SPRY353 May   2026 AWR2188 , LMH13000

 

  1.   1
  2.   Overview
  3.   At a glance
  4.   Introduction
  5.   The forces at play
  6.   Perception: Understanding the environment
    1.     Sensor considerations for ADAS design
    2.     Applying sensor fusion to accurately discern objects
    3.     Computation: Turning perception data into driving decisions
    4.     Communication: The backbone of the vehicle
    5.     Response: Turning data into actions
    6.     The hidden enablers: IC technologies that make it all possible
  7.   Conclusion: The road to higher autonomy
  8.   About the author

Communication: The backbone of the vehicle

As important as reliable sensing and high-performance computation are, without a network of high-bandwidth, low-latency and long-distance communication systems, these functions remain isolated. As shown in Figure 3, a range of communication protocols, each tailored to specific data input types and support mission requirements, connect systems across the vehicle.

High-speed, data-dense applications such as cameras require cable lengths of >10m to reach the computing unit. Serializers and deserializers (SerDes) provide an asymmetric pipeline for data, with forward channel speeds reaching >7Gbps and employing low-speed, back-channel protocols such as serial peripheral interface (SPI) or inter-integrated circuit (I2C) for bidirectional communications.

While SerDes technology has traditionally only been used in camera applications, the increasing evolution of radar and lidar system technologies has led to streaming architectures that send raw data directly to the central computing unit, which also requires the high-speed performance of SerDes. As these applications expand, time synchronicity becomes essential to prevent radar or lidar interference and to enable higher-resolution cascaded radar configurations that rely on very tight time coupling between sensors. For a specific example, see the Single-Chip mmWave Streaming Radar Reference Design for Automotive Front Radar.

 Diagram of
                                        various networking protocols and bandwidths to support
                                        different vehicle architectures Figure 3 Diagram of various networking protocols and bandwidths to support different vehicle architectures

In addition to SerDes, Automotive Ethernet has emerged as a reliable protocol for in-vehicle networking. Ethernet connects domain controllers and enables distributed processing and zone architectures. For ADAS and infotainment applications requiring separate compute units, Ethernet provides a long-distance bidirectional signal chain with integrated security features such as media access control security (MACsec) to meet growing cybersecurity requirements.

Ethernet also supports tight time synchronization, keeping data and decision-making consistent throughout the vehicle. In zone architectures, Ethernet aggregates multiple data sources such as corner radar, seat motor controls and audio signals, along with legacy protocols such as controller area network (CAN) and local interconnect network (LIN), into a single output to the centralized compute unit, reducing cabling costs and complexity. Standards such as 10BASE-T1S extend Ethernet connectivity to lower-bandwidth end nodes within the zone, further simplifying wiring and reducing cost. This zone networking architecture directly enables software-defined vehicles where over-the-air (OTA) updates push new features and capabilities to the vehicle post-production, reducing the need for costly hardware revisions. To learn more about this topic, see the article, "How Ethernet Accelerates the Move to Software-Defined Vehicles."

Although zone architecture and streaming ADAS sensors drive the need for Ethernet and SerDes, CAN and LIN remain essential for control systems throughout the vehicle, including seat motors, post-processed radar data, and lighting. For high-performance applications such as ADAS sensing or powertrains, CAN Flexible Data-Rate (CAN FD) offers a multicommander topology for bidirectional commands. And for simpler applications such as heating, ventilation and air conditioning, and seat motors, CAN FD Light provides a fixed bit rate and a single commander topology, reducing costs and complexity at the end nodes and zone controllers.

As vehicle autonomy increases and architectures evolve, designers need flexible networking options that support precise time synchronization, functional safety and cybersecurity protections.