Train-ground wireless retransmission method and system based on network tunnel technology

Abstract

The invention discloses a train-ground wireless retransmission method based on a network tunnel technology. The train-ground wireless retransmission method comprises the following steps: capturing a network IP message which is appointed by a user and needs to be retransmitted; adding a pre-designed retransmission protocol header to the front of a user network IP message, reassembling into a UDP message, and sending the UDP message to a receiving end; and a response packet returned by the receiving end is obtained, the response packet calculates a time interval RTT from network message sending to response receiving according to the response packet, the timeout retransmission time RTO is dynamically calculated according to the RTT, and the message is retransmitted according to the timeout retransmission time RTO. The method is suitable for a single network card scheme, the problems of packet loss and overlarge balance switching delay in an 802.11 R switching protocol can be effectively solved, and the network delay is not obviously improved under the normal operation condition.

Description

Technical Field

[0001] The present invention relates to the field of communications, and in particular to a vehicle-to-ground wireless retransmission method and system based on network tunnel technology. Background Art

[0002] In order to solve the problem of packet loss during handover, the train-to-ground wireless system in subway tunnels usually adopts a dual network card solution, such as Figure 1 As shown in the figure. In this scheme, the Master Radio plays the role of communication and the Slave Radio plays the role of scanning. When the switching conditions are met, the Slave Radio first connects to the target AP2, and then the Master Radio and the Slave Radio switch roles. This ensures that the communication link is switched only after connecting to the new AP, thereby achieving a switching-free packet loss situation. However, the disadvantages of this switching scheme are: (1) Dual network cards will inevitably lead to increased hardware costs and complexity; (2) Dual network cards will inevitably require more antennas, and during the construction process, antenna installation is more complicated and difficult. Summary of the invention

[0003] The main purpose of the present invention is to provide a vehicle-to-ground wireless retransmission method and system based on network tunnel technology, which can realize the packet loss retransmission of single network card network data messages and can achieve extremely low packet loss rate or even zero packet loss rate.

[0004] The technical solution adopted by the present invention is:

[0005] A vehicle-to-ground wireless retransmission method based on network tunnel technology is provided, comprising the following steps:

[0006] Capture the network IP packets that the user specifies need to be retransmitted;

[0007] Add the pre-designed retransmission protocol header to the front of the user network IP message, reassemble it into a UDP message, and send it to the receiving end;

[0008] The response packet returned by the receiving end is obtained. The time interval RTT from sending the network message to receiving the response is calculated based on the response packet. The timeout retransmission time RTO is dynamically calculated based on the RTT, and the message is retransmitted based on the timeout retransmission time RTO.

[0009] According to the above technical solution, the retransmission protocol header includes multiple fields: packet sequence number sn, used to determine the packet order; sending packet timestamp ts; fast retransmission number hack; message type cmd; checksum, which is the retransmission protocol header checksum; response port number ap, used to resolve the port number of the receiving end to send the response message; its own receiving window size window; payload data length len.

[0010] According to the above technical solution, the response packet returned by the receiving end includes a response protocol header and a data packet, wherein the response protocol header includes a hack field, which records the minimum packet sequence number sn that the receiving end has received, indicating that all data packets smaller than the hack field are sent successfully; the data packet records the packet sequence number and receiving timestamp of the packet received by the receiving end.

[0011] According to the above technical solution, after receiving the response packet, the receiving end will mark the sequence number of the lost packet according to the response packet. When the number of markings is greater than a certain threshold, the packet will be directly retransmitted again.

[0012] According to the above technical solution, when a packet is retransmitted multiple times by the sending end and still no corresponding response packet is received, it will not be retransmitted.

[0013] Following the above technical solution, the sending end dynamically calculates the RTO through a low-pass filter and in combination with the RTT variance, specifically including the steps of: calculating the actual RTT measurement value at the current moment according to the sending packet timestamp in the retransmission protocol header and the receiving packet timestamp in the response packet; calculating the average estimated value of the RTT at the current moment; calculating the RTT measurement value at the current moment according to the RTT measurement value and the average estimated value of the RTT; calculating the average estimated value of the jitter value at the current moment; calculating the RTO at the current moment according to the average estimated value of the RTT and the average estimated value of the jitter value.

[0014] Following the above technical solution, the sending end dynamically calculates the RTO through the Kalman filter algorithm, which specifically includes the following steps: calculating the RTT measurement value at the current moment through the prediction process of the Kalman filter algorithm; calculating the RTT estimation value at the current moment through the update process; dynamically updating the current RTT, which is the time interval after the Kalman filter at the current moment, based on the RTT measurement value and the RTT estimation value; and calculating the RTO at the current moment based on whether the packet is sent for the first time, whether it has timed out, or whether it has been retransmitted, combined with the RTT.

