Yu.V. Koltzov

Nizhegorodskiy Research Institute (Nizhny Novgorod, Russia)

The article is dedicate to unique radars – Radar on Chip (RoC) – vast possibilities which allowed in practice.

Driven by the demand for automotive assistance and safety systems, millimeter–wave (mm–wave) radars have received a lot of attention in the last decade, leading to highly sophisticated radar sensors being used in a large variety of different applications especially in automotive. These sensors are mostly based on frequency–modulated continuous wave (FMCW) or chirp–sequence modulation schemes. However, a disruptive development is observable in the field of mm–wave radars: the focus is on digital radars. The digital signal generation not only allows the flexible adoption of modulation parameters during operation but even enables the use of multiple waveforms and modulation schemes. Since the transmit signal is designed precisely in terms of time and spectrum utilization, it further enables simplified cooperation between sensors. This helps to avoid interference and enables the operation of multiple sensors in a sensor network. Additionally, unfiltered and uncut channel information is available at the receiver such that interference detection and its precise characterization are simplified, and the existence of other radars may even be exploited in the form of passive radar operation. This development is closely related to the evolution of orthogonal frequency–division multiplexing (OFDM) and phase– modulated continuous wave (PMCW) as modulation formats since those schemes can benefit most from a digital implementation. At the same time, the analog front–end hardware is simplified, and the combination of independent front– and back–end blocks modularizes radar design.

Yet there are reasons why this development is still research. The lack of affordable high–speed data converters, which are required to have a sufficiently good range resolution, is an important point to mention here, as well as the necessity of a digital processing engine that can handle the immense amount of data that need to be processed. For a sampling rate of 1 GHz, four in–phase/quadrature (I/Q)–channels, and a resolution of 14 bits, the data rate is about 112 Gb/s for both the transmitter and receiver. This leads to 14 GB of receive data per second, which, even for a short radar frame with a duration of 10 ms, sums up to 140 MB. Therefore, recent publications have mostly focused on optimized signal processing and the development of suitable transceiver monolithic microwave integrated circuits (MMICs), while the systems that have been realized and used to verify the concepts are rather bulky and rely on software–based offline processing. Only a few examples exist that take the next step toward a complete system. The most advanced integrated PMCW radar system–on–chip (SoC) given in our article, where the complete PMCW processing is integrated in an application–specified IC (ASIC).

In the following, we present a highly flexible all–digital 12×16 multiple–input, multiple–output (MIMO) mm–wave radar. It fulfills functional automotive–grade specifications and is capable of real–time evaluation of OFDM and PMCW signals. The article describes all hardware and software building blocks, starting with system aspects, the front–end hardware, and details of the radar MMICs. The modulation formats OFDM and PMCW and their signal processing chains are explained. The functionality of the whole system is verified with measurements for both PMCW and OFDM.

Compared to analog frequency–modulated radars, digital signal generation offers the possibility to realize modulation schemes such as OFDM and PMCW, which are instantaneously wideband. Both schemes employ a coded waveform, leading to a large processing gain and making it more robust against interference artifacts and hardware errors.

Despite the small size the automotive integrated radar have tremendous opportunities, previously unattainable, providing four– dimensional information (4D): three coordinates plus speed.

Koltzov Yu.V. Car radars on chip. Achievements of modern radioelectronics. 2021. V. 75. № 10. P. 50–75. DOI: https://doi.org/10.18127/ j20700784-202110-04 [in Russian]

