the processor, wherein the processor implements the method when executing the computing program.

[0014] On the other hand, the present invention also provides a computer-readable storage medium storing a computer program that, when executed by a processor, implements the method.

[0015] Compared with the prior art, the present invention has the following advantages and technical effects: (1) Improve the accuracy of network congestion status judgment and solve the problem that the existing technology uses a single congestion signal when the topology of the integrated space-ground network changes dynamically, which leads to the sending end misjudging the congestion status.

[0016] (2) To address the phenomenon of a sudden increase in estimated bandwidth caused by link failure and failure link recovery, optimize the transmission rate control strategy to prevent the sending end from sending data at a transmission rate exceeding the bottleneck link bandwidth.

[0017] (3) The proposed method is lightweight and has low deployment cost. This invention only optimizes the transmission control logic of the sending end, which can be implemented based on existing acknowledgment message parameters, making it easy to deploy and maintain. Attached Figure Description

[0018] The accompanying drawings, which form part of this application, are used to provide a further understanding of this application. The illustrative embodiments and descriptions of this application are used to explain this application and do not constitute an undue limitation of this application. In the drawings: Figure 1 This is a flowchart of the integrated space-ground network transmission control method based on multi-dimensional congestion signals according to an embodiment of the present invention; Figure 2 This is a schematic diagram illustrating the changes in round-trip time caused by inter-satellite link failure and failure link recovery according to an embodiment of the present invention; Figure 3 This is an experimental result diagram of the embodiment of the present invention, which verifies the effectiveness of the proposed method through a single transmission to a long-current experiment; Figure 4 The figure shows the experimental results of the proposed method's performance in various transmission experiments on different stream sizes, according to an embodiment of the present invention. Detailed Implementation

[0019] It should be noted that, unless otherwise specified, the embodiments and features described in this application can be combined with each other. This application will now be described in detail with reference to the accompanying drawings and embodiments.

[0020] It should be noted that the steps shown in the flowchart in the accompanying drawings can be executed in a computer system such as a set of computer-executable instructions, and although a logical order is shown in the flowchart, in some cases the steps shown or described may be executed in a different order than that shown here.

[0021] This invention discloses a transmission control method for an integrated space-ground network based on multi-dimensional congestion signals. Transmission control refers to the dynamic adjustment of the transmission rate by the transmitting end based on network feedback information, thereby making full use of network bandwidth resources while avoiding network congestion. Bandwidth-based transmission control (BBR) is a widely used transmission control method. It predicts the size of the network bottleneck bandwidth by continuously sampling the delivery rate to update the maximum estimated bandwidth. BBR continuously samples the data delivery rate and records it as the estimated bandwidth, storing it in a sliding window. Then, it uses the largest estimated bandwidth sample in the sliding window as the maximum estimated bandwidth to update the data transmission rate. In an integrated space-ground network, round-trip delay fluctuations and sudden increases in estimated bandwidth are not entirely caused by changes in network congestion status; some are caused by dynamic changes in the satellite-to-ground and inter-satellite topologies. Therefore, using the estimated bandwidth alone as a congestion signal in an integrated space-ground network will lead to BBR misjudging the network congestion status, causing the transmitting end to send data at a transmission rate exceeding the actual network bottleneck link bandwidth, ultimately triggering network congestion. How to design a transmission control method adaptable to integrated space-ground networks is the technical problem this invention aims to solve.

[0022] Example 1 In this invention, the network topology consists of user terminal nodes, ground station nodes, and satellite nodes. User terminals communicate directly with ground stations, ground stations exchange data with connected satellites, and satellite nodes handle routing and forwarding functions. User terminal nodes collect network connection information and adjust transmission rates. During the experiment, the inter-satellite link bandwidth was 1000 Mbps, and the satellite-to-ground link bandwidth was 200 Mbps; therefore, the network bottleneck link bandwidth was 200 Mbps. Inter-satellite link failures and failure recovery are triggered by disconnecting a specified path from the transmitter to the receiver and restoring it after a certain period.