[0015] According to the above technical solution, when roaming occurs, the roaming timeout retransmission time is recalculated based on the RTO of the previous moment and the RTO dynamically calculated at the current moment, and the retransmission rate is reduced to retransmit the message; when the roaming ends, the message is retransmitted at a rate lower than the RTO dynamically calculated at the current moment and at a higher retransmission rate.

[0016] The present invention also provides a wireless retransmission module based on network tunnel technology, comprising:

[0017] The message acquisition submodule is used to capture the network IP messages that the user specifies need to be retransmitted;

[0018] The message reassembly submodule is used to add the pre-designed retransmission protocol header to the front of the user network IP message, reassemble it into a UDP message, and send it to the receiving end;

[0019] The message retransmission submodule is used to obtain the response packet returned by the receiving end, calculate the time interval RTT from the time the network message is sent to the time the response is received based on the response packet, dynamically calculate the timeout retransmission time RTO based on the RTT, and retransmit the message based on the RTO.

[0020] The present invention also provides a vehicle-to-ground wireless retransmission tunnel system, comprising a vehicle-mounted motion unit and a ground server, and establishing a communication tunnel between the two. Both the vehicle-mounted motion unit and the ground server are provided with a wireless retransmission module, which is the wireless retransmission module based on network tunnel technology as described in the above technical solution.

[0021] The present invention also provides a computer storage medium, which stores a computer program that can be executed by a processor, and the computer program executes the vehicle-to-ground wireless retransmission method based on network tunnel technology described in the above technical solution.

[0022] The beneficial effects of the present invention are as follows: the present invention adds a pre-designed retransmission protocol header to the network IP message that needs to be retransmitted, reassembles it into a UDP message and then sends it to the receiving end, and calculates the time interval RTT from the network message being sent out to the receiving of the response according to the information in the response packet returned by the receiving end, dynamically calculates the timeout retransmission time RTO according to the RTT, and retransmits the message according to the timeout retransmission time RTO. The present invention is applicable to a single network card solution, and can effectively solve the problem of packet loss and excessive balanced switching delay in the 802.11R switching protocol. Under normal operation, the network delay is not significantly improved.

[0023] Furthermore, based on the Kalman filter algorithm and the actual situation of rail transit, the optimal retransmission time of network messages is calculated to implement a high-performance timeout retransmission algorithm. By calculating the packet loss rate in real time, combined with link delay and roaming time, the communication quality is evaluated, and the automatic rate adjustment algorithm is used to dynamically optimize the maximum allowed communication speed to prevent a large number of timeout retransmission data packets from appearing at the moment of roaming, causing more congestion in the link, thereby ensuring the link quality.

[0024] Of course, any product implementing the present invention does not necessarily need to achieve all of the advantages described above at the same time. BRIEF DESCRIPTION OF THE DRAWINGS

[0025] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the following briefly introduces the drawings required for use in the embodiments or the description of the prior art. Obviously, the drawings described below are some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.

[0026] Figure 1This is a schematic diagram of the structure of a conventional dual network card solution for a subway tunnel vehicle-to-ground wireless system in the prior art;

[0027] Figure 2 is a schematic diagram of a subway line network topology structure according to an embodiment of the present invention;

[0028] Figure 3 yes Figure 2 A schematic diagram of a specific network tunnel topology structure;

[0029] Figure 4 is the relationship between the retransmission data message and the user data message in the embodiment of the present invention;

[0030] Figure 5 It is the specific structure of the retransmission protocol header in the embodiment of the present invention;

[0031] Figure 6 It is a transmission flow chart of a retransmission message in a network tunnel according to an embodiment of the present invention;

[0032] Figure 7 is a flow chart of calculating RTT based on the Kalman filter algorithm according to an embodiment of the present invention;

[0033] Figure 8 is a flow chart of calculating RTO based on RTT under the Kalman filter algorithm of an embodiment of the present invention;

[0034] Fig. 9 This is a flow chart of the roaming timing strategy under the Kalman filter algorithm of an embodiment of the present invention;

[0035] Fig.10 is a schematic diagram of the HACK algorithm according to an embodiment of the present invention;

[0036] Fig.11 is a schematic diagram of the SN algorithm according to an embodiment of the present invention;

[0037] Fig.12 is a rate modulation control flow chart of an embodiment of the present invention;

[0038] Fig.13 is a schematic diagram of multi-channel connection according to an embodiment of the present invention;

[0039] Fig.14 It is a flow chart of a vehicle-to-ground wireless retransmission method based on network tunnel technology according to an embodiment of the present invention. DETAILED DESCRIPTION

[0040] In order to make the purpose, technical solution and advantages of the present invention more clearly understood, the present invention is further described in detail below in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present invention and are not used to limit the present invention.

