February 26, 2020
Automotive radar systems operate in complex and dynamic environments that can have a significant impact on their performance. In particular, the material properties of the vehicle bumpers and location of the radar behind or within the bumper can negatively impact operation. An example is that insertion loss or phase change of the material in front of a car radar can influence the radar’s detection range.
Engineers designing these radars in systems such as collision avoidance and adaptive cruise control (ACC) must determine the impact materials in front of the radar will have on the radar detection range. The test tool of choice for most engineers is a vector network analyzer (VNA) to conduct transmission and phase measurements. In today’s labs, the VNA system must not only be accurate but make measurements as efficiently as possible.
Radar System Categories
Radar systems are the technology of choice for collision avoidance because of their inherent advantages. They operate in all-weather conditions, conduct better measurements, detect more objects, and are easily incorporated into the car design. Radar systems typically fall into three categories: short-range radar (SRR), ultra-wideband SRR (UWB SRR), and long-range radar (LRR) (Figure 1).
Originally operating in the 21.65 GHz to 26.65 GHz frequency range, these systems have transitioned to the 77 GHz to 81 GHz band. This creates new issues with mounting the SRR and UWB SRR sensors on the car chassis. Materials and coatings of radomes, emblems, and bumper skins can also interfere with the radar systems.
Car Radar Technology
Nearly all ACC systems use a 77 GHz LRR system that is typically mounted behind the car emblem. Transmit and receive patch antennas of the radar are focused by a dielectric lens and operate at a 4 mm wavelength range. Because the radar beam looks through the car emblem, the reflected signal from the target is exposed twice to the influence of the radome.
The advantage of a frequency-modulated continuous wave (FMCW) radar is that it measures the distance of other vehicles, as well as directly measures the vehicles’ speeds. Based on design principles, the range of a car radar can cover up to 200 m.
When using an FMCW radar for distance measurements, the output signal frequency continuously changes over the transmission time. Plus, it is linearly modulated to reach a maximum of 81 GHz over a given time period. This waveform is called a chirp. A frame consists of numerous chirps, each lasting for a given chirp time, known as TChirp. The bandwidth and slope of each chirp is also crucial to the performance of the FMCW radar. These parameters have a direct influence on the maximum range and velocity and their corresponding resolutions.
An example of these factors is shown in Figure 2. A 76.15 GHz radar wave starts at 0 ms and 30 ms later a reflection is received that is influenced by the distance to the target and its speed (the frequency shifts are depicted by the red arrows for ∆f1 and ∆f2). The comparison between the transmitted signal generated by the radar to the reflected signal by the object gives an indication of the distance. The transmitted and received signals are mixed into an intermediate frequency signal where a Fourier transfer range (FFT) on this intermediate frequency signal reveals information on the distance of the object with high accuracy. Based on this, the frequency difference is a direct measure of the distance.
A second effect has to be included in the calculation – the relative speed of the object. In Figure 2, the emitted frequency that is received is reduced, indicating that the object is moving away from the radar.
By selecting a triangle edge (i.e., increasing and decreasing frequencies), the distance and the velocity of an object after one cycle (in this case 200 ms) can be determined. The radar speed measurement considers the Doppler effect, which states that there will be an increase or decrease in frequency wave as the vehicle moves towards or away from an object. In the figure, the emitted frequency received is reduced, indicating that the object is moving away from the radar.
A New VNA Configuration
VNAs for mmWave applications such as these have been historically large, heavy, complicated, and very expensive. This scenario is less than ideal for automotive radar designs. Fortunately, a new approach that shows significant advantages for these applications can be seen with the Anritsu ShockLine™ MS46522B VNA, which can be configured as a dedicated E-band VNA. The ShockLine MS46522B E-band VNAs configuration consists of small tethered source/receiver reflectometer modules and a base chassis.
The remote, small form factor reflectometer modules have native WR12 waveguide interfaces for convenient connection to typical waveguide devices. They also have short/long-term thermal stability due to the vanishing thermal gradient across the modules, high amplitude and phase stability, and raw directivity. Most importantly, placing the sampling directional bridge closer to the AUT/DUT provides long-term amplitude and phase stability.
Attenuation Measurements
Benefits of this system can be seen in the following example. Figure 3a shows the attenuation at 77 GHz. The middle of the emblem is marked by a vertical and a horizontal line. The attenuation of both lines at 77 GHz is displayed in Figure 3b. In the region of interest, the attenuation is fairly flat (between 0 dB and –1 dB) and increases at the edges of the emblem due to the shape and characteristics of the material.
To give an overview of the attenuation of the whole frequency range, figure 4a has five marked positions. The corresponding lines are shown in the same color in figure 4b. The electromagnetic wave hits the central position orthogonally, and therefore, the signal has a higher attenuation. Each position over the whole bandwidth has attenuation between 0 dB and –1 dB. Based on this, the signal loss caused by the shape and material characteristics is minimal and does not affect the radar significantly.
Value of Phase Shift Measurements
Phase shift provides additional information. Figure 5a shows a phase map of the emblem at 77 GHz. Figure 5b shows that the vertical red line changes over the distance, which might be due to an unevenly positioned emblem. The horizontal blue line has an average of –40° except on the edges of the emblem. The constant phase over the horizontal length indicates a uniform emblem shape and characteristic over the observed distance, so the phase is not heavily affected.
Figures 6a and 6b display an unwrapped phase to see continuous progress. Figure 6b displays the various frequency sweeps. Due to the dispersion, the phase changes with increasing frequency, so the shape of the emblem does not affect the phase of the transmitted signal. The red and yellow line have a different starting point because the emblem used for this measurement was uneven.
To learn more about the challenges of conducting E-band car radar emblem measurements and how they can be overcome with the proper VNA-based test system, download this Anritsu application note.