[0023] like Figure 1 As shown, this embodiment provides a method for integrated space-ground network transmission control based on multi-dimensional congestion signals, including: During data transmission in the integrated space-ground network, the sending end samples and counts at least two different types of congestion signals from the received acknowledgment messages based on the characteristics of dynamic changes in network topology. The congestion signals include the estimated bandwidth signal, the interval between adjacent acknowledgment messages, and the difference in acknowledgment sequence numbers between adjacent acknowledgment messages. Based on the comprehensive judgment result of the estimated bandwidth signal, the adjacent acknowledgment message interval, and the difference in the acknowledgment sequence number of the adjacent acknowledgment message, the current network congestion state is divided into congestion state, topology change disturbance state, and non-congestion state. Based on the current network congestion status, the data transmission rate of the sending end is dynamically adjusted. When it is determined to be a topology change disturbance, the sending end is prohibited from updating the maximum estimated bandwidth parameter it maintains.

[0024] Specifically, the following steps are included: Step S1: Based on the feature sampling and statistical analysis of the dynamic changes in the topology of the integrated space-ground network, multi-dimensional congestion signals are collected and analyzed.

[0025] In an integrated space-ground network, topology dynamics can cause changes in the end-to-end data transmission path, resulting in fluctuations in round-trip delay and a sudden increase in estimated bandwidth. This can cause the BBR to misjudge the network congestion state and send data at a rate exceeding the actual network bottleneck link bandwidth, thus triggering network congestion.

[0026] To address the issue of BBR misjudging network congestion due to round-trip delay variations caused by the dynamic topology of integrated space-ground networks, this invention introduces additional congestion signals to accurately determine network congestion status in addition to the estimated bandwidth. During data transmission, the sending end continuously receives acknowledgment messages and calculates the interval between adjacent acknowledgment messages based on the timestamps of two adjacent acknowledgment messages, and calculates the difference in acknowledgment sequence numbers between adjacent acknowledgment messages based on the acknowledgment sequence numbers carried by the two adjacent acknowledgment messages. Specifically, to avoid the impact of delay variations caused by topology dynamics on the BBR, this invention introduces two additional statistical pieces of information—the interval between adjacent acknowledgment messages and the difference in acknowledgment sequence numbers between adjacent acknowledgment messages—to distinguish from estimated bandwidth changes caused by topology dynamics and estimated bandwidth changes caused by actual network congestion status changes from multiple dimensions, thereby preventing the sending end from sending data at a rate exceeding the actual network bottleneck bandwidth.

[0027] This invention uses the adjacent acknowledgment message interval to detect the increased round-trip time (RTT) caused by topology dynamics. The adjacent acknowledgment message interval refers to the time interval between two adjacent acknowledgment messages received by the sender. When the end-to-end path is stable, adjacent acknowledgment messages arrive at the sender at a relatively stable interval. However, when topology dynamic events lengthen the end-to-end path, the RTT increases, and acknowledgment messages take longer to reach the sender, leading to a sharp increase in the adjacent acknowledgment message interval. Figure 2As shown in (a), when the inter-satellite link between satellite Sa and satellite Sb fails, the end-to-end data transmission path changes, increasing the round-trip delay. The acknowledgment message carrying acknowledgment sequence number 101 takes longer to return to the sender, leading to a larger interval between adjacent acknowledgment messages as calculated by the sender. The sender calculates the interval between adjacent acknowledgment messages after each received acknowledgment message. ; parameter It is the interval between adjacent acknowledgment messages, parameter This parameter represents the time when the sender last received an acknowledgment message. This represents the time when the sender received the previous acknowledgment message.

[0028] This invention uses the difference in acknowledgment sequence numbers between adjacent acknowledgment messages to detect the reduction in round-trip time (RTT) caused by topology dynamics. The difference in acknowledgment sequence numbers between adjacent acknowledgment messages refers to the difference in the acknowledgment sequence numbers carried by two adjacent acknowledgment messages. When the end-to-end path remains unchanged, the difference in acknowledgment sequence numbers between adjacent acknowledgment messages is relatively fixed. However, when topology dynamic events shorten the end-to-end path, the RTT decreases, and some acknowledgment messages arrive at the sending end earlier, causing a sharp increase in the difference in acknowledgment sequence numbers between adjacent acknowledgment messages. For example... Figure 2 As shown in (b), after the faulty inter-satellite link between satellite Sa and satellite Sb is restored, the end-to-end data transmission path changes, the round-trip delay decreases, and acknowledgment messages carrying acknowledgment sequence number 105 and later arrive at the sender earlier than the acknowledgment message carrying acknowledgment sequence number 100. This will cause a sudden increase in the difference in acknowledgment sequence numbers between adjacent acknowledgment messages counted by the sender. The sender counts the difference in acknowledgment sequence numbers between adjacent acknowledgment messages after each acknowledgment message received: ; Where parameters This represents the difference in acknowledgment sequence numbers between adjacent acknowledgment messages. This represents the acknowledgment sequence number carried in the most recent acknowledgment message received by the sender. (Parameter) This indicates the acknowledgment sequence number carried in the previous acknowledgment message received by the sender.