[0041] It should be noted that the illustrations provided in the embodiments of the present invention are only used to illustrate the basic concept of the present invention in a schematic manner. Therefore, the drawings only show components related to the present invention rather than being drawn according to the number, shape and size of components in actual implementation. In actual implementation, the type, quantity and proportion of each component may be changed arbitrarily, and the component layout type may also be more complicated.

[0042] In the present invention, it is also necessary to explain that, if the terms "center", "upper", "lower", "left", "right", "vertical", "horizontal", "inner", "outer", etc. appear, the orientation or position relationship indicated is based on the orientation or position relationship shown in the drawings, which is only for the convenience of describing the present application and simplifying the description, and does not indicate or imply that the device or element referred to must have a specific orientation, be constructed and operated in a specific orientation, and therefore cannot be understood as a limitation on the present application. In addition, if the terms "first" and "second" appear, they are only used for description and distinction purposes, and cannot be understood as indicating or implying relative importance.

[0043] like Fig.14 As shown, the vehicle-to-ground wireless retransmission method based on network tunnel technology in an embodiment of the present invention mainly includes the following steps:

[0044] S1, capture the network IP message that the user specifies needs to be retransmitted;

[0045] S2, add the pre-designed retransmission protocol header to the front of the user network IP message, reassemble it into a UDP message, and send it to the receiving end; Figure 4 As shown, the originally acquired network message and the reassembled retransmitted network message.

[0046] S3. Obtain the response packet returned by the receiving end, calculate the time interval RTT from the time the network message is sent out to the time the response is received according to the response packet, dynamically calculate the timeout retransmission time RTO according to the RTT, and retransmit the message according to the timeout retransmission time RTO.

[0047] The retransmission protocol header pre-designed by the present invention includes multiple fields, such as Figure 5 As shown, it mainly includes: packet sequence number sn, used to determine the packet order; sending packet timestamp ts; fast retransmission number hack; message type cmd; checksum, which is the retransmission protocol header checksum; response port number ap, used to resolve the port number of the receiving end to send the response message; its own receiving window size window; payload data length len.

[0048] The response packet includes a response protocol header and a data packet, wherein the response protocol header includes a hack field, which records the minimum packet sequence number sn that the receiving end has received, indicating that all data packets smaller than the hack field have been sent successfully; the data packet records the packet sequence number and receiving timestamp of the packet received by the receiving end.

[0049] After receiving the response packet, the sender will mark the sequence number of the lost packet according to the response packet. When the number of markings is greater than a certain threshold, the packet will be retransmitted directly.

[0050] When a packet is retransmitted multiple times by the sender but still no corresponding response packet is received, it will not be retransmitted.

[0051] The sending end dynamically calculates the RTO through a low-pass filter and in combination with the RTT variance, specifically including the following steps: calculating the actual RTT measurement value at the current moment according to the sending packet timestamp in the retransmission protocol header and the receiving packet timestamp in the response packet; calculating the average estimated RTT value at the current moment; calculating the RTT measurement value at the current moment according to the RTT measurement value and the average estimated RTT value; calculating the average estimated value of the jitter value at the current moment; calculating the RTO at the current moment according to the average estimated RTT value and the average estimated value of the jitter value.

[0052] The sending end dynamically calculates the RTO through the Kalman filter algorithm, which specifically includes the following steps: calculating the RTT measurement value at the current moment through the prediction process of the Kalman filter algorithm; calculating the RTT estimation value at the current moment through the update process; dynamically updating the current RTT, which is the time interval after the Kalman filter at the current moment, based on the RTT measurement value and the RTT estimation value; and calculating the RTO at the current moment based on whether the packet is sent for the first time, whether it has timed out, or whether it has been retransmitted, combined with the RTT.

[0053] When roaming occurs, the roaming timeout retransmission time is recalculated based on the RTO of the previous moment and the RTO dynamically calculated at the current moment, and the retransmission rate is reduced to retransmit the message; when roaming ends, the RTO is lower than the RTO dynamically calculated at the current moment, and the retransmission rate is increased to retransmit the message.

[0054] Furthermore, users can specify data traffic for retransmission, and based on network tunnel technology, the retransmitted traffic can also be encrypted; the network delay is automatically calculated through the Kalman filter algorithm to accurately control the network message retransmission time; the adaptive sending rate control can effectively solve the network congestion caused by the sudden drop in network quality at the moment of switching; the HACK and SN algorithms can achieve fast retransmission instead of waiting for timeout retransmission.

