Vehicle-mounted ethernet tsn network performance test system and method

Abstract

The application discloses a kind of vehicle-mounted Ethernet TSN network performance test system and method, the system includes multiple vehicle-mounted Ethernet simulator, passive TAP test tool, traffic generation module and characteristic rule matching module;Multiple vehicle-mounted Ethernet simulator is used to simulate multiple application vehicle-mounted terminal, traffic generation module generates specific data stream and sends to application vehicle-mounted terminal, arrives destination application vehicle-mounted terminal by TSN network forwarding;Passive TAP test tool obtains the mixed data stream formed by the data stream of sending and receiving;Characteristic rule matching module obtains the time information of specific data stream to reach corresponding TAP test point by matching specific data stream with the mixed data stream obtained, the performance index of TSN network is calculated by comparing the sending time of specific data stream and the time of reaching corresponding TAP test point.This application can realize the non-inductive test of TSN network, and accurately obtain the network transmission performance of TSN network.

Description

Technical Field

[0001] This invention belongs to the field of network performance testing, specifically relating to a vehicle-mounted Ethernet TSN network performance testing system and method. Background Technology

[0002] Time-Sensitive Networking (TSN) is a set of Ethernet enhancement standards developed by the IEEE 802.1 working group. It aims to provide deterministic data transmission services for Ethernet-based networks. The application of TSN in automotive Ethernet will become one of the core competitive advantages of future intelligent vehicles. Current performance testing of automotive Ethernet TSN networks often uses active TAP (Test Access Point) testing technology. Traditional TAP testing techniques (including active and passive TAP testing) are typically applied in full-duplex Ethernet networks with multiple twisted pairs, where TX (transmit) and RX (receive) are transmitted over different twisted pairs. However, automotive Ethernet uses the same twisted pair for full-duplex communication, causing the TX and RX signals to overlap on the same pair. If passive TAP testing is used, it will be impossible to demodulate the data transmitted in the east and west directions separately. Therefore, at present, only active TAP testing can be used in the field of automotive Ethernet. However, active TAP testing uses active devices in the network link, making it impossible to perform unobtrusive testing. The intervention of active TAPs affects the original TSN network state, leading to inaccurate capture of TSN communication performance. Furthermore, active TAP testing introduces additional time delays into the TSN network system. As the number of TAP nodes in the TSN network increases, the related time-sensitive performance measurements become increasingly inaccurate, with a widening deviation range. Summary of the Invention

[0003] The purpose of this invention is to provide a vehicle-mounted Ethernet TSN network performance testing system and method, which can realize seamless testing of TSN networks and accurately obtain the network transmission performance of TSN networks.

[0004] One aspect of the present invention provides a vehicle Ethernet TSN network performance testing system, wherein the vehicle Ethernet TSN network uses a single pair of twisted-pair cables for full-duplex communication, and the system includes multiple vehicle Ethernet simulators, a passive TAP testing tool, a traffic generation module, and a feature rule matching module;

[0005] The multiple vehicle Ethernet simulators are used to simulate multiple application vehicle terminals. The vehicle Ethernet simulators can convert data streams into physical electrical signals of the vehicle Ethernet standard and send them in the TSN network, and can receive physical electrical signals of the vehicle Ethernet standard from the TSN network and convert them into received data streams.

[0006] The traffic generation module is used to generate a specific data stream and send it to one of the multiple application vehicle terminals. The generated data stream is then forwarded by the application vehicle terminal to the TSN network, and then forwarded by the TSN network to another of the multiple application vehicle terminals.

[0007] The passive TAP test tool is set at a TAP test point in the TSN network. At the TAP test point, the transmitted and received data streams form a full-duplex mixed data stream. The passive TAP test tool is used to acquire the mixed data stream.

[0008] The feature rule matching module is used to match a specific data stream with a mixed data stream obtained from the TAP test point to obtain the time information of the specific data stream arriving at the corresponding TAP test point. By comparing the sending time and arrival time of the specific data stream at the corresponding TAP test point, the performance index of the TSN network is calculated.

