November 02, 2023

Share:

The Revolution Driving In-Vehicle Networking

Modern vehicles are rapidly becoming more sophisticated, with countless sensors, processors and actuators being deployed to enhance functionality, safety, and efficiency. As automotive technology has evolved, the demand for a vehicle’s networking system has increased, with higher bandwidths and lower latency communication needed to maintain functionality and safety.

 

Over the years, several protocols have been developed for use primarily (or exclusively) in vehicles

 

A number of communication protocols have been developed for in-vehicle use
A number of communication protocols have been developed for in-vehicle use

 

While each protocol has unique attributes, they each struggle to address the needs of modern vehicles due to changing architectures and the sheer amount of data being transferred within the in-vehicle network (IVN). As a result, automakers are looking for new solutions to provide the necessary performance and bandwidth.

 

Automotive Ethernet

Ethernet was an obvious choice because of its widespread adoption in computing, relatively high bandwidth, and reasonable cost. However, one significant drawback for automotive use was the inability to operate in a time-sensitive or deterministic mode. This was due to the carrier sense multiple access with collision detection (CSMA/CD) inherent in Ethernet’s operation.

To enable the automotive industry to leverage the benefits of Ethernet, a new protocol was developed. Known as 10BASE-T1S, this automotive-specific variant replaces CSMA/CD with physical layer collision avoidance (PLCA) to allow for the deterministic operation that is critical to drive-by-wire use and advanced driver assistance system (ADAS) applications.

Automotive Ethernet was increasingly used for in-vehicle infotainment systems as well as ADAS system thanks to high bandwidth and low latency. Ethernet plays a major role in enabling connectivity features, such as vehicle-to-vehicle(V2V) and vehicle-to-infrastructure (V2I) communication, which are vital features for enhanced safety management.

As vehicles become more data-driven, there will be a growing need for higher bandwidth to support advanced features like autonomous driving, high- definition/ 4K video streaming, and augmented reality applications. Ethernet networks will need to provide even faster data transfer rates. And future automotive Ethernet networks must offer ultra-low latency to enable rapid decision-making and response times for autonomous driving.

 

Vehicle Architecture and Imaging Solutions

The organization and interconnection of the different sub-systems within a vehicle are crucial considerations for automakers. Typically, the sub-systems are organized by function (e.g., drivetrain, chassis, comfort) regardless of their placement in the vehicle. This results in significant amounts of cabling, which increases the vehicle’s cost and weight.

Recently, the preferred approach is to organize the sub-systems into “zones” depending on where they are in the vehicle. This results in clusters of non-similar functions that are often designated by vehicle locations, for example, “front right,” “rear left,” or similar. Zonal architecture is both scalable and flexible, allowing changes such as removing, adding, or upgrading sub-systems to be performed relatively easily. It also allows for the deployment of redundant and fault-tolerant elements that are essential to achieving the functional safety levels required for critical systems.

While its design alleviates the need for cabling, it significantly increases the data to be transmitted on the IVN backbone, requiring more bandwidth and performance as well as low latency. For ADAS functions, such as automatic emergency braking (AEB), sensors and control electronics will be distributed across the vehicle, and safe and reliable system operation is dependent on time-sensitive networking (TSN) to remove any latency discrepancies between. Even systems such as in-cabin noise canceling that rely on multiple microphones also require TSN to operate correctly.

Undoubtedly, deterministic (10BASE-T1S) Ethernet will play a significant role in future vehicles, especially for the backbone of a zonal architecture. Protocols such as MOST and FlexRay are unlikely to be used in new designs, but it is expected that LIN and CAN will continue to play a role, especially within the various “zones.”

There will also continue to be further development of other protocols, including camera serial interface 2 (CSI-2) and display serial interface (DSI-2), from the MIPI Alliance that are essential to ADAS and infotainment systems to connect high-resolution cameras, sensors and displays in modern vehicles. Furthermore, there is a move towards a standardized SerDes solution under development by the MIPI Alliance and Automotive SerDes Alliance (ASA). As part of this collaboration, research is also being conducted on improving the security of MIPI protocols and implementing asymmetrical ethernet for cameras, which involves transmitting at high bandwidths and receiving lower ones. But the most significant architecture change is that CAN will no longer be the default protocol for the main vehicle communication’s backbone – with Ethernet assuming that role.

 

onsemi and IVN

onsemi has been heavily involved in IVN for over 30 years, with a broad product portfolio backed by excellent customer and application support. Our product portfolio covers all the main existing IVN technologies such as LIN, CAN and FlexRay, while we continue to provide enhanced IP to exceed automotive requirements and demands.

With the importance of 10BASE-T1S ethernet to the automotive industry continuing to increase, much of our development resources will be concentrated here. Following up on recently released solutions, the second-generation products with elevated performance are already in development to help support the industry’s continued push towards zonal architectures and automated driving.

Featured

TND6015/D
Technical Paper
Basics of In-Vehicle Networking (IVN) onsemi products
ADAS

ADAS, or Advanced Driver Assistance Systems, are transforming modern vehicles into safer, smarter, and increasingly semi-autonomous platforms. These automotive safety systems help reduce driver workload, improve reaction time, and support accident prevention by continuously monitoring the vehicle’s surroundings. Common ADAS applications include park assist, surround-view cameras, rearview cameras, lane departure warning, lane keeping assist, automatic emergency braking, adaptive cruise control, blind spot monitoring, traffic sign recognition, auto-pilot functions, and collision avoidance. As vehicle automation continues to evolve, these features are becoming essential for improving road safety and enabling the path toward higher levels of autonomous driving.

Automotive ADAS systems rely on multiple sensing technologies to detect, classify, and respond to real-world driving conditions. Image sensors provide the visual intelligence needed to identify lanes, vehicles, pedestrians, cyclists, road markings, traffic lights, and traffic signs, while LiDAR, radar, and ultrasonic sensors add depth, distance, speed, and short-range object detection capabilities. Hyperlux™ automotive image sensor technology is designed for high-performance ADAS camera systems, offering high-resolution imaging, high dynamic range HDR, strong low-light sensitivity, low-noise performance, and LED Flicker Mitigation. These features help improve camera visibility and machine vision accuracy in challenging environments such as direct sunlight, tunnel entry and exit, nighttime driving, shadows, bright headlights, reflective surfaces, and flickering LED traffic signals or digital road signs.

Ultrasonic sensors play a key role in short-range automotive sensing, especially for park assist, automated parking, curb detection, rear obstacle detection, side obstacle detection, and low-speed collision avoidance. By using ultrasonic time-of-flight measurement, these sensors detect nearby objects around the vehicle even in dark garages, tight parking spaces, and low-visibility environments. The data from Hyperlux™ image sensors, ultrasonic sensors, LiDAR, radar, and vehicle control inputs is processed by a machine vision processor or ADAS system controller, which enables sensor fusion, object detection, environmental awareness, and real-time safety decisions such as driver alerts, braking, or steering assistance. For safety-critical automotive designs, ISO 26262 and ASIL-capable components help support functional safety, diagnostics, fault detection, and reliable ADAS operation.