[0055] The wireless retransmission module based on the network tunnel technology in the embodiment of the present invention is mainly used to implement the above method embodiment. The wireless retransmission module mainly includes:

[0056] The message acquisition submodule is used to capture the network IP messages that the user specifies need to be retransmitted;

[0057] The message reassembly submodule is used to add the pre-designed retransmission protocol header to the front of the user network IP message, reassemble it into a UDP message, and send it to the receiving end;

[0058] The message retransmission submodule is used to obtain the response packet returned by the receiving end, calculate the time interval RTT from the network message being sent to the receipt of the response based on the response packet, and dynamically calculate the timeout retransmission time RTO based on the RTT, and retransmit the message based on the RTO.

[0059] like Figure 2 As shown, the network topology of the entire train-to-ground wireless retransmission tunnel system is realized. The present invention aims to achieve network communication between the internal network of the train and the ground control center to ensure the network quality. The train-to-ground wireless retransmission tunnel system includes a vehicle-mounted motion unit and a ground server, and a communication tunnel is established between the two. The vehicle-mounted motion unit and the ground server are both provided with a wireless retransmission module, which is a wireless retransmission module based on the network tunnel technology of the above-mentioned embodiment. The structure is to achieve:

[0060] (1) Uplink traffic STA->AC: The STA end can set specified traffic according to the vehicle user, such as specified UDP, specified port, or specified dscp tag traffic, which can be intercepted by the STA retransmission system, enter the retransmission tunnel, and be retransmitted and encrypted (this function is optional);

[0061] (2) Downlink traffic AC->STA: Similarly, the AC side can specify traffic according to the user settings of the ground control center, such as traffic with specified UDP, specified port, or specified dscp tag, which can be intercepted by the AC retransmission system, enter the retransmission tunnel, and be retransmitted and encrypted (this function is optional);

[0062] (3) Regardless of uplink or downlink traffic, as long as it enters the tunnel retransmission system, it can achieve packet retransmission of network data packets after packet loss, ensuring a packet loss effect of 0.001% or even 0% during normal operation or roaming.

[0063] like Figure 3 As shown, Figure 2The specific network structure of the system, in which the network tunnel technology can use the Linux tun device driver, combined with iptables technology to achieve network message interception; use the Linux socket API on the STA or AC side to send and receive retransmitted data packets and response packets; when receiving the network message of the tun device, you can choose to encrypt or not. If encrypted, the retransmitted data packet will be decrypted after being received by the other end and then sent to the user, realizing lossless restoration of user data messages and sending them to the user. The main work of this system is to use a series of retransmission algorithms to transmit user network messages with high quality.

[0064] like Figure 4 As shown in the figure, it is a comparison diagram of the structure of the retransmitted data message and the user data message. Since the user-specified network message is to be retransmitted, a specific control field must be added, such as Figure 4 The retransmission protocol header in the program allows the program to obtain certain transmission information and perform retransmission control.

[0065] The specific definition of the retransmission protocol header is as follows Figure 5 As shown, it mainly includes:

[0066] a) sn: packet number, used to determine the order of the packet;

[0067] b)ts: timestamp of receiving packet;

[0068] c)hack: fast retransmission number;

[0069] d) cmd: message type;

[0070] e) Checksum: 8-bit retransmission protocol header checksum;

[0071] f)ap: response port number, used to resolve the port number of the receiving end to send the response message;

[0072] g)window: receives the window size by itself;

[0073] h))len: payload data length.

[0074] Retransmitting data packets is to pack the entire IP packet of the user as a whole into the retransmitted network packet. The so-called retransmission is to retransmit the IP packet of the user.

[0075] The user specifies that network packets enter the retransmission system, which mainly uses Linux Tun device driver and Iptables technology to control traffic. Figure 6As shown, the specific process is as follows: Use Iptables to MARK the network message specified by the user; Note: Here, the PREROUTING chain of the mangle table is used to set the MARK rule. Import the MARKed message into the custom routing table tuntable; Set the custom routing table rule to use the tun device for forwarding by default; After the above steps, the user-specified network IP message can be captured from the tun device. Add a custom retransmission protocol header to the front of the user network IP message; Send the reassembled UDP message to the receiving end, Note: Encryption processing can be performed at this step; Parse the retransmission protocol header at the receiving end and perform retransmission algorithm analysis; Remove the retransmission protocol header and send the user network IP message out, Note: If it is encrypted, decryption processing is required here.

[0076] The calculation of the network message timeout retransmission time RTO is based on RTT (the time interval from when the network message is sent out to when the response is received). RTT (the time interval from when the network message is sent out to when the response is received) can directly reflect the quality of the network link. How to accurately calculate RTT and predict RTO is particularly important. The present invention provides two calculation methods: (1) an RTO calculation method based on a low-pass filter and combined with RTT variance; (2) an RTO calculation method based on Kalman filtering.

