Multiple-input multiple-output (MIMO) is normally discussed in terms of wireless networking and telecom, but it is a transmission and reception technique that can be used with any signal, as long as that signal’s emitter and receiver have some well-defined spatial resolution. Some upcoming applications of FMCW MIMO radar are focused in highly accurate automotive radar modules and robotics, both of which require a suite of sensors for interfacing with the outside world. If you’re interested in integrating FMCW MIMO radar into your RF system, I’ll discuss its implementation in three critical areas: aerospace and defense electronics, automotive, and robotics.

What is MIMO Radar?

A MIMO radar system applies the same concepts used in MIMO antennas for telecom and simply applies them to higher frequencies. A MIMO system uses an M x N antenna array, where the TX side of the system broadcasts multiple orthogonal signals. The elements in the TX side might be phased arrays, and orchestrated beamforming could be used to steer the beams from this array of TX arrays. The interfered beam that is produced in MIMO radar has much higher angular resolution than the beam produced by a single array, and the orthogonality of different Tx signals allows the receive side to distinguish between signals that are broadcast over different angles. In this way, you can perform much higher resolution scanning, giving more accurate target tracking, because each signal received by the Rx array is linked to a specific Tx element.

The use of MIMO is not normally discussed in the context of small modules for automotive radar. MIMO is often discussed in the context of antenna and systems design for wireless technologies like 5G, but it really applies to any sensing technology, including radar. As an example, with spatially distributed optical transceivers, there is even the possibility of developing FMCW MIMO lidar. Applications that need precise angular resolution with mmWave sensors can use MIMO radar with off-the-shelf components and a unique system architecture. Principal application areas include:

• mmWave imaging. This sounds like principally a medical application, but this is not really the case. The point cloud generated with scanned radar measurements can be used for a type of imaging, where object outlines are detected and motion of objects can be tracked. The recognition of tracking occurs through signal processing using multiple measurements.


• Drone/UAV radar. MIMO isn’t required for drone radar, but it is useful for smaller drones and robots that need to get closer to target objects and have more accurate detection methods that are not based on image recognition.


• More accurate ADAS. The advanced driver assistance systems (ADAS) used in automobiles currently rely on multiple radar modules to provide short-range detection of nearby objects with broad angular sweep.


• 4D radar. This is important for car radar, UAV radar, and robotics radar as it incorporates elevation as one of the elements that can be distinguished from signals received on the Rx side.

I mention imaging and image recognition above for multiple reasons. First, robotics and modern automobiles rely on a suite of sensors to interact with the natural world, in the same way the human brain uses inputs from our five senses. Using radar for clear object recognition and tracking aids tasks in computer vision, namely object recognition and image segmentation.

Second, anyone who has worked on imaging systems will tell you that the "images" used by computers to interpret the world are almost never the same types of "images" we think of as humans. They are almost never photographs; they are some 2D map of information extracted from a sensor, whether that sensor be a CCD camera or otherwise.

Designing FMCW MIMO Radar

FMCW MIMO radar design starts by choosing an orthogonal signal basis to be used with different portions of the antenna array. When we say "orthogonal" in this context, we’re referring to multiple signals in the same sense as in orthogonal frequency division multiplexing (OFDM), where multiple data streams are broadcast over multiple subcarrier signals in a broadcast pulse. Instead of broadcasting a signal with a defined data rate, we’re broadcasting multiple FMCW signals (linear chirp). Each signal can then be decoded using the same techniques as in conventional FMCW radar with off-the-shelf components.

System Architecture

The block diagram below shows schematically how an FMCW radar system can be constructed. In these systems, the TX elements are broadcasting distinct carrier frequencies, all of which can be emitted simultaneously. The TX elements are normally individual antennas, although they could each be a phased array, where each array uses beamforming to scan the field of view. FMCW signals are synthesized with linear chirp for each TX element.

 

 

The RX array will receive multiple frequencies that are broadcast by the TX elements, and the signal detected by each RX element is used to determine the angle of arrival. The received are distinguished using a multiplexer, and the beat frequency from each signal is recovered with a standard heterodyning technique with a mixer. The mixer element is important here for detecting a Doppler shift in the recovered signal, which will be used to determine the speed of the target.