[0029] Step S2: Determine the current network congestion status based on multidimensional congestion signals.

[0030] This invention introduces two additional statistical pieces of information—the interval between adjacent acknowledgment messages and the difference in acknowledgment sequence numbers between adjacent acknowledgment messages—as congestion signals in addition to the estimated bandwidth. This allows for a multi-dimensional assessment of network congestion status, categorizing it into three states: congested, topology-change-disruptive, and non-congested. The proposed method detects the interval between adjacent acknowledgment messages and the difference in acknowledgment sequence numbers in real time based on acknowledgment message feedback information. If the interval exceeds a threshold, a link failure is considered to have occurred, resulting in increased round-trip time and a sudden increase in estimated bandwidth. However, this sudden increase in estimated bandwidth does not necessarily indicate that the network has more available bandwidth resources, and in this case, the method classifies the network as being in a topology-change-disruptive state. Conversely, if the difference in acknowledgment sequence numbers between adjacent acknowledgment messages exceeds a threshold, the faulted link is considered to have recovered, resulting in decreased round-trip time and a sudden increase in estimated bandwidth. Again, this sudden increase in estimated bandwidth does not necessarily indicate that the network has more available bandwidth resources, and in this case, the method classifies the network as being in a topology-change-disruptive state.

[0031] Based on the interval between adjacent acknowledgment messages and the difference in acknowledgment sequence numbers between adjacent acknowledgment messages, each is compared with its corresponding preset threshold, and the comparison result is used to determine whether the current network is under topology change disturbance. The setting of the corresponding thresholds when determining network congestion status based on the threshold detection method is described in detail below.

[0032] This invention uses the adjacent acknowledgment message interval to detect the increased round-trip delay caused by topology dynamics and to determine network congestion status. In the Linux kernel TCP / IP protocol stack, the data receiver sends an acknowledgment message to the sender for every two data packets received. Therefore, when the network link state is stable, the interval between acknowledgment messages received by the sender is also relatively stable and related to the network bottleneck bandwidth. The adjacent acknowledgment message interval when the link state is stable can be calculated by the following formula: ; in The parameter MSS represents the interval between adjacent acknowledgment messages; it indicates the maximum length of the application layer data portion in a single TCP segment, which is negotiated and determined when the two communicating parties establish a TCP connection. The meaning is the available network bandwidth. Therefore, when the network link is stable, the interval at which the sending end receives acknowledgment messages is... When topology dynamic events increase round-trip time, acknowledgment messages take longer to reach the sender, and the interval between the sender receiving acknowledgment messages also increases. Therefore, this invention will... The first threshold is used to detect the increased round-trip delay caused by topological dynamics, denoted as . In the experimental environment of this invention The minimum RTT is 100 milliseconds. The BBR probes the minimum RTT at 10-second intervals. During the probe, the BBR sends only 4 MSS of data and then stops sending to empty the network queue and obtain the minimum RTT. The probe lasts for 200 milliseconds. During the 200 milliseconds of the BBR probing the minimum RTT, because the sender stops sending after sending 4 MSS of data, no further acknowledgment messages will be received until the probe ends. That is, if the interval between adjacent acknowledgment messages received by the sender exceeds 200 milliseconds, this phenomenon is caused by the BBR probing the minimum RTT and does not indicate a network link failure. Therefore, this invention uses the second threshold... Set to 200 milliseconds. The sending end continuously checks whether the interval between adjacent acknowledgment messages meets the condition. ; If the interval between adjacent acknowledgment messages If the conditions are met, it is considered that a link failure has occurred on the end-to-end network transmission path. The increase in round-trip delay and the sudden increase in estimated bandwidth obtained from the sending end are both caused by the link failure. At this time, this method determines that the network is in a topology change disturbance state.