[0077] Among them, the RTO calculation method based on low-pass filter and combined with RTT variance is mainly as follows:

[0078] 1) Get the actual RTT measurement value at time k: z k =t2-t1, where t1: timestamp of sending data packet, t2: timestamp of receiving response packet, z k : The actual RTT measurement value at time k.

[0079] 2) Calculate the average estimated RTT at time k: Among them, β: takes the value of 1 / 8, The average estimated value of RTT at time k-1, Average estimated RTT at time k;

[0080] 3)3) Calculate the jitter value of the RTT measured at time k: e k : The jitter of RTT measured at time k.

[0081] 4) Calculate the average estimated value of the jitter of RTT at time k: x k =(1-α)*x k-1 +α*e k , where α: takes the value of 1 / 4, x k-1 : The average estimated value of the jitter of RTT at time k-1;

[0082] 5) Calculate rto at time k: Where γ: takes the value of 4, rto: rto at time k.

[0083] The RTO calculation method based on Kalman filtering is mainly as follows:

[0084] Kalman filtering is a recursive predictive filtering algorithm that provides an efficient and computable method to estimate the state of a process and minimize the estimated mean square error.

[0085] Kalman filtering includes prediction process and update process:

[0086] 1) Prediction process

[0087]

[0088] P k - =AP k-1 A T +Q

[0089] 2) Update process

[0090] G k =P k - H T (HP k - H T +R) -1

[0091]

[0092] P k =(1-G k H)P k -

[0093] The meaning of variables in the above formula: z k : measured value at time k, The prior estimate of time k; Posterior state estimate at time k; The posterior state estimate at time k-1; P k - : The prior estimated covariance at time k, that is The covariance of k : The posterior estimated covariance at time k, that is covariance; A: state transfer matrix, B: control input matrix, u k : measurement noise at time k, R: measurement noise covariance, G k: Kalman gain, Q: covariance of the system process, this parameter is used to represent the error between the state transition matrix and the actual process, H: conversion matrix from state variables to measurements.

[0094] In addition, the technical terms involved include: RTT (Round-Trip Time): the time interval from when a network message is sent to when a response is received; RTO: timeout retransmission time; KRTT: RTT value after Kalman filtering; KHRTT: observation matrix; KARTT: state matrix; KQRTT: state matrix variance; KRRTT: observation matrix variance; KPRTT: prediction error; KBRTT: RTT measurement coefficient; KURTT: RTT measurement noise; KGAIN: Kalman gain.

[0095] Among them, the process of calculating RTT based on Kalman filtering is as follows Figure 7 As shown, the timestamps t1 and t2 are stored in the ts field. Figure 7 As shown, it mainly includes the following steps:

[0096] 1) Measurement process, calculate the measurement value at time k: z k = t2-t1, where z k : The measured value at time k, t1: The timestamp of sending the data packet, t2: The timestamp of receiving the response.

[0097] 2) Kalman prediction process, calculation:

[0098]

[0099] P k - =AP k-1 A T +Q

[0100] in The a priori estimate of time k is obtained by iterative calculation; The posterior state estimate at time k-1, i.e. the last output filter value, is obtained by iterative calculation; A: state transfer matrix, determined at the initialization time; B: control input matrix, determined at the initialization time; u k : k-time measurement noise, determined at the initialization time; P k - : The prior estimated covariance at time k, that is The covariance of is calculated iteratively; P k-1 : The prior estimated covariance at time k-1, obtained by iterative calculation; Q: The covariance of the system process, determined at the initialization moment.

[0101] 3) Kalman update process:

[0102] Gk =P k - H T (HP k - H T +R) -1

[0103]

[0104]

[0105] Among them G k : Kalman gain at time k, obtained by iterative calculation; H: conversion matrix from state variables to measurements, determined at the initialization time; R: measurement noise covariance, determined at the initialization time. After calculation, the desired filtered KRTT value will be obtained. Then, according to different operating states, combined with KRTT, different strategies are used to update the RTO calculation.

