• Automotive E/E
  • Our Resources
  • Blog
  • Can Hardware Solve Next Gen Architecture’s Major Challenges?
Connected Vehicles

Can Hardware Solve Next Gen Architecture’s Major Challenges?

May 2, 2022

It’s no secret that the automotive industry is shifting to the use of zonal architecture, in which the number of ECUs - and associated wiring - are consolidated. While this is good news for the cost and weight of new automobiles, there are new challenges that arise as current technologies are not equipped to support zonal architecture while keeping up with the next-generation car’s software demands.

Cars will need superfast secure multi-interface routing with high throughput to run smoothly with a large amount of onboard software. While zonal architecture certainly goes far in easing the burden, its design lends itself to timing, routing, and communication challenges. Zonal routers use a combination of microcontrollers (MCUs) and Ethernet switches to relay data from the ECUs to the backbone.

The main challenges include:

  1. Low Latency - as we consolidate the number of ECUs in a vehicle into fewer but more capable High Powered Computes (HPC), all messages must be routed through the HPC backbone resulting in additional hops where the HPC or Zonal Gateway must understand what the message is saying and where it should go next
  2. Multi-interface routing - there are multiple communication interfaces in the vehicle from CAN to Ethernet to LIN to GPIO that need to be routed together and quickly
  3. High throughput - as vehicle capabilities continue to grow, applications require more bandwidth (i.e. cameras, ADAS) and the communication backbone must be able to support a higher level of throughput

In order to minimize and eliminate these bottlenecks, the different messages should be combined into a container that includes multiple data types to optimize the speed at which a gateway can decipher a message and send it forward. These containers reduce both the volume of messages and the time it takes for them to be sent from one HPC to another. 

These challenges like everything else in a software-defined vehicle, need to be addressed with security considerations in mind. Zonal gateways must be able to assess each data packet deterministically and confirm whether it should be communicated to the rest of the vehicle and how best to route that communication. The communication backbone needs to be built securely by design.

The Need for Speed (and Low Latency)

In a zonal architecture, signals must pass through several hops, each one causing a small amount of latency as the ECU or Gateway understands what the message is and where it needs to go. These small amounts add up quickly and the end result is a latency that exceeds the limits set by the system, especially when there are unpredictable fluctuations. 

A solution is needed that can allow for low latency routing decisions while also supporting multi-gigabit traffic and a variety of interfaces. 

In essence, the Backbone HPCs have to become high-performance routers that can securely and independently manage the traffic generated by multiple processes involving numerous ECUs in different zones all at the same time. Traditional methods that involve using an MCU and Ethernet switch work, but they don’t do enough to solve the communication bottlenecks that arise when trying to balance a cost-effective design with a huge amount of interfaces and traffic requirements.


PDU Routing

Vehicles have multiple analog and digital sensors that use various interfaces (i.e. Ethernet, CAN, LIN, GPIO, etc.) for collecting and sharing different types of information. The volume of the messages and the assortment of interfaces create bottlenecks around the gateways as they decipher each different message and interface and push them to the correct destination.

In order to minimize and eliminate these bottlenecks, the different messages should be combined into a container that includes multiple data types to optimize the speed at which a gateway can decipher a message and send it forward. These containers reduce both the volume of messages and the time it takes for them to be sent from one HPC to another. 

They also require less computing resources to decipher each message as it is deciphering bundles of messages at once and can send these bundles to various locations simultaneously.

CommEngine Traditional vs. Single

Demand for Higher Throughput

As the amount of applications required to run a vehicle continues to increase and their complexity rises, the vehicle will need a way to provide the potential for higher throughput with lower latency to support increasingly data-heavy applications. 

There is always going to be a limit to the amount of throughput that will be possible when relying on software to manage data routing. A hardware solution supports higher throughput and lower latency as it can increase the number of routing decisions that are possible simultaneously (or as close as) and provides more bandwidth for a higher throughput allowing programs to operate quickly and simultaneously.

NXP Webinar blog banner

Why is the Need for Low Latency with Higher Throughput so Important?

It is not merely the increase in the number of applications that will power the vehicles of the future, but also the types, and data demands, of these services applications like augmented reality and advanced collision avoidance, for example, require much faster routing times and very low latency. 

In some cases, applications may experience latency with no discernable impact on the user. For example, if the infotainment system takes an extra few microseconds to load a song, it is unlikely that someone will even notice. But when a program that is critical to the safety of the car’s operation (i.e. autonomous driving) has even a tiny delay, the result can be catastrophic. 

As more advanced driver assistance systems (ADAS) are implemented in vehicles, latency, throughput, and processing speed will become crucial issues. As quickly as sensors pick up data, that data must be instantly processed and acted upon to ensure that drivers, passengers, and pedestrians all remain safe as they relinquish control to the car itself.

The GuardKnox Solution

There’s no question that the automotive industry is “going zonal.” But the innovation cannot stop there. Applications and services in vehicles will only continue to grow and develop, eating up more bandwidth and requiring smooth uninterrupted routing. Now is the time to consider solutions to the challenge and ensure that the necessary applications can access the high throughput and low latency routing they need to operate smoothly.

GuardKnox’s F.A.S.T.E.R. CommEngine™ is a single-chip solution that can be used to implement zonal functionality, allowing for ultra-low latency with multi-gigabit bandwidth. It replaces the traditional method of using an MCU and Ethernet switch with one single-chip solution to manage the huge amounts of data that must be routed and processed in next-generation architecture.

The CommEngine™ is secure by design - built with deterministic security in the design as well as providing deep packet inspection - and ensures the right information is allowed while malicious or illegitimate packets are dropped. It addresses the automotive industry’s needs for connectivity and scalability, allowing OEMs to reap the full benefits of zonal architecture.

Lets’s Talk