4D imaging radar is a breakthrough technology that significantly enhances radar performance by extending its capabilities from measuring distance, speed and horizontal azimuth to cover distance (Range), bearing (Azimuth), pitch (Elevation) and relative velocity measurements in any light or weather conditions. In layman's terms, the 4D imaging radar has greatly improved resolution in the horizontal and elevation directions compared to existing conventional millimetre wave radars, enabling it to 'understand' not only the horizontal plane but also the vertical plane, and to truly profile the object detected ahead. This allows the vehicle to determine whether it is driving "under" or "over" an object.
Typical application scenarios
As mentioned earlier, the main feature of 4D imaging radars is their very high angular resolution, which can be as high as 1° azimuth and 2° pitch for forward-facing 4D imaging radars. When radars with this capability are available, the reflected point of a car or person will no longer be just a simple point, but an image consisting of hundreds or even thousands of points, thus showing the entire outline of the object.
Let us first look at a few typical application scenarios associated with imaging radar:

One is if an imaging radar system can recognise when smaller objects (such as pedestrians or motorcyclists) are mixed with or obscured by larger objects (such as trucks or other obstacles) and can determine whether objects in an area are stationary or moving and in which direction they are moving and can provide real-time data with a detection range of more than 300m.

Secondly, in the case of cars driving in tunnels, for example, thanks to improved pitch angle detection, imaging radar can now measure the entire length and width of a tunnel and sense the geometry of objects, providing high-resolution images when other traffic participants, such as trucks, motorbikes and cars, are present in the tunnel. In other words, the car's vision is even more advanced. 

Thirdly, if a car is travelling at 80 km/h on a motorway and a motorbike (a small object with low reflectivity) is coming up from behind at 200 km/h. Unlike cameras and LIDAR, imaging radar can recognise a motorbike at an initial distance and can identify that the two objects are moving at two different speeds.
This means that the imaging radar not only offers multimode functionality, but also extends the currently available L2+ autonomous driving functions such as high-speed cruising and lane change assistance by providing ultra-high resolution images that enable precise environmental mapping and scene perception. And as the autonomous driving level rises to L3 and beyond, 4D imaging radar will be able to perform multiple functions such as map building, localisation, object contouring and classification of the object under test. When combined with cameras, or pattern recognition and machine learning, the imaging radar system will be able to sense the surrounding environment at ±60° FoV (or 100° FoV) at a high resolution of 1° azimuth and 2° elevation, an enhanced 'sensing capability' that is essential for fully automated driving in complex driving environments. This enhanced "sensing capability" is essential for fully automated driving in complex driving environments.

In general, as the number of radar sensors fitted in each vehicle increases with the level of autonomous driving, a single imaging radar that wishes to achieve very high angular resolution requires multiple radars to be cascaded at the RF front end to form a larger array of radar antennas to achieve the imaging capability. As a result, a single forward-facing 4D imaging radar module requires more sensor chips and a more powerful back-end processor.