[0106] The main process of calculating RTO based on RTT is as follows Figure 8 As shown in the figure, different update algorithms are adopted mainly based on the KRTT obtained by Kalman filtering and different operating states. Figure 8 As shown, specifically: determine whether the packet is sent for the first time, if so, RTO = KRTT + KRTT / 3, where the round trip time of the data packet is KRTT. If the set RTO time is less than KRTT, a timeout retransmission will occur prematurely. If it is much larger than the KRTT time, the sender will not be able to retransmit in time. Therefore, when the packet is sent for the first time, it is expected that the first RTO will be slightly larger than KRTT (the value is: KRTT / 3), so the present invention takes the value of KRTT + KRTT / 3; if the packet is not sent for the first time, continue to determine whether the packet has timed out, if so, RTO + = Last RTO / 4. If the packet has timed out and retransmitted, the expected RTO increases with the number of retransmissions. For example, the first RTO1=50, the second RTO2=RTO1+RTO1 / 4=62.5, the third RTO3=RTO2+RTO2 / 4=78.125, and the nth RTO n =RTO n-1 +RTO n-1 / 4. If the packet has not timed out, continue to determine whether the packet is quickly retransmitted. If so, RTO = last RTO, that is, if the packet is sent for fast retransmission, it means that the network is temporarily jittering, resulting in timeout. At this time, the packet is directly retransmitted quickly without waiting for RTO, and there is no need to modify the RTO value. The last RTO can be directly used. If the packet does not need to be quickly retransmitted, the process ends.

[0107] There are several calculation methods according to different situations:

[0108] 1. Roaming Time RTO Algorithm

[0109] According to actual tests, when the above update state is used under normal communication conditions, the data packet delay is very low and there is zero packet loss. However, during roaming, a large number of retransmitted data packets will appear, and the RTO update is obviously not suitable. Therefore, a unique RTO calculation algorithm is required.

[0110] like Fig. 9 As shown in the figure, during roaming, the network quality is not good. Here, it is necessary to increase the RTO to a certain extent and adjust the sending window to ensure that the inbound traffic is temporarily stored in the device and wait for the roaming moment to end. To reduce the RTO, the device cache traffic should be sent as soon as possible. Specifically, the following steps are included: determine whether roaming occurs. If so, RTO = last RTO + KRTT + KRTT / 5, adjust the sending rate (decrease), if roaming occurs, the roaming process takes a certain amount of time, assuming Troam, theoretically expect to immediately suspend data transmission, and send data immediately after roaming ends, so a special RTO calculation method RTO + KRTT + KRTT / 5 is used here. According to the actual experience after optimizing 80211r, the calculated RTO is larger than Troam at this time, so at least during the Troam time, no data will be transmitted, meeting the above expectations; if not, continue to determine whether roaming ends, if so, RTO = KRTT / 2, adjust the sending rate (increase), as expected above, after roaming ends, send data immediately, and send the data packets accumulated during roaming as soon as possible, so here RTO will be shorter than the normal KRTT, which is half of KRTT. If roaming has not ended, continue to determine until roaming ends.

[0111] 2. ACK algorithm

[0112] like Figure 5 As shown in the figure, there is a hack field in the retransmission protocol header, which records the smallest sn number that the receiving end has received. When the sending end receives the hack field of the response packet, it will mark all data packets in the sending cache that are smaller than the hack field as successfully sent, telling the sending end that it does not need to wait for the timeout. Fig.10 As shown in the figure, the sender has sent out packets 0 to 4, and the receiver has actually received packets 0 to 2. When the receiver sends a response packet of 0 to the sender, it will carry hack=3, telling the sender that all packets less than 3 have been received and can be deleted. In this way, after the receiver only receives the response packet of 0, it also knows that packets 1 to 2 have been sent successfully, so 1 and 2 are deleted from the sending cache without waiting for timeout.

[0113] 3. SN algorithm

[0114] There is a sn field in the retransmission protocol header. When the sender receives the response packet, it will mark that the data packets smaller than the sn have been lost. When the number of markings is greater than a certain threshold, it means that the packet has been lost and there is no need to wait for timeout retransmission. The packet can be retransmitted directly. This reduces many timeout packet problems and response packet loss problems.

[0115] like Fig.11 As shown: If the sender has sent packets 0 to 4, the receiver will not respond to the sender because it has not received the packet 1. The receiver will only send the response packet of sn=2 to the sender. After receiving it, the sender finds that 1 has been skipped once, so it marks it once. When it receives the response packet of sn=3 again, the sender finds that 1 has been skipped again. When the number of skips is greater than a certain threshold, it does not need to wait for RTO, and directly sends 1 packet, and marks the state as fast retransmission. Note: There is also a TTL update algorithm here, and the strategy is: when a packet is retransmitted many times and still no response packet is received, it means that the packet does not need to be sent again, and the TTL is directly aged.

[0116] 4. Rate modulation control

[0117] The purpose of rate modulation control is to ensure that the maximum allowed sending rate is adjusted under different operating conditions to control network congestion and take into account the requirements of low latency and high bandwidth.

