In my earlier post, I shared the exciting emergence of new possibilities as 5G and automotive technologies merge.
Among the use cases, what interests me most is how to reduce latency in network communications. Currently, the latency on a 4G network varies from 30 - 100 ms, half of that of a 3G network. Developers are now working towards 1 ms latency for 5G – and that is critical to realize the ubiquitous mobile network for highly automated or autonomous vehicles.
So why the need for speed? Low latency doesn’t mean we are speeding up our lives – it’s about minimizing the transit time between end-points in a communications system.
There are any number of use cases where improved latency will benefit future commuters using autonomous vehicles. These include:
- How autonomous vehicles operate and react to potentially dangerous situations such as maneuvers at junctions or merging lanes.
- Collision prevention in the event of a traffic light system failure.
- High-density platooning for space and fuel economy.
As an example, my car is driving along and the car in front of me suddenly puts on the breaks for some reason – whether there is a stalled truck in front of it, an accident ahead or roadworks.
My car will get a message from the car in front of me, or even from two or more cars ahead that can automatically trigger the brakes on my car or take evasive action – fully automatically and safely. This can only be achieved if the time taken for the ‘safety message’ from the cars ahead get to my cars on-board systems in time to take the evasive action! Ultra-low latency will ensure the time-critical information is transferred well in advance of dangerous situations.
In another scenario, current detection technologies depend on cameras or radar and work better for detection within a line or radius of detection. To enable better accident-prevention capabilities, if there is something around the corner, the on-board detection system in my car must be able to broadcast those hidden potential dangers to me. Lower latency means minimizing the time for sending that critical message from one car to the other, and triggering a cohesive series of automated safety measures, like putting on the brakes hard, pre-tensioning the seat belts, to avoid, or prepare for an imminent crash.
Human errors have been documented to account for some 90 percent of all fatal accidents, but each time there is a serious or fatal crash involving a self-driving car, public trust in this technology suffers and the automotive industry are aware of this. That is why consumer confidence in this technology is paramount, and every device and connection must be tested thoroughly before being released to market.
The experienced human driver relies so much on familiarity, habits and instincts without consciously thinking about how to drive, but there is a lot going on in our heads! The car of the not-too-distant future will need to be able to replicate and improve on our driving ability at the same time as perform efficient navigation from A to B.
The key challenges for automotive engineers and designers is to design in enough capability and performance to carry out all of these tasks safely, in comfort and at acceptable speeds (consumers would not be willing to have an ultra-cautious vehicle that will travel in relative safety but can only travel at walking speed!)… all within time and cost budgets.
Processing speed, power and data volume are increasing all the time, with new design and test requirements.
Data from sensors, V2X systems, road conditions and navigations all must be processed concurrently with minimum latency, with on-board systems and artificial intelligence to process the information and predict the possibilities moving forward.
Keysight’s task is to focus on improving automotive wireless connectivity, be it cellular connection or DSRC using 802.11p standards. For 5G cellular V2X, Keysight is working with stakeholders to identify how best to test each channel, measure each connection’s accessibility, latency, message decodes, and overall performance. This is where we can contribute our experience to test and evaluate use cases with our rich experience in the following connected car and radar test technologies:
- 5G Mobile Communication
- Automotive Serial Bus Test
- Wireless Power Transfer
- Wireless Technologies to Connect Cars
- Connectivity and RF Simulation Test
- Automotive Radar
If you have any insights to share with our readers on your best practices for latency reduction use cases, I welcome you to share your comments here.