IEEE Solid-States Circuits Magazine - Spring 2022 - 45

buildings, and eventually, to enable
the mapping of the environment,
while cost-effectiveness and size
reduction complying to car design
will remain essential.
In this article, we take a renewed
look at the design approach that
nowadays has FMCW waveforms
and CMOS transceiver as the epicenter
of distance and velocity detection.
We shift our attention to the MIMO
antenna and transceiver array, pointing
out new means to use optimally
available physical and IC technology
resources for angular resolution and
interference before the physical limits
imposed by the 76-81-GHz band
necessitate, once more, moving to a
new band.
FMCW Radar Resolution Space
Radar sensing uses modulated electromagnetic
pulses to measure distance
and velocity simultaneously.
The transmitted chirped pulses in
Figure 2 are reflected by an object
and received back with a delay proportional
to the time of flight, which
is a measure of distance to the radar.
A chirp is a tone with a frequency f0
Level 1
Driver Assistance
at time t0 that increases linearly to
fB 0 + over a pulse duration of T .chirp
It employs pulse compression [1] to
decouple range resolution (the ability
of the radar to distinguish targets
in range), determined by modulation
bandwidth B, from sensitivity, determined
by the signal-to-noise ratio
(SNR) for a given range as a function
of the pulse emission duration
T ,chirp
emitted power, and receiver
noise. Small frequency variations
introduced by the motion of targets
relative to the radar source known
as
the Doppler effect
[1] provide
velocity estimation using sequential
pulses to measure phase changes of
reflected waves. The wavelength and
pulse repetition rate determine the
maximum unambiguous velocity,
whereas the wavelength and total
measurement time determine the
velocity resolution (the minimum
speed differences a radar can detect).
Table 1 summarizes the main
relationships. Figure 2 also shows
the FMCW transceiver architecture.
A fractional-N phase-locked loop
and the frequency translation circuits
generate the 76-81-GHz chirp.
Level 2
Partial Automation
Level 2+ to 3
Conditional Automation
A modulator sets the chirp phase
before amplification and transmission
via the antenna. Figure 2 shows
a black chirp frame bouncing off a
target. The reflected chirps in red are
received, amplified, and mixed with
the transmitted chirp sequence for
correlation. Echo delays x translate
to beat frequencies (blue) proportionally
to the modulation bandwidth
over the chirp duration as described
in (5) in Table 1. The beat tones are
filtered and amplified prior to digitization
at a Nyquist sampling rate
equal to twice the maximum beat
frequency. Large dynamic range differences
between echoes due to free
space loss in the forward and return
paths (for example, 52 dB for two targets
at 10 and 200 m) combined with
radar cross-sections varying between
−5 and 30 dB (e.g., for debris compared
to a truck) in the presence of
noise require the use of coherent integration
using multiple sequences and
optimal matched filtering in the form
of FFTs.
The range FFT provides quantized
distance information in range bins
operating on NR
Level 4-5
High/Full Automation
samples from each
Seeing
Other
Cars
Seeing Bicycles
and Pedestrians
Seeing
Smaller
Objects
Seeing
Around
the Car
Imaging
Radar
24-24.5 GHz
2D
Speed, Distance
Low
Resolution
76-81 GHz
SiGe
3D
+ Azimuth
Higher
Resolution
Corner
Radar
Multiple
Small Modules
76-81-GHz CMOS + Advanced Processing
4D + Elevation
Resolution Boost
FIGURE 1: Radar resolution expansion in the mm-wave bands.
IEEE SOLID-STATE CIRCUITS MAGAZINE
SPRING 2022
45
Long-Range
Radar
Highest
Resolution and
Performance
Front and Rear
Higher
Performance
How to Expand
Resolution Further?
How Far Before Moving
Again to a New Band?
Which Technologies?
Mapping of the Environment
IEEE Solid-States Circuits Magazine - Spring 2022

Table of Contents for the Digital Edition of IEEE Solid-States Circuits Magazine - Spring 2022