[0118] like Fig.12 As shown, the modulation transmission rate varies with the different operating states; K2 = K / a, K1 = last transmission rate / b, a and b are adjustment coefficients, the empirical values ​​are a = 2, b = 2, and the parameters can be set; after adjustment, the transmission rate will be in a dynamic balance process. When 80211R fast roaming occurs, retransmission packets will continue to appear. At this time, the system will run in the startup phase most of the time, that is, constantly trying to increase the transmission rate, but not too high, thereby reducing congestion caused by large amounts of data during the fast roaming phase. The specific steps include: startup phase, 发The sending rate increases exponentially every RTT time. In the startup phase, since no data has been sent, the actual network quality cannot be detected. Here, the sending speed is increased exponentially until it reaches the threshold K; the sending rate increases linearly. When the sending rate reaches the threshold K, the sending rate needs to be reduced and the sending rate is allowed to increase linearly to slowly detect the bandwidth limit capacity until packet loss occurs; when packet loss occurs, determine whether to retransmit quickly. If so, quickly recover and adjust the starting threshold K1=K / 2, and adjust the current sending rate to K1+last sending rate / 3. If not, continue to determine whether to retransmit after timeout. If not, return to the previous step to determine whether to retransmit quickly. If so, packet loss recovery adjusts the starting threshold K2=K / 2, and adjusts the current sending rate to =last sending rate / 5. That is, in the linear growth stage, if a timeout retransmission occurs, it means that the bandwidth limit capacity has been exceeded, and it is expected that the rate will be quickly reduced. Therefore, the start threshold is reduced to half of the previous one, that is, K1=K / 2, and the sending rate is adjusted to / 5 of the previous sending rate, and then the restart state is entered. The difference between this process and the fast retransmission is that the current sending rate adjustment is different. Fast retransmission: the current sending rate Vf=K1+V_last / 3, after adjustment Vf is above K1, and an offset V_last / 3 is added, packet loss recovery: the current sending rate Vl=V_last / 5, after adjustment Vl is only V_last / 5. In summary: Vf is greater than VI, and compared with VI, Vf does not reduce the rate much, while Vl reduces a lot, which meets the above expectations.

[0119] 5. Multi-channel connection

[0120] like Figure 5 As shown: There is an ap (ack port) field in the protocol header, which is used to solve the following problems:

[0121] like Fig.13 As shown in the figure: There must be many STAs on a subway line. When all STAs communicate with AC through AP to establish a retransmission tunnel, AC needs to support multiple channels and know which STA device the response packet should be sent to. Then the ap field comes into play. When each STA sends data to AC, it will carry the ap field to tell AC its own response port. Then when AC receives the data packet, it will immediately establish an ACK retransmission channel based on the STA address and ap field, thus ensuring that AC can support multiple STAs.

[0122] The present invention uses a communication tunnel established between a vehicle-mounted motion unit (STA) and a ground server (AC) to intercept important data messages on the communication link for tunnel retransmission. The present invention collects link delays in real time, calculates the optimal retransmission time of network messages based on the Kalman filter algorithm and the actual situation of rail transit, and implements a high-performance timeout retransmission algorithm; by calculating the packet loss rate in real time, combining the link delay and roaming time opportunity, evaluating the communication quality, and dynamically optimizing the maximum communication speed allowed according to the automatic rate adjustment algorithm, a large number of timeout retransmission data packets are prevented from appearing at the moment of roaming, causing the link to be more congested, thereby ensuring the link quality; through the HACK response and SN algorithm, packet loss is quickly retransmitted; through the development of the above algorithm and actual engineering optimization, the single network card solution of the vehicle-ground wireless system and the problem of excessive packet loss and balanced switching delay in the 802.11R switching protocol are solved.

[0123] The present application also provides a computer-readable storage medium, such as a flash memory, a hard disk, a multimedia card, a card-type memory (e.g., an SD or DX memory, etc.), a random access memory (RAM), a static random access memory (SRAM), a read-only memory (ROM), an electrically erasable programmable read-only memory (EEPROM), a programmable read-only memory (PROM), a magnetic memory, a disk, an optical disk, a server, an App application store, etc., on which a computer program is stored, and when the program is executed by a processor, a corresponding function is realized. When the computer-readable storage medium of this embodiment is executed by a processor, the vehicle-to-ground wireless retransmission method based on network tunnel technology of the method embodiment is realized.

[0124] It should be pointed out that, according to the needs of implementation, the various steps / components described in this application can be split into more steps / components, and two or more steps / components or partial operations of steps / components can be combined into new steps / components to achieve the purpose of the present invention.

[0125] The order of execution of each step in the above embodiment does not mean the order of execution. The execution order of each process should be determined by its function and internal logic, and should not constitute any limitation on the implementation process of the embodiment of the present application.