[0009] Preferably, the TSN network includes a TSN switch, a routing gateway, and a background vehicle terminal. The background vehicle terminal is simulated by an in-vehicle Ethernet simulator. The traffic generation module is also used to generate a background flow for the background vehicle terminal. The background flow is the natural traffic of the TSN network during actual operation. The feature rule matching module can identify the specific data flow generated by the traffic generation module without being affected by the background flow.

[0010] Preferably, the system further includes a host computer, with the traffic generation module and feature rule matching module set in the host computer, and the vehicle Ethernet simulator and passive TAP testing tool connected to the host computer via cables.

[0011] Another aspect of the present invention provides a method for testing the performance of an in-vehicle Ethernet TSN network, which utilizes the above-described system to perform performance testing of an in-vehicle Ethernet TSN network, including:

[0012] The traffic generation module generates a specific data stream and sends it to one of the multiple application vehicle terminals. The generated data stream is then forwarded by the application vehicle terminal to the TSN network, and then forwarded by the TSN network to another of the multiple application vehicle terminals.

[0013] The feature rule matching module matches a specific data stream with a mixed data stream obtained from the TAP test point to obtain the time information of the specific data stream arriving at the corresponding TAP test point. By comparing the sending time and arrival time of the specific data stream at the corresponding TAP test point, the performance index of the TSN network is calculated.

[0014] Preferably, the matching of specific data streams and mixed data streams is performed as follows:

[0015] Define multiple target sequences for a specific data stream. The length of each target sequence is greater than a predetermined number of bits. Each target sequence is extracted from each data packet of the specific data stream and contains the sequence number of that data packet.

[0016] Detection sequences are continuously acquired through TAP test points. The difference between the detection sequence and the target sequence is used to determine the legality of the match and whether the match is successful.

[0017] Record the sequence number of the successfully matched data packet and the time it arrived at the TAP test point.

[0018] Preferably, the validity of the difference is determined as follows:

[0019] For each position i, calculate the difference sequence di = Mi - Ni between the detection sequence Mi and the target sequence Ni;

[0020] If there exists any If so, it is determined that the detected sequence cannot be a sequence generated by superimposing and coupling the target sequence;

[0021] If all di satisfy If so, it is determined that the detection sequence is generated by superimposing and coupling the target sequence.

[0022] Preferably, when the vehicle Ethernet is a vehicle 100 Mbps Ethernet using PAM3 encoding, the predetermined number of bits is 60 consecutive bits; when the vehicle Ethernet is a vehicle gigabit Ethernet using PAM3 encoding, the predetermined number of bits is 63 consecutive bits.

[0023] Preferably, assuming the target sequence has n state bits, and after matching, the detected sequence has 'a' fully matched state bits, 'b' partially matched state bits, and 'c' compatible state bits, where a + b + c = n, then in the case of PAM3 encoding for vehicular Ethernet, the predetermined number of bits is: The probability of random sequence coupling collisions occurring for

[0024] ,

[0025] Assuming that at least X bits need to be transmitted before a random sequence coupling collision occurs once, X is calculated as follows:

[0026] .

[0027] Preferably, a full match means that when the state bit corresponding to the target sequence is -1, the state bit of the detected sequence is -2; when the state bit corresponding to the target sequence is 1, the state bit of the detected sequence is 2.

[0028] Half-matching means that when the state bit corresponding to the target sequence is -1 or 0, the state bit of the detection sequence is -1; when the state bit corresponding to the target sequence is 1 or 0, the state bit of the detection sequence is 1.

[0029] The compatibility state refers to the state bit of the detection sequence being 0 when the state bit of the target sequence is -1, 0, or 1.

[0030] Preferably, the performance metrics include throughput, one-way latency, and latency jitter.

