RADAR Systems for ADAS (Advanced Driver Assistance System)

Introduction

RADAR remains the technology of choice for detection, tracking and, more recently
classification of un-cooperative targets, over huge volumes of space and over the widest
possible range of environmental and operational conditions.Radar is the primary sensor
used to detect the vehicle's surroundings, along with the front-facing cameras.
For fully autonomous driving vehicles with the security level 4 or 5 we have to the use
RADAR, Camera and LIDAR systems all three together. As per the gradation of automated
driving by SAE Federal Highway Research Institute, we have 6 levels, from level-0 to level-5
as stated below.
Level 0: Driver only: the human driver controls everything independently, steering, throttle,
brakes, etc.
Level 1: Assisted driving: assistance systems help during vehicle operation (Cruise Control,
ACC).
Level 2: Partial automation: the operator must monitor the system at all times. At least one
system, such as cruise control and lane centring, is fully automated.
Level 3: Conditional automation: the operator monitors the system and can intervene when
necessary. Safety-critical functions, under certain circumstances, are shifted to the vehicle.
Level 4: High automation: there is no monitoring by the driver required. Vehicles are
designed to operate safety-critical functions and monitor road conditions for an entire trip.
However, the functions do not cover all every driving scenario and are limited to the
operational design of the vehicle.
Level 5: Full automation: operator-free driving.
RADAR Systems for ADAS (Advanced Driver Assistance System)
ADAS (Advanced Driver Assistance Systems) require several radar sensors for overall
function of autonomous driving. Current radar systems are either based on 24 GHz or 77
GHz. The advantages of the 77 GHz lie mainly in the higher accuracy for distance and speed
measurements as well as in the more precise angular resolution.
Further advantages over 24 GHz are the smaller antenna size as well as the lower
interference problem. The main differences lie between SRR (Short-range radar) applications
and MRR/LRR (mid-range radar, long-range radar) applications.
Short-range radio applications include:
a) Blind Spot Detection (Blind Spot Monitoring)
b) The lane and the lane-change assistant
c) Rear end radar for collision warning or collision avoidance
d) Park Assist
e) Cross-traffic monitoring
Examples for MRR/LRR applications are
a) Brake Assist
b) Emergency braking
c) Automatic distance control
The SSR (Solid State Radar) applications are essentially designed to replace ultrasonic
sensors and to support highly automated driving.
Sensors are placed at each corner of the vehicle, and a forward-looking sensor for long range
detection is positioned on the front of the vehicle. For a ‘cocoon’ radar system, extra sensors
are placed on each side mid-body. Ideally, these radar sensors would use the 79-GHz
frequency band with a 4-GHZ bandwidth; however, global frequency specifications so far
allow only 1 GHZ bandwidth at 77 GHz.
Today, a common partitioning for a radar MMIC (monolithic microwave integrated circuit)
are three transmission channels (TX) and four receive channel (RX) to be monolithically
integrated. We can either integrate base band processing in the MMIC or we can simply use
the raw data radar sensor. The difference is that the output of the baseband processor
provides so called pre-targets, which means it is pre-processed data such as unverified
information on speed, distance, signal strength, horizontal angle, and vertical angle for each
detected object. The raw data radar sensor provides unfiltered raw data, which the ECU
then processes.

In this case, the base band is integrated with the radar process controller. The radar sensor
provides unfiltered raw data to the process controller. This approach has several
advantages. First, the integration of base bands within the process controller saves silicon
surface space and associated costs. The reason is that a relatively simple CMOS-fine-line can
be used rather than a specially optimized technology for RF applications.
5
A further advantage is the relocation of the power loss from the radar sensor into the control
unit. Since the controller has much more space than the radar sensor, it is easier to control
the power loss at this point. Finally, because there is no loss of data by filtering or
compression, the ability to access the radar sensor’s unfiltered raw data provides more
possibilities for signal processing and flexibility. Even the required data rate for such a raw
data radar sensor is not a problem since data can be transferred using a MIPI CSI-2
communication interface.
This interface is already in use today, for example, in video surround systems. This
architecture fits in very well with the raw data radar sensor, because the interface includes
four data lines, which fit the four receiver outputs, each with 12 bits of the radar MMIC. The
bandwidth of the communication interface is also well suited with 1 to 1.5 Gbit/s. This
partitioning of the radar sensor then simplifies the data fusion of the video and radar data
and future LIDAR data since the same communication interface can be used.
A prerequisite for the development of MMICs are dedicated high-frequency (HF)
technologies to realize the frequencies (24 GHz or 77 GHz) and the corresponding output
power. Today, SiGe hetero bipolar transistors are already in use for the HF part, and
monolithic 130-nm CMOS processors are used for logic integration. The development of the
77/79-GHz baseband uses the new BiCMOS9MW technology with a minimum 130 Nm
CMOS structure width.
For future radar systems with higher frequencies such as 122 GHz, B55-technology is
developed. This technology provides SiGe hetero bipolar transistors with a transit frequency
of more than 320 GHz and makes it possible to integrate the corresponding CMOS digital
logic in 55 nm. The A770 is a monolithic integrated transceiver that includes three
transmitters, four receivers, a configurable ramp generator, an integrated ADC (Analog to
Digital Converter), and a MIPI CSI II Interface. This block can be cascaded even more
transmit and receive channels as needed for the application. The block is in a ceramic EWLB
housing with a size of 9 mm x 9 mm. The A770 is suitable for medium and long-range
applications.
Currently under feasibility study for future radar sensors is a monolithic SoC integrating
both the radar and baseband functions. As described before, the advantages of a raw
data radar sensor in such a highly integrated solution are being studied in great detail as
well as possible disadvantages, using market studies looking at commercial considerations.
The following objectives can be achieved using the RADAR based ADAS:
1) Mission planning: The vehicle is able to automatically generate a route from the user’s
current location to his desired destination. There can be penalties for low speed limit,
number of traffic light / stop signs, number of lane changes, etc. to find the path with the
soonest estimated time of arrival. After the destination is chosen, autonomous driving can be
enabled by pushing the auto/manual switch button.
2) On-road driving: On the road, the vehicle is able to perform lane-keeping and distancekeeping
to slower traffic. It can perform a lane change if the vehicle needs to merge
into another lane to exit the road or if the neighboring lane is faster. The autonomous vehicle
is also able to avoid on-road obstacles by either stopping if the road is blocked
or circumventing if permitted.
6
3) Work zone behavior: The autonomous vehicle can detect work zones. It notifies the
human driver to be ready to take over and also slows down to the work zone speed limit.
The vehicle can also follow a shifted or alternate lane indicated by the work zone
channelizer.
4) Pedestrian & Bicyclist Handling: The vehicle is able to detect and stop for pedestrians
crossing the road. If there is a bicyclist in front of the car, it will keep a safe distance and look
for a safe opportunity to perform a lane change or circumvent.
5) Intersections Handling: The vehicle is able to traverse intersections with both stop signs
and traffic lights. For instance, for a four-way-stop intersection, it can estimate precedence
and decide when it is its turn to proceed.
6) Intersection Traversal using Vehicular Networks: The vehicle is capable of
communicating with other vehicles using DSRC, and it can cross an intersection with
minimal stopping in two different ways: 1) using V2I (Vehicle-To- Infrastructure) or 2) V2V
(Vehicle-To-Vehicle) communication.
The vehicle will use V2I communication to cross an intersection governed by a traffic light
equipped with DSRC and V2V communication to cross a four-way-stop intersection. This
can provide higher traffic throughput and shorter delay.