In-vehicle 10GBASE-T1 Ethernet Intelligent Test Solution

High-speed in-vehicle networks increasingly rely on 10 Gigabit Ethernet communications with bandwidths up to 10 Gbit/s. However, high-bandwidth networks will inevitably push the performance of test systems to their limits. How can this challenge be overcome with existing software and hardware test tools? This article will introduce you to the implementation solution.

Nowadays, the vehicle-mounted high-performance computing controller HPC, ADAS sensors and infotainment systems on the vehicle platform need to interact with a large amount of data in real time, and the 100BASE-T1 or 1000BASE-T1 vehicle Ethernet can no longer meet the bandwidth required for vehicle network communication. 10G-T1, based on the IEEE 802.3ch specification, has a transmission rate of up to 10 Gbit/s and will play an increasingly important role in high-speed network data transmission, such as for transmitting high-resolution sensor and camera signals, as well as high-performance backbone network communications.

01

Typical Ethernet test environment

Whether analyzing, simulating, testing or recording in-vehicle Ethernet communications, physical access to the network is required. However, as bandwidth reaches the 10 Gigabit range, the uplink from the network interface hardware to the PC quickly reaches its technical limit. How to use the standard interface of the test PC to effectively test 10 Gigabit data and achieve specific simulation and analysis requirements requires a carefully designed test system framework.

Typical Ethernet test tasks:

>

"Residual" bus simulation for automotive Ethernet communication

>

Supports stimulation and error state injection of data at each communication layer

>

Record all data traffic on the communication network

Figure 1 Example of vehicle network topology

Figure 1 shows an example system of an in-vehicle network topology consisting of multiple ECUs with switches and 100BASE-T1 or 1000BASE-T1 Ethernet connections. Each ECU sends a large amount of data, which is transmitted over the network and aggregated and forwarded by the switch, resulting in a high uplink load. In this example, the data rate totals more than 2.5 Gbit/s, which exceeds the bandwidth of the PC's USB 2.0 or 1000BASE-T interface.

The problem can be solved by using a higher bandwidth PC interface, such as USB 3.0 or 10GBASE-T. However, if the sensors (SENL and SENR) in the network topology are replaced with sensors that can send high-resolution sensor data at a higher rate, the USB 3.0 or 10GBASE-T interface will also reach a transmission bottleneck. Therefore, it is necessary to build an expandable test system based on the application scenario and communication rate to conveniently and effectively expand the data transmission from the network interface hardware to the PC.

02

Ethernet Test System Requirements

When building a test system, you must consider the overall test framework in layers and clarify the system's requirements for multiple network interfaces. Key points of the test system:

>

Network hardware interface cards can be interconnected and easily expand the number of ports

>

Time-synchronized plug-and-play communication interface card ensures consistent time base for data acquisition on all ports

>

The collected data can be distributed and transmitted to different host computer systems in parallel

>

Integrate real-time hardware filters to filter required data on demand

Actual in-vehicle Ethernet analysis and testing first requires examining the system network itself. In the network topology of Figure 1, there are 7 communication links, among which the HP2 ECU is the "hotspot" with the highest data rate. The network hardware interface card is the link between the network system and the test tool. In addition to selecting the number of channels and supported PHY technologies as needed, the combination of different hardware interface cards and the form of data recording should also be considered, such as through a PC or a data recorder. This places many requirements on the hardware interface card. Whether based on the network under test or the test host computer, its interface must be flexible enough to meet all requirements and effectively solve the actual test tasks.

03

Typical Ethernet test environment

Figure 2: Scalable test solution based on multiple hardware interface cards, data recorders and test host computers

Figure 2 shows the test solution for the above in-vehicle network topology example. Each link in the network transmits data to the test PC or data recorder through the corresponding Uplink. In this example, the 1000BASE-T1 link is forwarded to the data recorder (such as VP7500) at 3.2 Gbit/s through the network hardware interface card. While collecting and recording data, the test data can be forwarded to other network links in parallel.