[0031] According to the vehicle-mounted Ethernet TSN network performance testing system and method of the present invention described above, it is possible to achieve seamless testing of TSN networks and accurately obtain the network transmission performance of TSN networks. Attached Figure Description

[0032] To more clearly illustrate the technical solutions of the present invention, the accompanying drawings used in the description of the embodiments of the present invention will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort:

[0033] Figure 1 This is a schematic diagram of the structure of an in-vehicle Ethernet TSN network performance testing system according to an embodiment of the present invention;

[0034] Figure 2 This is a schematic diagram of data stream transmission at a TAP test point according to an embodiment of the present invention;

[0035] Figure 3 This is a flowchart of a feature rule matching algorithm according to an embodiment of the present invention;

[0036] Figure 4 This is a schematic diagram illustrating the principle of a vehicle-mounted Ethernet TSN network performance testing method according to an embodiment of the present invention. Detailed Implementation

[0037] To make the objectives, technical solutions, and advantages of this invention clearer, the technical solutions of this invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of this invention, and not all embodiments. Based on the embodiments of this invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of this invention.

[0038] One embodiment of the present invention provides a performance testing system for an in-vehicle Ethernet TSN network, which combines active measurement technology and passive TAP testing technology. It identifies data flows during active measurement through feature rule matching, thereby performing performance tests such as throughput, one-way latency, and latency jitter of the in-vehicle Ethernet TSN network without affecting performance. Active measurement technology refers to actively generating data flows and injecting them into the network under test while simultaneously monitoring their transmission. In contrast, passive measurement only monitors data flow transmission without actively initiating data flows. This invention actively sends specific application flow data packets using active measurement technology. Because the packets are actively sent, the sequence and size of each data packet are known. Knowing the sequence and time of the data packets arriving at the destination allows calculation of performance indicators such as throughput, one-way latency, and latency jitter. Throughput refers to the amount of data transmitted per second; one-way latency is the difference between the time a data packet is sent and the time it arrives at the destination; and latency jitter is the difference between the maximum and minimum latency of multiple data packets.

[0039] In this embodiment of the invention, the vehicular Ethernet TSN network, i.e., the physical layer of the vehicular communication link, uses a single pair of twisted-pair cables for full-duplex communication and features time-sensitive networking with time synchronization, traffic scheduling and shaping, redundancy and fault tolerance, and resource management. The vehicular Ethernet includes standards such as 1000BASE-T1, 100BASE-T1, 2.5G / 5G / 10GBASE-T1, and 10BASE-T1S. A network with TSN functionality must meet the functional protocols of standards such as IEEE 802.1AS / AS-Rev (time synchronization protocol), IEEE 802.1Qbv (time-aware shaping), IEEE 802.1Qav (credit shaping), IEEE 802.1Qbu (frame preemption), IEEE 802.1CB (frame duplication and cancellation), IEEE 802.1Qat (flow reservation protocol), and IEEE 802.1Qcc (centralized configuration protocol).

[0040] like Figure 1 As shown, an embodiment of the vehicular Ethernet TSN network performance testing system of the present invention includes multiple (at least two) application vehicular terminals simulated by a vehicular Ethernet simulator (test instrument), a passive TAP testing tool, a traffic generation module, and a feature rule matching module. The multiple application vehicular terminals include application vehicular terminal 1, application vehicular terminal 2, ..., application vehicular terminal N. The passive TAP testing tool is set at TAP test points in the TSN network under test, for example... Figure 1 The test includes TAP test points 1 and 2. At each TAP test point, the transmitted and received data streams form a full-duplex mixed data stream, and the passive TAP test tool is used to acquire this mixed data stream.

[0041] The traffic generation module generates a specific data stream and sends it to the application vehicle terminal (the sender). This data stream is then forwarded by the application vehicle terminal to the TSN network under test, and finally forwarded by the TSN network to the application vehicle terminal (the receiver). The feature rule matching module matches the known specific data stream with a mixed data stream obtained from the TAP test points to obtain the arrival time information of the specific data stream at the corresponding TAP test point. By comparing the sending time and arrival time of the specific data stream at the corresponding TAP test point, the performance indicators of the TSN network are calculated. The traffic generation module and the feature rule matching module can be traffic generation software and feature rule matching software installed on a host computer. The vehicle Ethernet simulator and the passive TAP testing tool are connected to the host computer via network cable, GPIB (General Purpose Interface Bus), or other cables, and are uniformly controlled by the host computer.

