The connected car is a driving force in the ever-expanding Internet of Things (IoT) world. One of the many technologies that is integrated into these smart cars is collision avoidance systems that utilize automotive radar to help prevent traffic accidents. Due to the high frequency of automotive radar – typically in the E-band – and the need to detect objects at both short and long range, design verification for engineers developing these systems poses obstacles.
A tried and true instrument, the spectrum analyzer, remains a valuable piece of equipment in the lab or on the production floor in these high-frequency applications. To accurately conduct tests on automotive radar, however, the spectrum analyzer must achieve a new level of performance and have specialized features.
In addition to collision avoidance, automotive radar is currently used for Adaptive Cruise Control (ACC). In both cases, it is used to detect the distance and velocity of large objects such as vehicles traveling in front. Automotive radar is an element of Advanced Driving Assistant Systems (ADAS), which also use a variety of sensors to detect objects.
One automotive radar requirement for emerging ADAS is the ability to detect small objects with a high-resolution of distance, velocity and angle. In order to achieve this, wideband Frequency Modulated Continuous Wave (FMCW) modulation is becoming a key technology. FMCW modulation transmits signals to sweep from lower to upper frequencies with constant amplitude, as well as detect the velocity and distance using beat frequencies generated by mixing transmitted signals and received reflected signals.
Another aspect of these automotive radar designs is maximum continuous 4 GHz radar bandwidth, which has been adopted by the industry. In the near future, there will be demand for high-resolution millimeter wave (mmWave) radar using the 79 GHz band. Measurement instruments that employ conventional harmonic mixers might not accurately test wideband modulation signals that have a 4 GHz maximum bandwidth at mmWave frequencies.
The FCC and ETSI have developed specifications that outline radio characteristics tests for 79 GHz radar, including verification of phase noise performance and FMCW frequency. Both can be measured using a spectrum analyzer, however because of the E-band frequency and wide bandwidth, engineers have a new set of considerations.
Overcoming Image Response
A spectrum analyzer is usually used with a harmonic mixer when measuring mmWave signals for two reasons – lower cost of test and a simplified configuration. One potential drawback is that a harmonic mixer can cause its image response to be displayed on the spectrum analyzer, leading to inaccurate measurements and faulty designs. Because image response of a mixer occurs at an offset of twice the IF frequency above and below the desired signal, selecting a spectrum analyzer/harmonic mixer that has a high IF frequency can eliminate this issue.
For these reasons, the ideal spectrum analyzer for these designs will feature a very high IF frequency, such as 1.875 GHz, so engineers can conduct accurate spectrum measurements on wideband modulated signals. Another factor is the measurement span of the spectrum analyzer. With this high IF frequency, an image free CW span of 7.5 GHz allows a wider bandwidth to be measured in a single sweep, rather than multiple sweeps that add test time and complexity. Choosing an appropriate pre-selector at the mmW frequency can also help reject the image frequencies.
Tx Power and Spurious Emissions Measurement
In many automotive radar designs, engineers will need test equipment with good sensitivity to conduct Over The Air (OTA) tests. If the test antenna is 50 cm from the device under test (DUT), the free-space loss for a 79 GHz signal is 65 dB. Since Maximum Radiated Average Power Spectral Density defined in the ETSI standard requests measurements of <-40 dBm/MHz, the requirement for test equipment is approximately -142 dBm/Hz at 79 GHz with 23 dB test antenna gain.
To meet the specification outlined in the ETSI standard, a spectrum analyzer will have to have a wide dynamic range that can produce excellent sensitivity to create a very low noise floor. For example, a noise floor of -150 dBm/Hz will allow for the measurement of very low level signals. Additionally, the wide dynamic range can help achieve the necessary 1 dB compression point (P1dB) performance, so small signal measurements can be conducted to determine the true performance of the DUT. These spectrum analyzer features are supported by a harmonic mixers with low conversion loss of <15 dB to eliminate the need for noise cancelling functionality when conducting TX power and spurious emission measurements in accordance with the specification.
Wideband FMCW Modulation Test
To test FMCW modulation quality, phase noise characteristics and frequency linearity need to be verified. When the automotive radar detects the adjacent objects, the time and frequency difference between transmitted and received signals is smaller. If phase noise performance is insufficient, both signals cannot be separated because the received signal might be hidden in the phase noise of the transmitted signal. A 79 GHz high-resolution radar will have -90 dBc/Hz (100 kHz offset) and -100 dBc/Hz (1 MHz offset) phase noise performance. Therefore, the spectrum analyzer needs to have performance that exceeds these parameters to achieve measurement accuracy. Instruments with close-in phase noise performance of <-100 dBc (100 kHz offset), <-110 dBc/Hz (1 MHz offset) at 79 GHz can achieve the measurement accuracy necessary for engineers to have design confidence.
Conclusion
The demand for 79 GHz high-resolution radar with a maximum 4 GHz bandwidth has created test challenges for engineers involved in automotive radar designs. To overcome these obstacles, engineers should select a spectrum analyzer that features a high IF frequency and pre-selector functions to measure wideband signals, as well as have excellent sensitivity performance.
