Next Generation In-vehicle Networking (IVN): Enabling Connected and Autonomous Driving

 

Modern vehicles are increasingly becoming more connected and automated with advanced driver assistance systems. However, the traditional in-vehicle networks are proving to be inadequate to support the data and communication requirements of next generation connected and autonomous vehicles. Next generation IVN standards are being developed to overcome the limitations of current networks and enable new autonomous driving technologies.

Introduction to Current IVN Technologies

Currently, most vehicles rely on Controller Area Network (CAN) bus as the main in-vehicle network for connecting electronic control units (ECUs) and transmitting data. CAN bus has served vehicle networking needs well since its introduction but it has some drawbacks. CAN bus has a limited data transmission rate of only around 125 Kbps to 1 Mbps which is insufficient for handling vast amounts of sensor and system status data in connected and autonomous vehicles. It also does not support advanced features like prioritization of real-time critical safety messages.

Other legacy networks like Local Interconnect Network (LIN) and Media Oriented Systems Transport (MOST) have also proved incapable of meeting the requirements of next generation vehicles with their limitations in bandwidth, reliability and security. With new advanced driver assistance systems relying on high resolution cameras, radars, lidars and ultrasound sensors generating massive data streams, the need for high-speed networking has become imperative.

Emergence of Ethernet and TSN Standards

To overcome the bandwidth and real-time communication constraints of CAN and other traditional networks, automakers are looking at Ethernet as the potential replacement for Next Generation In-Vehicle Networking. Ethernet provides gigabit bandwidth and advanced TCP/IP communication capabilities required for sensor fusion, over-the-air updates and infotainment data streaming in connected cars.

However, regular Ethernet is not well-suited for real-time safety critical applications due to variability in latencies. This gave rise to the development of Time Sensitive Networking (TSN) standards which enhance Ethernet with real-time communication features using time awareness and scheduling mechanisms. TSN enables hard real-time guarantees, latency control and security features essential for autonomous driving applications. Prominent TSN standards gaining traction in automotive include IEEE 802.1AS, 802.1Qbv, 802.1Qbb, 802.1Qch, 802.1CB and AVB.

Advent of 5GV and the Impact on IVN

Introduction of 5G vehicular (5GV) cellular technology will also influence next generation IVN architectures. Advanced 5G networks with high bandwidth, low latency and vehicle-to-everything (V2X) communication facilitate integration of vehicles with urban infrastructure and the transport ecosystem. This brings opportunities for updating HD vehicle maps, remote telemetry and over-the-air updates via the cloud.

However, integrating large volumes of 5G and V2X data into vehicles requires networking platforms capable of multi-gigabit throughput, latency control and security hardening. The emergence of technologies like multi-Gbps Ethernet, TSN, Security Network Virtualization and software defined networking principles are aimed at resolving these new challenges posed by 5G connected driving experiences.

Transitioning to Ethernet and TSN Based IVN Platforms

Major automakers have already started transitioning from traditional CAN and MOST networks to Ethernet and TSN based platforms. Premium manufacturers like BMW, Audi and Mercedes introduced partial Ethernet networks over the past years. However, their next gen models will feature comprehensive Ethernet and TSN networks across entire vehicle domains.

Ethernet switches are replacing CAN controllers and aggregating data from numerous gigabit ECUs. Systems like flexible domain/zonal/cluster/central Ethernet architectures along with virtual TSN bridges allow consolidation of data and deterministic forwarding. Cloud connected ECUs running on powerful network processors powered by programmable hardware and APIs accelerate sensor fusion and decision making for autonomous functions.

Challenges in Adopting Next Gen IVN Platforms

While next generation IVN standards promise higher bandwidth and real-time capabilities, their widespread adoption faces some challenges:

- Maintaining backward compatibility with legacy ECUs over extended phase-out timelines of over 10 years creates complex heterogeneous networks.

- Ensuring robust cybersecurity against attacks on safety critical systems connected over high speed networks demands sophisticated security architectures involving firewalls, intrusion prevention and embedded device authentication.

- Developing unified open specifications and software stacks takes time given involvement of multiple standards bodies, automakers and tier 1 suppliers.

- Managing vast amounts of real-time data from advanced sensors and edge computing devices requires higher performance network processors and switches than traditionally used.

- Validating functional safety and providing ASIL D certification for autonomous driving Ethernet and TSN networks demands rigorous testing procedures over automobile lifespan under various failure conditions.

While in-vehicle networking technologies have come a long way from low bandwidth proprietary solutions, next generation autonomous driving pushes performance requirements to new frontiers. Consolidating advanced sensors and computing on high speed real-time Ethernet and TSN networks is crucial to deliver safe assisted and self-driving experiences envisioned for future mobility. Standardization efforts and partnerships across the automotive-tech industry will help overcome current challenges towards enabling transformative innovation in connected vehicles.

 

Get More Insights On This Topic: Next Generation In-vehicle Networking