[0033] This invention uses the difference in acknowledgment sequence numbers between adjacent acknowledgment messages to detect the reduction in round-trip delay caused by topology dynamics and to determine network congestion status. The acknowledgment sequence number refers to the sequence number of the next byte the receiver expects to receive; that is, data before the acknowledgment sequence number has been successfully received. The difference in acknowledgment sequence numbers between adjacent acknowledgment messages can be calculated using the following formula: ; Where parameters Indicates the difference in acknowledgment sequence numbers between adjacent acknowledgment messages; Parameter This represents the acknowledgment sequence number carried in the most recent acknowledgment message received by the sender; Parameter The meaning is the acknowledgment sequence number carried in the previous acknowledgment message received by the sender. In the Linux kernel TCP / IP protocol stack, the receiving end returns an acknowledgment message to the sender every 2MSS of data received. Therefore, when the network link state is stable, the difference in acknowledgment sequence numbers between adjacent acknowledgment messages counted by the sender is also relatively stable, usually 2MSS. However, when topology dynamic events reduce round-trip time, some acknowledgment messages arrive at the sender earlier in a shorter time, causing a sudden increase in the difference in acknowledgment sequence numbers between adjacent acknowledgment messages counted by the sender. Therefore, this invention enables the data sender to continuously count the difference in acknowledgment sequence numbers between adjacent acknowledgment messages, and to enhance the robustness of network congestion state judgment, this invention sets a third threshold for the difference in acknowledgment sequence numbers between adjacent acknowledgment messages. Set to 20 MSS. The sender continuously counts the difference in acknowledgment sequence numbers between adjacent acknowledgment messages and checks whether it meets the following conditions: ; If the difference in the confirmation sequence number between adjacent confirmation messages If the conditions are met, it is considered that a fault link has been restored on the end-to-end network transmission path. The reduction in round-trip delay and the estimated bandwidth increase obtained from the sending end are both caused by the restoration of the fault link. At this time, this method determines that the network is in a topology change disturbance state.

[0034] The method proposed in this invention detects increased round-trip delay caused by topology dynamism by using the interval between adjacent acknowledgment messages based on estimated bandwidth; and detects decreased round-trip delay caused by topology dynamism by using the difference in acknowledgment sequence numbers between adjacent acknowledgment messages. This method uses a threshold detection method to comprehensively judge the network congestion status, thereby avoiding misjudgment of network congestion status by the BBR due to changes in round-trip delay caused by satellite-to-ground and inter-satellite topology dynamism.

[0035] In summary, when the interval between adjacent acknowledgment messages does not trigger the topology change disturbance judgment condition and the difference in the acknowledgment sequence number of adjacent acknowledgment messages does not trigger the topology change disturbance judgment condition, the network is determined to be in a congested or non-congested state based on the comparison result between the estimated bandwidth signal and the preset bandwidth threshold.

[0036] Step S3: Adjust the data transmission rate based on the current network congestion status.

[0037] To avoid the impact of round-trip delay variations caused by the dynamic nature of the integrated space-ground network topology on the Broadband Buffer (BBR), the method proposed in this invention dynamically adjusts the transmission rate based on the network congestion status. During operation, the BBR maintains an estimated bandwidth sliding window, storing estimated bandwidth samples acquired in the most recent 10 rounds. The BBR dynamically maintains a maximum estimated bandwidth parameter to adjust the transmission rate; the maximum estimated bandwidth parameter is set to the largest estimated bandwidth sample value within the sliding window. The BBR adjusts the transmission rate using the maximum estimated bandwidth parameter and a periodically changing transmission rate gain coefficient. When a topology dynamic event causes the estimated bandwidth sample counted by the transmitter to exceed the network bottleneck bandwidth, the BBR updates the transmission rate using the largest estimated bandwidth sample in the estimated bandwidth sliding window. This results in the BBR adjusting the transmission rate using an estimated bandwidth exceeding the network bottleneck bandwidth, causing the transmitter's transmission rate to exceed the network bottleneck bandwidth, leading to network congestion and severely impacting network transmission performance. Based on the judgment of the network congestion status, this invention further determines whether updating the maximum estimated bandwidth parameter maintained by the transmitter is allowed based on the judgment of whether the current network is in a topology change disturbance state; when the current network is judged to be in a topology change disturbance state, updating the maximum estimated bandwidth parameter is prohibited. The sending end's transmission rate gain coefficient changes at a preset period. After determining that the current network is under topology change disturbance, the state prohibiting updates to the maximum estimated bandwidth parameter is maintained for a preset duration, longer than the preset period. After the duration ends, updates to the maximum estimated bandwidth parameter resume. The method proposed in this invention uses the adjacent acknowledgment message interval and the difference in adjacent acknowledgment message sequence numbers to detect both the increased round-trip time due to link failure and the decreased round-trip time due to link recovery, avoiding misjudgment of network congestion by the BBR due to topology dynamics.