[0042] In this embodiment of the invention, the vehicular Ethernet TSN network under test needs to have at least two hops of forwarding or routing links, and the network includes at least a TSN switch, a routing gateway, and multiple background vehicular terminals. These background vehicular terminals include background vehicular terminal 1, background vehicular terminal 2, ..., background vehicular terminal M. The background vehicular terminals can be simulated using a vehicular Ethernet simulator or real vehicular controller nodes. They are mainly used to construct background flows in the vehicular Ethernet TSN network, which can be created by a test program or by native service programs. The application vehicular terminals are simulated using a vehicular Ethernet simulator and are mainly used to construct application flows in the vehicular Ethernet TSN network. These application flows are created by a test program or can also be created by simulating actual service flows. The application flows are traffic actively generated by measurement tools or applications in active network measurements for specific measurement purposes, such as measuring network throughput, latency, packet loss rate, and other performance indicators. The background flows refer to normal service traffic generated by other users or applications in the network, other than application flows. These flows are not for measurement purposes but are natural traffic generated during actual network operation.

[0043] Figure 1 In the test, the vehicle Ethernet TSN network under test includes a TSN switch, a routing gateway, and a background vehicle terminal. The TSN switch is responsible for Layer 2 (data link layer) switching in the network, and the routing gateway is responsible for Layer 3 (network layer) routing in the network. During the test, TAP test points are embedded in the network under test in series as needed, and the vehicle terminal is used as the terminal to access the network under test.

[0044] In one embodiment, the passive TAP test tool uses a copper interface electrical splitter to capture the electrical signals of the mixed data stream in the vehicle Ethernet line and send them to the host computer. The host computer is equipped with a high-speed AD data acquisition card. After sampling and quantization, the electrical signal data is transmitted to the FPGA. The FPGA uses feature rule matching software to perform real-time matching and encoding processing on the electrical signal data. After successful matching, the timestamp of the specific data stream when it flows through the corresponding TAP test point is recorded.

[0045] The traffic generation software is installed on the host computer as application layer software. It can call the Ethernet protocol stack through the socket interface to send a specific data stream to the vehicle Ethernet simulator (e.g., application vehicle terminal 1). The generated data stream is forwarded by the vehicle Ethernet simulator and sent to the TSN network under test. After TSN rectification and forwarding, it reaches the destination vehicle Ethernet simulator (e.g., application vehicle terminal 2). After receiving the specific data stream, application vehicle terminal 2 calculates the network transmission performance such as throughput, packet loss rate, latency, and jitter when the data stream reaches the application vehicle terminal 2 node by comparison and statistics.

[0046] The in-vehicle Ethernet simulator can convert data streams into physical electrical signals of the corresponding in-vehicle Ethernet standard and send them in the TSN network; it can also receive physical electrical signals of the corresponding in-vehicle Ethernet standard from the TSN network and convert them into the received data stream; the in-vehicle Ethernet simulator contains a complete in-vehicle Ethernet encoding and decoding architecture (including RELAY, PHY, MAC, etc.).

[0047] Embodiments of the present invention also provide a method for testing the performance of an in-vehicle Ethernet TSN network, which utilizes the system described in the above embodiments to perform in-vehicle Ethernet TSN network performance testing.

[0048] In the method of this embodiment of the invention, the traffic generation module generates a specific data stream and sends it to one of the multiple application vehicle terminals. The generated data stream is then forwarded by the application vehicle terminal to the TSN network, and then forwarded by the TSN network to another of the multiple application vehicle terminals.

[0049] The feature rule matching module matches known specific data streams with mixed data streams obtained from TAP test points to obtain the time information of the specific data stream arriving at the corresponding TAP test point. By comparing the sending time and arrival time of the specific data stream at the corresponding TAP test point, the performance of the corresponding TSN network is judged. Under the corresponding functions of traffic shaping and forwarding scheduling of TSN network operation, performance tests such as throughput, one-way latency, and latency jitter are carried out.