[0126] It should be understood that those skilled in the art can make improvements or changes based on the above description, and all these improvements and changes should fall within the scope of protection of the appended claims of the present invention.

Claims

1. A vehicle-to-ground wireless retransmission method based on network tunnel technology, characterized in that: The following steps are involved: Capture the network IP packets that the user specifies need to be retransmitted; Add the pre-designed retransmission protocol header to the front of the user network IP message, reassemble it into a UDP message, and send it to the receiving end; The response packet returned by the receiving end is obtained. The time interval RTT from sending the network message to receiving the response is calculated based on the response packet. The timeout retransmission time RTO is dynamically calculated based on the RTT, and the message is retransmitted based on the timeout retransmission time RTO.

2. The vehicle-to-ground wireless retransmission method based on network tunnel technology according to claim 1 is characterized in that: The retransmission protocol header includes multiple fields: the packet sequence number sn, which is used to determine the packet sequence; Sending packet timestamp ts; Fast retransmission number hack; message type cmd; checksum, which is the retransmission protocol header checksum; The response port number ap is used to determine the port number of the receiving end to send the response message; the size of its own receiving window window; Payload data length len.

3. The vehicle-to-ground wireless retransmission method based on network tunnel technology according to claim 1 is characterized in that: The response packet includes a response protocol header and a data packet, wherein the response protocol header includes a hack field, which records the minimum packet sequence number sn that the receiving end has received, indicating that all data packets smaller than the hack field have been sent successfully; the data packet records the packet sequence number and receiving timestamp of the packet received by the receiving end.

4. The vehicle-to-ground wireless retransmission method based on network tunnel technology according to claim 1 is characterized in that: After receiving the response packet, the sender will mark the sequence number of the lost packet according to the response packet. When the number of markings is greater than a certain threshold, the packet will be retransmitted directly.

5. The vehicle-to-ground wireless retransmission method based on network tunnel technology according to claim 1 is characterized in that: When a packet is retransmitted multiple times by the sender but still no corresponding response packet is received, it will not be retransmitted.

6. The vehicle-to-ground wireless retransmission method based on network tunnel technology according to claim 1 is characterized in that: The sending end dynamically calculates the RTO through a low-pass filter and in combination with the RTT variance, specifically including the steps of: calculating the actual RTT measurement value at the current moment based on the sending packet timestamp in the retransmission protocol header and the receiving packet timestamp in the reply packet; Calculate the average estimated RTT value at the current time; Calculate the current RTT measurement value based on the RTT measurement value and the average RTT estimation value; Calculate the average estimated value of the jitter value at the current moment; The RTO at the current moment is calculated based on the average estimated value of the RTT and the average estimated value of the jitter value.

7. The vehicle-to-ground wireless retransmission method based on network tunnel technology according to claim 1 is characterized in that: The sending end dynamically calculates the RTO through the Kalman filter algorithm, which specifically includes the following steps: calculating the RTT measurement value at the current moment through the prediction process of the Kalman filter algorithm; calculating the RTT estimation value at the current moment through the update process; dynamically updating the current RTT, which is the time interval after the Kalman filter at the current moment, based on the RTT measurement value and the RTT estimation value; and calculating the RTO at the current moment based on whether the packet is sent for the first time, whether it has timed out, or whether it has been retransmitted, combined with the RTT.

8. The vehicle-to-ground wireless retransmission method based on network tunnel technology according to claim 1 is characterized in that: When roaming occurs, the roaming timeout retransmission time is recalculated based on the RTO of the previous moment and the RTO dynamically calculated at the current moment, and the retransmission rate is reduced to retransmit the message; when roaming ends, the RTO is lower than the RTO dynamically calculated at the current moment, and the retransmission rate is increased to retransmit the message.

9. A wireless retransmission module based on network tunnel technology, characterized in that: include: The message acquisition submodule is used to capture the network IP messages that the user specifies need to be retransmitted; The message reassembly submodule is used to add the pre-designed retransmission protocol header to the front of the user network IP message, reassemble it into a UDP message, and send it to the receiving end; The message retransmission submodule is used to obtain the response packet returned by the receiving end, calculate the time interval RTT from the time the network message is sent to the time the response is received based on the response packet, dynamically calculate the timeout retransmission time RTO based on the RTT, and retransmit the message based on the RTO.

10. A vehicle-to-ground wireless retransmission tunnel system, characterized in that: It includes a vehicle-mounted motion unit and a ground server, and a communication tunnel is established between the two. Both the vehicle-mounted motion unit and the ground server are provided with a wireless retransmission module, which is the wireless retransmission module based on network tunnel technology as described in claim 9.

Images