Figure 3: Scalable test system solution for 2.5G/5G/10G-T1

If the sensors in the system of Figure 1 are replaced with sensors that send several gigabits of data, the configuration must be expanded in a different way (see Figure 3). Using multiple Ethernet interface cards and recorders that support MultiGBASE-T1 PHY technology, the network simulation can be performed using CANoe on the PC while recording the bus data. Since the data traffic in such a network cannot be accurately estimated, the highest communication rate from the uplink to the recorder is as high as possible to avoid data loss. The data in the example system is as high as 6.4Gbit/s. If the bandwidth of the uplink is not enough, a second uplink needs to be added. The core of this test system construction is to consider how to expand the network channel according to actual needs and effectively allocate it to different hardware, while using data filtering mechanisms to classify and assign it to specific uplinks, such as using specific fields of the data frame or communication protocols to filter triggers and forward relevant data to the corresponding Uplink interface. This avoids duplicate data storage and improves the operating efficiency of the system.

04

Automotive Ethernet bus simulation in test systems

The remaining bus simulation test based on the in-vehicle 10 Gigabit Ethernet can simulate all signal interactions on the vehicle bus or stimulate the needs of specific test conditions. In the above example, if you want to test the PDU data flow of the Layer2 switch and the HP2 node, you can use CANoe as a simulation and test tool to simulate the entire network system, or send specific PDU data frames for comprehensive testing of special content. If the system communication uses the ARXML database of the AUTOSAR specification, the simulation can be realized quickly and the data flow of each layer can be observed in detail. However, the remaining bus simulation with large bandwidth and multiple channels requires a high-performance processor environment. If a single test PC cannot meet the performance requirements, a distributed simulation solution can be adopted, that is, the bus simulation is distributed on multiple PCs. CANoe provides the "MultiCANoe" function to meet this application scenario.

Figure 4 Multi-CANoe distributed test system

05

Re-HiL/Replay Test of Ethernet System

Figure 5 Ethernet backflow test example

The replay test for the ADAS system is an important test method for verifying the system under test. As shown in Figure 1, there is a large amount of sensor data in the network system. In the actual bus simulation, there is almost no control data simulation involved. Therefore, in order to reduce the complexity of the system, the actual recorded sensor data can be directly used for playback testing.

06

Is Ethernet data verification online or offline?

Finally, the type of analysis of the test results must also be considered: must it be done online or can offline analysis be used? The final decision will have a great impact on the entire test process. Usually, the results are verified online through the CANoe software tool, and the test results can be obtained immediately and do not require a large amount of data storage. However, offline verification is also an important means for verifying in-vehicle Ethernet such as switches. At this time, the stored data is used, combined with the necessary remaining bus simulation, to build an offline test system.

07

CANoe Option Ethernet and 10G-T1S

Bus interface card VN5650

Figure 6 CANoe.Ethernet is equipped with a variety of Ethernet devices to meet various test requirements

The automotive industry is facing the challenge of increasing communication rates in vehicle networks, among which 10Gbit/s 10GbE is becoming increasingly important for high-speed data transmission. Strategically splitting test tasks and adopting a scalable test system architecture can meet the data acquisition, analysis, simulation and testing of high-speed networks. Vector provides users with a variety of interface cards to meet the integrated test environment of different vehicle Ethernet rates (10BASE-T1S, 100BASE-T1, 1000BASE-T1, 2.5/5/10G BASE-T1) and other buses. CANoe Option Ethernet provides simulation, development, testing and diagnostic functions for ECUs with Ethernet communication, providing database creation, encryption/decryption of Ethernet communication, and visual hierarchical analysis. Based on the embedded test and simulation API, complex test scenarios can be developed through CAPL, C# and Python. VN5650 can be equipped with 1~3 10GbE PHYs, each PHY provides 2 10GbE ports, and the cables are available in three types: loose wire, H-MTD Zj and H-MTD Zp.

Figure 7 VN5650 10G PHY

- VNmodule60 2AE10G BCM89890