[0050] Taking 100BASE-T1 vehicular Ethernet as an example, a specific data stream is generated by the traffic generation module, and its bit sequence order is a known condition. After conversion to PAM3 (three-level pulse amplitude modulation) encoding corresponding to 100BASE-T1, the logic state sequence is also a known condition. Here, the known logic state sequence of the specific data stream is set as the target sequence N. The electrical signal detected by the TAP detection point is the superposition of the electrical signals from both the transmitting and receiving nodes, presenting five logic states. Simultaneously, the arrival time of the specific data stream at the TAP detection point is unknown; here, the unknown logic state sequence detected by the TAP is set as the detection sequence M. In the case of... Figure 2 The logical state correspondence between the TAP detection points shown below, where data streams are sent from point A and from point B to point C, is as follows:

[0051]

[0052] Assuming the target sequence N is transmitted from end A to end B, if a -2 state is obtained from end C, it means there is a 100% probability that end A sent a -1 state; if a -1 state is obtained from end C, it means there is a 50% probability that end A sent a -1 state and a 50% probability that end A sent a 0 state; if a 0 state is obtained from end C, it means that the probability of end A sending a -1, 0, or 1 state is 33.3%. Therefore, the correspondence between the A-end transmitting status bit and the C-end coupling status bit can be divided into four levels: ① Full match: when the status bit corresponding to the target sequence N is -1, the status bit of the detection sequence M is -2; when the status bit corresponding to the target sequence N is 1, the status bit of the detection sequence M is 2; ② Half match: when the status bit corresponding to the target sequence N is -1 or 0, the status bit of the detection sequence M is -1; when the status bit corresponding to the target sequence N is 1 or 0, the status bit of the detection sequence M is 1; ③ Compatible state: when the status bit corresponding to the target sequence N is -1, 0, or 1, the status bit of the detection sequence M is 0; ④ Mismatch state: when the status bit corresponding to the target sequence N is -1, the status bit of the detection sequence M is 1 or 2; when the status bit corresponding to the target sequence N is 1, the status bit of the detection sequence M is -2 or 1; when the status bit corresponding to the target sequence N is 0, the status bit of the detection sequence M is -2 or 2.

[0053] Based on the above principle, each corresponding state bit of the target sequence N and the detection sequence M is matched one by one. Firstly, there are no mismatched states in the sequences. Secondly, the more full and partial matches there are, the greater the probability that the detection sequence M represents the target sequence N reaching the TAP test point. The quantification process of the probability of successful matching is as follows:

[0054] I. Basic Condition Verification: Judgment of the Legality of Differences

[0055] First, verify whether the M sequence can be generated by superposition and coupling of the N sequence, that is, whether each corresponding state bit satisfies di=Mi-Ni ( ):

[0056] 1. Calculate the difference sequence: For each position i, calculate di = Mi - Ni.

[0057] 2. Legality filtering: If any... We can directly determine that sequence M cannot be a sequence generated by superposition and coupling of sequences N, with a probability of 0. If all di are valid (i.e....), Then proceed to the next step of probability calculation.

[0058] II. Probability Calculation of Coupled Collisions of Assuming Random Sequences

[0059] If the detection sequence M passes the difference validity check, it means that each corresponding state bit has only three possibilities: full match, half match, and compatible state. At this point, it can be basically determined that sequence M is generated by superimposed coupling of sequences N. However, further analysis reveals two other possibilities: one is that sequence M is generated by superimposed coupling of sequences N, which is called the accurate identification case; the other is that sending random sequences simultaneously from ends A and B also happens to satisfy the difference validity check, which is called random sequence coupling collision. Therefore, if the probability of random sequence coupling collision can be quantified and calculated at this point, the quantified probability of the accurate identification case can be further explained.

[0060] For the case of random sequence coupling collision, the probability of sending -1, 0, and 1 states at both ends A and B is 1 / 3. The probabilities of -2, -1, 0, 1, and 2 after coupling at end C are -2 (1 / 9), -1 (2 / 9), 0 (1 / 3), 1 (2 / 9), and 2 (1 / 9), respectively. Assuming sequence N has n state bits, and after matching, sequence M has a fully matched state bits (including -2 and 2), b partially matched state bits (including -1 and 1), and c compatible state bits (including 0), then the probability of this (i.e., random sequence coupling collision) is:

[0061]

[0062] Where a + b + c = n, c < n / 2 (because PAM3 does not encode consecutive 0s / 0s in the status bit encoding); in PAM3-encoded vehicular Ethernet, every two status bits represent 3 bits, that is, one status bit represents 1.5 bits. Therefore, the probability of a random sequence coupled collision is... At a given time, at least X bits need to be transmitted before a random sequence coupling collision will occur once. X can be calculated using the following formula:

[0063]

[0064] For automotive 100Mbps Ethernet 100BASE-T1, it can transmit a maximum of 100M bits per second. Given that the lifespan requirement for automotive electronic components is 10,000 hours, the maximum number of bits that can be transmitted over its entire lifespan is [not specified]. .