[0038] Based on the method's assessment of network congestion, when the network is in a topology-changing disturbance state, this method limits the data transmission rate by freezing the estimated bandwidth, thereby preventing the BBR from transmitting data at a rate greater than the network bottleneck bandwidth due to topology dynamic events. Specifically, if the round-trip delay change and estimated bandwidth surge are caused by a topology dynamic event in the integrated space-ground network, i.e., the network is in a topology-changing disturbance state, the transmission control method proposed in this invention will pause updating the maximum estimated bandwidth parameter. Even if a new estimated bandwidth sample larger than the current maximum estimated bandwidth parameter is added to the sliding window, the maximum estimated bandwidth parameter remains unchanged, thus achieving the purpose of freezing the estimated bandwidth and limiting the transmission rate. The estimated bandwidth freeze lasts for 10 rounds. This is because when topology dynamics cause an increase in round-trip delay, the sending end's sending window is exhausted and it stops transmitting data, and the token bucket begins to accumulate tokens. When the sending end resumes transmitting data, there are additional tokens in the token bucket that allow for data bursts. If the BBR uses a transmission rate gain coefficient of 1.25 to probe the network bottleneck bandwidth at this time, the sending end will obtain an estimated bandwidth sample that exceeds the actual network bottleneck bandwidth. The BBR transmit rate gain coefficient changes dynamically in 8-round cycles. When topology dynamic events cause increased round-trip delay and the BBR uses a 1.25 transmit rate gain coefficient to probe the bottleneck bandwidth, the estimated bandwidth sample obtained by the transmitter suddenly increases and exceeds the bottleneck bandwidth. This will lead to excessive transmit rate and network congestion. To solve this problem, it should be ensured that when a link failure causes increased round-trip delay and the BBR uses a 1.25 transmit rate gain coefficient to probe the network bottleneck bandwidth, the estimated bandwidth is frozen. That is, the duration of the estimated bandwidth freeze should be greater than the BBR transmit rate gain coefficient change cycle (8 rounds). Therefore, the transmission control mechanism proposed in this invention sets the estimated bandwidth freeze duration to 10 rounds.

[0039] On the other hand, this embodiment also provides an electronic device, including a memory, a processor, and a computing program stored in the memory and executable on the processor, wherein the processor implements the method when executing the computing program.

[0040] On the other hand, this embodiment also provides a computer-readable storage medium storing a computer program that, when executed by a processor, implements the method.

[0041] The present invention also conducted the following experiments: This embodiment implements the proposed method in the Linux kernel and conducts experiments on a container-based integrated space-ground network simulation platform to verify the effectiveness and perform performance tests on the proposed method.

[0042] Figure 3 The effectiveness of the proposed method is demonstrated through a single-transmission long-current experiment. Figure 3(a) shows that when an inter-satellite link failure causes a sudden increase in delivery rate, the method proposed in this invention detects the phenomenon and sets the OLOR (On-Path Link Outage or Recovery) flag to 1 to freeze the maximum estimated bandwidth of the BBR, thereby avoiding the maximum estimated bandwidth from exceeding the actual bottleneck link bandwidth due to abnormal delivery rate. After freezing for 10 rounds, the OLOR flag is set to 0 to unfreeze the system. Figure 3 (b) in the figure demonstrates that the method proposed in this invention effectively avoids the phenomenon that the delivery rate surge caused by inter-satellite link failure causes the transmission rate to exceed the actual bottleneck link bandwidth, thereby avoiding network congestion. Figure 3 (c) shows that when the delivery rate suddenly increases due to the recovery of the faulty link, the method proposed in this invention detects this phenomenon and sets the OLOR flag to 1 to freeze the maximum estimated bandwidth of the BBR, thereby avoiding the maximum estimated bandwidth from exceeding the actual bottleneck link bandwidth due to abnormal delivery rate. The freeze is lifted after 10 rounds. Figure 3 (d) in the figure demonstrates that the method proposed in this invention effectively avoids the phenomenon that the delivery rate surge caused by the recovery of a faulty link causes the transmission rate to exceed the actual bottleneck link bandwidth, thereby avoiding network congestion.