FMCW MIMO Radar and Angular Resolution

Because each TX element has a defined angle with respect to the RX element, you only need the standard angle-of-arrival and time-of-flight measurements to determine the location of the target with respect to the RX array. The standard constant false alarm rate method can then be used to resolve targets for each of the received carriers. Here, you’re relying on the fact that individual TX elements are broadcasting at orthogonal frequencies for precise angular resolution. This fact, combined with the use of a greater number of TX elements than in a typical car radar module, makes these systems very precise and suitable for unique applications like radar imaging.

Today, Texas Instruments is arguably the leader in the integrated mmWave radar transceiver/sensor space. Their component options are flexible enough to implement the block diagram shown above. They currently support cascaded radar with some of their reference designs, and we’ve worked with some aerospace and defense electronics companies to implement these reference designs into custom products.

Upcoming component from many vendors will be highly integrated and support dozens of antenna elements for highly accurate imaging, target tracking, and object detection using FMCW MIMO radar. However, there is still the question of the role of lidar, which is seeing similar innovation with FMCW and MIMO being applied to these systems.

Will We Go Beyond Lidar?

Advanced 4D FMCW MIMO radar developers claim their systems can surpass lidar, or they could completely eliminate the need for lidar and image recognition systems with standard cameras. I’m skeptical we’ll ever get to that point, especially in applications like UAVs where images still need to be relayed back to operators. For automotive, the story is a bit different, and it remains to be seen how the safety regulatory environment will influence design of these systems in terms of sensor requirements. Currently, commercial lidar still beats radar in terms of resolution, as can be seen in the example image below.

 

 

Just like MIMO techniques are applicable to radar, they are also applicable to lidar, and it will be interesting to see if FMCW MIMO lidar systems become commercialized. If silicon photonics transceivers proliferate in this space, it’s possible to have a compact transceiver element that broadcasts IR pulses through a spatially multiplexed nonlinear emitter, where self-phase modulation applies a linear chirp to pulses via the Kerr effect. These systems still require significant effort until they can be made more compact, and fabrication of nonlinear multiplexed emitters/receivers is still a challenge.

If you’re interested in reviewing some progress on FMCW lidar, take a look at the IEEE articles I’ve linked below:

• Chung, SungWon, Makoto Nakai, Samer Idres, Yongwei Ni, and Hossein Hashemi. "19.1 Optical Phased-Array FMCW LiDAR with On-Chip Calibration." In 2021 IEEE International Solid-State Circuits Conference (ISSCC), vol. 64, pp. 286-288. IEEE, 2021.


• Zhang, Hongxiang, Kai Chen, Zhongyang Xu, Dan Zhu, and Shilong Pan. "Frequency-modulated continuous-wave lidar using a phase-diversity coherent optical receiver for simultaneous ranging and velocimetry." In CLEO: Applications and Technology, pp. AW4K-3. Optical Society of America, 2019.


• Lukashchuk, Anton, Johann Riemensberger, Maxim Karpov, Junqiu Liu, Erwan Lucas, and Tobias J. Kippenberg. "Microresonator Dual-Comb Coherent FMCW LiDAR." In 2020 Conference on Lasers and Electro-Optics (CLEO), pp. 1-2. IEEE, 2020.


• Qi, Baoling, Qingyan Li, Dongbing Guo, Bin Zhang, and Chunhui Wang. "A High-precision FMCW Lidar Ranging System Based on Dual-path Error Compensation Algorithm." In 2020 IEEE International Conference on Information Technology, Big Data and Artificial Intelligence (ICIBA), vol. 1, pp. 1428-1432. IEEE, 2020.

 

If your company wants to push the limits of UAV, aerospace, and defense electronics with innovative FMCW MIMO radar, it pays to work with the best electronics design firm. NWES helps electronics OEMs, aerospace OEMs, and defense primes design advanced RF PCBs and create cutting-edge embedded technology. We've also partnered directly with EDA companies and advanced ITAR-compliant PCB manufacturers, and we'll make sure your next high speed digital system is fully manufacturable at scale. Contact NWES for a consultation.

 

---

---

---

---