[0065] Therefore, when X > 3.6E+15 bits, the random sequence coupling collision is almost impossible to occur within the life cycle of automotive electronic components, so it is assumed that only accurate identification occurs at this time.

[0066] Depend on The calculation formula shows that the probability weight of the half-matched state bits is the largest, meaning that the more fully matched state bits there are, the higher the probability weight will be. The smaller the value, the better the compatibility, and the worst is the half-match state. That is, in the extreme state (a=c=0), when b=40, n=40, X=7.33588E+15bit>3.60E+15bit. Therefore, for the 100BASE-T1 vehicular Ethernet, selecting a target sequence N with more than 40 status bits (60 consecutive bits) and judging its validity through the difference can be considered as an accurate identification situation.

[0067] Similarly, for the automotive gigabit Ethernet 1000BASE-T1, it can transmit a maximum of 1000M bits per second, and a maximum of 3.6E+16 bits over its entire lifecycle. Therefore, by selecting a target sequence N with more than 42 status bits (63 consecutive bits) and verifying its validity through difference, it can be considered an accurate identification case (i.e., in the extreme state (a=c=0), when b=42, n=40, X=3.89948E+16 bits > 3.60E+15 bits).

[0068] like Figure 3 As shown, the feature rule matching module performs feature rule matching between specific data streams and mixed data streams as follows, until the timestamp data of all data packets flowing through the TAP test points is matched:

[0069] Define multiple target sequences N1, N2, N3... for a specific data stream, where the length of sequence N should be greater than the length defined above. Each sequence N should be extracted from each data packet and contain the sequence number in that data packet. Sequence M is continuously detected through the TAP test point. The difference between sequence M and the target sequence N is used to determine the legality of the difference and to check if they match successfully. If they match successfully, the packet sequence value and the time of arrival at the TAP test point of the data packet are recorded, and the next batch of data packets is tracked. Otherwise, sequence M is shifted backward and the matching continues.

[0070] Determine if all sequences have been traced. If yes, output the data stream time tracing conclusion. Otherwise, check if the receiving end has received all data packets. If not, shift sequence M forward and continue matching. If all data packets have been received, determine if the data packets of the unmatched target sequence have not flowed through the corresponding TAP test point. Record the unmatched and matched data packets at the corresponding TAP test point, output the data stream time tracing conclusion, and record the timestamp of the data packets arriving at the TAP test point.

[0071] The following is passed Figure 4 The specific examples further illustrate the vehicle Ethernet TSN network performance testing method of the present invention.

[0072] In this example, the CBS traffic shaping performance of video streams in a binocular vision scenario is tested. The left and right visual sensors serve as the application vehicle terminals, the LiDAR as the background vehicle terminal, the left and right domain controllers as TSN switches, and the intelligent driving domain controller as the routing gateway. Image data acquired by the left visual sensor is synchronized and traffic shaped by the left domain controller before being transmitted to the intelligent driving domain controller; similarly, image data acquired by the right visual sensor is synchronized and traffic shaped by the right domain controller before being transmitted to the intelligent driving domain controller; the LiDAR generates point cloud data which is synchronized and traffic shaped by the right domain controller before being transmitted to the intelligent driving domain controller. TAP test points are set between the left and right domain controllers, and between the right and intelligent driving domain controllers, to check whether the data from both visual sensors is synchronously transmitted to the intelligent driving domain controller, unaffected by the background stream. The specific test scenario is as follows:

[0073] The left-side vision sensor, right-side vision sensor, and LiDAR are simulated by an in-vehicle Ethernet simulator. The traffic generation software actively generates left-side and right-side vision image data as application streams and injects them into the in-vehicle Ethernet TSN network. The traffic generation software also actively generates LiDAR point cloud data as background streams and injects it into the in-vehicle Ethernet TSN network.

