Developers of connected autonomous vehicles have many decisions to make about the physical design of the vehicle navigation system. One key consideration is the design and placement of the GPS/GNSS antenna, as well as antennas for other types of radio signals, like WiFi and 5G.
It’s critical because an autonomous vehicle needs to be able to receive continuous, accurate and reliable signals in order to calculate its own position on the road and in relation to other vehicles and objects. A connected vehicle must also be able to communicate its exact position and point in time to other vehicles (V2V) and intelligent transportation infrastructure (V2I).
Different designs must be tested, in different places on the vehicle, to understand how well the antenna receives signals and how changes in the driving environment might affect reception.
Antennas must be tested in realistic conditions – but real-world testing is impractical
Real-world drive testing is an expensive and inefficient way to gain an accurate understanding of how each antenna model, design and placement performs. As we’ve noted previously, fully autonomous vehicles need to rack up hundreds of millions of driving miles in a wide range of conditions in order to truly understand how they respond to the environment around them.
Simulation makes the job of racking up those miles faster and easier – and safer, since the car can be subjected to scenarios that would be risky to create in real life. But while receivers can be tested using conducted signals, when it comes to testing the antenna, the simulated RF signals – whether from GNSS satellites, cell towers or WiFi points – must be radiated to the vehicle over the air (OTA).
That has to be done without contaminating the real-world RF environment, since the presence of “false” signals can seriously affect GNSS-dependent equipment.
Historically, GNSS OTA testing has been done – mostly by the military – on open test ranges in remote locations, making it a highly expensive and logistically complex activity that requires many months to organise, set up and obtain the appropriate permissions (since interfering with live GPS signals is an illegal activity in most countries).
And even when the military has all the right permissions, and chosen a remote location for the test, there can still be serious repercussions for civilian and commercial users. The emerging autonomous vehicle industry needs a more accessible and practical solution.
The solution: over-the-air simulation in an anechoic chamber
One solution is to use a dedicated anechoic chamber (with appropriate licenses and regulatory approvals) and a GNSS, WiFi and/or 5G/LTE signal simulator to radiate signals without the risk of them leaking into the real world.
R&D organisations like WMG at Warwick University are investing in advanced simulation chambers for autonomous vehicle testing, allowing emulated signals to be safely radiated over the air as part of a complete simulation environment.
Over-the-air simulation in a chamber presents many benefits:
With full control over the signal environment, researchers can create any combination of signal conditions to test the vehicle in any scenario.
Signal power levels can be adjusted to understand how the antenna handles very low-power or very high-power signals, and at which point it becomes incapable of receiving a signal.
Different directions and angles of arrival can be simulated, to understand how the antenna copes with
multipath and obscuration effects
and accidental or deliberate spoofing of the signal.
Patterns of RF interference can be introduced, to evaluate how the antenna responds to factors like jamming, atmospheric effects and noise from other components.
Multiple positioning signals (GNSS, WiFi, cellular) can be simulated at the same time, to inform the design of hybrid positioning systems and indoor positioning systems.
With a test chamber, accurate calibration is essential
A properly set up anechoic chamber will be an essential element of the test environment for the coming generations of connected and autonomous vehicles.
It is possible to realistically emulate live signals from space-based satellites in a chamber, but to do so requires significant expertise in calibrating the power levels of the simulators to mimic signals from space, and accurately simulating the movement of the satellites.
The setup and calibration effort intensifies with the addition of other simulated RF signals – such as WiFi and cellular – in the same chamber. For all these reasons, enlisting a partner with specialist OTA GNSS simulation expertise is a sensible decision.
Find out how Spirent can help
Spirent draws on 30 years of experience in military and commercial testing to offer a complete solution for over-the-air simulation of GNSS signals in an anechoic chamber environment.
To learn more about how Spirent can help with over the air signal simulation for autonomous vehicle antenna testing, contact us.
Read Blog #5 called "Testing RTK and PPP for precise positioning in connected and autonomous vehicles"