[0043] Figure 4 The performance of the proposed method is demonstrated through multi-transmission experiments with different stream sizes. The proposed method (BBR-Link State Change Detection, BBR-LSCD) is compared with the conventional BBR method to test the improvements brought by the proposed method. Figure 4 (a) and Figure 4 (b) shows the comparison of the average flow completion time and average slowdown of BBR-LSCD and BBR as a function of load after an inter-satellite link failure. Figure 4 (c) and Figure 4 Figure (d) shows the comparison of average flow completion time and average slowdown of BBR-LSCD and BBR as a function of load after the faulty link is restored. The experimental results show that, compared to the traditional BBR method, the proposed BBR-LSCD method exhibits better performance in integrated space-ground networks with strong topology dynamics, and the improvement in transmission performance by BBR-LSCD becomes increasingly significant with increasing load.

[0044] The above are merely preferred embodiments of this application, but the scope of protection of this application is not limited thereto. Any variations or substitutions that can be easily conceived by those skilled in the art within the scope of the technology disclosed in this application should be included 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 method for integrated space-ground network transmission control based on multi-dimensional congestion signals, characterized in that, Includes the following steps: During data transmission in the integrated space-ground network, the sending end samples and counts at least two different types of congestion signals from the received acknowledgment messages based on the characteristics of dynamic changes in network topology. The congestion signals include the estimated bandwidth signal, the interval between adjacent acknowledgment messages, and the difference in acknowledgment sequence numbers between adjacent acknowledgment messages. Based on the comprehensive judgment result of the estimated bandwidth signal, the adjacent acknowledgment message interval, and the difference in the acknowledgment sequence number of the adjacent acknowledgment message, the current network congestion state is divided into congestion state, topology change disturbance state, and non-congestion state. Based on the current network congestion status, the data transmission rate of the sending end is dynamically adjusted. When it is determined to be a topology change disturbance, the sending end is prohibited from updating the maximum estimated bandwidth parameter it maintains.

2. The method according to claim 1, characterized in that, During data transmission, the sending end continuously receives acknowledgment messages, calculates the interval between adjacent acknowledgment messages based on the timestamps of two adjacent acknowledgment messages, and calculates the difference in acknowledgment sequence numbers between adjacent acknowledgment messages based on the acknowledgment sequence numbers carried by the two adjacent acknowledgment messages.

3. The method according to claim 2, characterized in that, Based on the interval between adjacent acknowledgment messages and the difference in the acknowledgment sequence number between adjacent acknowledgment messages, each is compared with its corresponding preset threshold, and the comparison result is used to determine whether the current network is under topology change disturbance.

4. The method according to claim 3, characterized in that, When the interval between adjacent acknowledgment messages is greater than the first threshold and less than the second threshold, it is determined that the current network is in a topology change disturbance state.

5. The method according to claim 3, characterized in that, When the difference in the confirmation sequence number between adjacent confirmation messages is greater than or equal to a preset third threshold, it is determined that the current network is in a topology change disturbance state.

6. The method according to claim 3, characterized in that, When the interval between adjacent acknowledgment messages does not trigger the topology change disturbance dynamic judgment condition and the difference in the acknowledgment sequence number of the adjacent acknowledgment messages does not trigger the topology change disturbance dynamic judgment condition, the network is determined to be in a congested or non-congested state based on the comparison result between the estimated bandwidth signal and the preset bandwidth threshold.

7. The method according to claim 3, characterized in that, Based on the result of determining whether the current network is under topology change disturbance, it is determined whether to allow updating the maximum estimated bandwidth parameter maintained by the sending end; when it is determined that the current network is under topology change disturbance, updating the maximum estimated bandwidth parameter is prohibited.

8. The method according to claim 7, characterized in that, The sending rate gain coefficient changes at a preset period. After determining that the current network is under topology change disturbance, the state of prohibiting the updating of the maximum estimated bandwidth parameter will be maintained for a preset duration, and the duration is greater than the preset period. After the duration ends, the updating of the maximum estimated bandwidth parameter will be resumed.

9. An electronic device comprising a memory, a processor, and a computing program stored in the memory and executable on the processor, characterized in that, When the processor executes the computing program, it implements the method of any one of claims 1-8.

10. A computer-readable storage medium storing a computer program, characterized in that, When the computer program is executed by a processor, it implements the method of any one of claims 1-8.

Images