[0074] For each application stream, a corresponding target matching sequence is generated. The target matching sequence is at least 65 bits in size, and each target matching sequence contains at least the packet sequence characteristics and generation timestamp characteristics of the data packets in the corresponding application stream.

[0075] The detection sequence of the corresponding application flow is monitored by TAP test points 1 and 2. TAP test point 1 monitors the application flow sent from the left domain controller to the intelligent driving domain controller, and TAP test point 2 monitors the application flow sent from the right domain controller to the intelligent driving domain controller.

[0076] The correspondence between the target matching sequence and the detection sequence is identified by the feature matching algorithm. The timing information of the application flow when it passes through the TAP test point is tested. The test requires that the arrival time of the left and right visual data at the TAP test point be synchronized. At the same time, the right visual data transmission process is not affected by the background flow. That is, no matter how the background flow changes, the application flow can be accurately identified by feature matching without being affected by the background flow.

[0077]

[0078] In summary, the vehicle-mounted Ethernet TSN network performance testing system and method of this invention combine active network measurement technology with passive TAP testing technology and utilizes a feature rule matching algorithm to accurately perform TSN network performance testing during traffic forwarding and scheduling across multiple nodes, including throughput testing, latency jitter testing, traffic shaping testing, and time synchronization testing, etc., and has the following beneficial effects:

[0079] (1) The passive TAP monitoring method has no impact on the network under test and will not cause additional latency and traffic collision overhead, thus meeting the requirements for conducting non-sensory performance testing of time-sensitive networks.

[0080] (2) Combining active network measurement with passive TAP can achieve seamless testing of TSN networks, ensuring that the test data is authentic and reliable from the source, and guaranteeing the accuracy of the test;

[0081] (3) Through the feature rule matching algorithm, the full-duplex mixed signal in the vehicle Ethernet can be identified and separated, ensuring that the passive TAP test can be applied on the TSN network;

[0082] (4) The active network measurement and feature rule matching algorithm can effectively distinguish and identify the full-duplex communication waveform of a single pair of twisted pairs, meeting the test requirements for data packet sequence and arrival time performance;

[0083] (5) A quantitative evaluation method for feature rule matching algorithm is proposed, which can realize the quantitative evaluation of the recognition matching accuracy and ensure the feasibility of implementing feature rule matching algorithm;

[0084] (6) Based on quantitative evaluation, the feature rule matching algorithm of the present invention achieves high-accuracy identification with a short bit, is easy to implement, and has a wide range of applications;

[0085] (7) A non-intrusive test method and test system for vehicle Ethernet TSN network were proposed, which realized the testing of performance indicators such as time synchronization, traffic shaping, throughput, latency and jitter of TSN network.

[0086] The foregoing has only described certain exemplary embodiments of the present invention by way of illustration. Undoubtedly, those skilled in the art can modify the described embodiments in various ways without departing from the spirit and scope of the present invention. Therefore, the foregoing drawings and descriptions are illustrative in nature and should not be construed as limiting the scope of protection of the claims of the present invention.

Claims

1. A vehicle-mounted Ethernet TSN network performance testing system, wherein the vehicle-mounted Ethernet TSN network uses a single pair of twisted-pair cables for full-duplex communication, characterized in that, Includes multiple in-vehicle Ethernet simulators, passive TAP testing tools, traffic generation modules, and feature rule matching modules; The multiple vehicle Ethernet simulators are used to simulate multiple application vehicle terminals. The vehicle Ethernet simulators can convert data streams into physical electrical signals of the vehicle Ethernet standard and send them in the TSN network, and can also receive physical electrical signals of the vehicle Ethernet standard from the TSN network and convert them into received data streams. The traffic generation module is used to generate a specific data stream and send it to one of the multiple application vehicle terminals. The generated data stream is then forwarded by the application vehicle terminal to the TSN network, and then forwarded by the TSN network to another of the multiple application vehicle terminals. The passive TAP test tool is set at a TAP test point in the TSN network. At the TAP test point, the transmitted and received data streams form a full-duplex mixed data stream. The passive TAP test tool is used to acquire the mixed data stream. The feature rule matching module is used to match a specific data stream with a mixed data stream obtained from the TAP test point to obtain the time information of the specific data stream arriving at the corresponding TAP test point. By comparing the sending time of the specific data stream and the arrival time at the corresponding TAP test point, the performance index of the TSN network is calculated. The feature rule matching module performs matching between specific data streams and mixed data streams as follows: Define multiple target sequences for a specific data stream. The length of each target sequence is greater than a predetermined number of bits. Each target sequence is extracted from each data packet of the specific data stream and contains the sequence number of that data packet. Detection sequences are continuously acquired through TAP test points. The difference between the detection sequence and the target sequence is used to determine the legality of the match and whether the match is successful. Record the sequence number of the successfully matched data packet and the time it arrived at the TAP test point; The following steps are used to determine the validity of the difference: For each position i, calculate the difference sequence di = Mi - Ni between the detection sequence Mi and the target sequence Ni; If there exists any di If {-1,0,1}, then the detection sequence cannot be a sequence generated by superimposing and coupling the target sequence; If all di satisfy di∈{-1,0,1}, then the detection sequence is determined to be generated by superposition and coupling of the target sequence.

2. The system as described in claim 1, characterized in that, The TSN network includes a TSN switch, a routing gateway, and a background vehicle terminal. The background vehicle terminal is simulated by an in-vehicle Ethernet simulator. The traffic generation module is also used to generate a background flow for the background vehicle terminal. The background flow is the natural traffic of the TSN network during actual operation. The feature rule matching module can identify the specific data flow generated by the traffic generation module without being affected by the background flow.

3. The system as described in claim 1 or 2, characterized in that, It also includes a host computer, with the traffic generation module and feature rule matching module set in the host computer. The vehicle Ethernet simulator and passive TAP test tool are connected to the host computer via cables.

4. A method for testing the performance of an in-vehicle Ethernet TSN network, characterized in that, Performing vehicle Ethernet TSN network performance testing using the system described in any one of claims 1-3, including: The traffic generation module generates a specific data stream and sends it to one of the multiple application vehicle terminals. The generated data stream is then forwarded by the application vehicle terminal to the TSN network, and then forwarded by the TSN network to another of the multiple application vehicle terminals. The feature rule matching module matches a specific data stream with a mixed data stream obtained from the TAP test point to obtain the time information of the specific data stream arriving at the corresponding TAP test point. By comparing the sending time and arrival time of the specific data stream at the corresponding TAP test point, the performance index of the TSN network is calculated.

5. The method as described in claim 4, characterized in that, When the vehicle Ethernet is a 100Mbps vehicle Ethernet using PAM3 encoding, the predetermined number of bits is 60 consecutive bits; when the vehicle Ethernet is a gigabit vehicle Ethernet using PAM3 encoding, the predetermined number of bits is 63 consecutive bits.

6. The method as described in claim 4 or 5, characterized in that, Assuming the target sequence has n state bits, and after matching, the detected sequence has 'a' fully matched state bits, 'b' partially matched state bits, and 'c' compatible state bits, where a + b + c = n, then in the case of PAM3 encoding in the vehicle Ethernet, with a predetermined number of bits, the probability ρ of random sequence coupling collision is... , Assuming that at least X bits need to be transmitted before a random sequence coupling collision occurs once, X is calculated as follows: .

7. The method as described in claim 6, characterized in that, A full match means that when the state bit corresponding to the target sequence is -1, the state bit of the detected sequence is -2; when the state bit corresponding to the target sequence is 1, the state bit of the detected sequence is 2. Half-matching means that when the state bit corresponding to the target sequence is -1 or 0, the state bit of the detection sequence is -1; when the state bit corresponding to the target sequence is 1 or 0, the state bit of the detection sequence is 1. The compatibility state refers to the state bit of the detection sequence being 0 when the state bit of the target sequence is -1, 0, or 1.

8. The method as described in claim 4 or 5, characterized in that, The performance metrics include throughput, one-way latency, and latency jitter.

Images