1. Schweizer B., Grathwohl A., Rossi G. et. al. The Fairy Tale of Simple All–Digital Radars: How to Deal With 100 Gbit/s of a Digital Millimeter–Wave MIMO Radar on an FPGA. IEEE Microwave Magazine. 2021. July. V. 22. № 7. P. 66–76.
2. Voskresenskiy D.I., Dobychina E.M. Tsifrovye antennye reshetki bortovykh sistem. M.: Radiotekhnika. 2020. [in Russian]
3. Campbell S.D., Jenkins R.P., O’Connor P.J., Werner D.H. The Explosion of Artificial Intelligence in Antennas and Propagation. Antennas & Propagation Magazine. 2021. June. 63. № 3. P. 16–27.
4. Zardi F., Nayeri P., Rocca P., Haupt R.L. Artificial Intelligence for Adaptive and Reconfigurable Antenna Arrays. Antennas & Propagation Magazine. 2021. June. 63. № 3. P. 28–38.
5. Pell R. Next step in automotive electronics: functional exteriors. eeNews Automotive. 2021. August 09.
6. Joosting J.-P. Automotive radar scans 100x more objects in 4D with 300m range. MWee. 2018. January 16.
7. Dunn J. Designing mmWave radar systems for next–gen smart vehicles. Smart2zero. 2019. March 21.
8. Giannini V., Goldenberg M., Eshraghi A. et. al. A 192–Virtual–Receiver 77/79 GHz GMSK Code–Domain MIMO Radar System–on–Chip. 2019 IEEE International Solid–State Circuits Conference (ISSCC), February 19 / Session 9 / High–Frequency Transceivers for Radar and Communications.
9. Hindle P. Uhnder Emerges with Highly Integrated Automotive Radar SoC. Microwave Journal. 2019. March 13.
10. Giannini V., Hegde M., Davis C. Digital Code Modulation MIMO Radar Improves Automotive Safety. Microwave Journal. 2019. August 10.
11. Taranovich S. A digital mmW radar IC for automotive use. EDN. 2019. May 13.
12. Bourdoux A., Ahmad U., Guermandi D. et. al. PMCW waveform and MIMO technique for a 79 GHz CMOS automotive radar. 2016 IEEE Radar Conference (RadarConf), 2–6 May 2016.
13. Stark W., Ali M., Maher M. Digital Sode Modulation (DCM) Radar for Automotive Application. Uhnder. – White Paper, January 2020.
14. Silverman H. An introduction to programming the Winograd Fourier transform algorithm (WFTA). IEEE Trans. Acoustics, Speech, and Signal Processing. 1977. April. V. 25. № 2. P. 152–165.
15. Shapira N. Radar Systems for Autonomous Driving – at L2/L2+ and Beyond. Microwaves & RF. 2021. June 23.
16. Ginsburg B.P., Subburaj K., Samala S. et. al. A Multimode 76–to–81GHz Automotive Radar Transceiver with Autonomous Monitoring. 2018 IEEE International Solid–State Circuits Conference / Session 9 / Wireless Transceivers and Techniques / 9.1 / P. 158–160.
17. Mangraviti G., Khalaf K., Shi Q. et. al. A 4–Antenna–Path Beamforming Transceiver for 60 GHz Multi–Gb/s Communication in 28 nm CMOS. 2016 IEEE International Solid–State Circuits Conference / Session 13 / Wireless Systems / 13.5 / P. 246–248.
18. Moorhead P. Qualcomm Officially Enters Self–Driving Market With Snapdragon Ride Platform And Extends Partnership With GM To Include ADAS. Forbes. 2020. Jan 6. 
19. Qualcomm ob"yasnila neozhidannyy vybor nazvaniya dlya Snapdragon 888. 2020. 2 dekabrya. URL: https://www.ixbt.com/ news/2020/12/02/qualcomm-snapdragon-888.html
20. Qualcomm Snapdragon 888 na konchike pal'tsa. 2020. 2 dekabrya. URL: https://www.ixbt.com/news/2020/12/02/qualcommsnapdragon-888-2.html
21. Shkodnik V. Qualcomm raskryla kharakteristiki Snapdragon 888 – pervogo flagmanskogo protsessora s super"yadrom ARM Cortex–X1. 3Dnews. 2020. 2 dekabrya. [in Russian]
22. Joosting J.-P. Latest Qualcomm mobile platform «sets benchmark» for 5G, AI and gaming. Smart2zero. 2020. December 03.
23. Moorhead P. Qualcomm Officially Enters Self–Driving Market With Snapdragon Ride Platform And Extends Partnership With GM To Include ADAS. Forbes. 2020. Jan 6. 
24. James L. Qualcomm’s new platform for all levels of automated driving, Snapdragon Ride. Power & Beyond. 2020. March 05. URL: https://www.power-and-beyond.com/qualcomms-new-platform-for-all-levels-of-automated-driving-snapdragon-ride-a-911130
25. Expanded roadmap and software ecosystem keep Snapdragon Ride moving forward. URL: https://www.qualcomm.com/products/ automotive/autonomous-driving
26. Automotive System–on–Chips (SoCs). URL: https://www.renesas.com/sg/en/products/automotive-products/automotive-system-chips-socs
27. Winning A. R–Car V3U ASIL D SoC offers 60 TOPS of deep learning power. eeNews Embadded. 2020. December 17.
28. Pell R. ASIL–D SoC speeds ADAS, automated driving development. Smart2zero. 2020. December 29.
29. Renesas Develops Automotive SoC Functional Safety Technologies for CNN Accelerator Cores and ASIL D Control Combining World– Class Performance and Power Efficiency. Design & Reuse. 2021. February 23.
30. Hammerschmidt C. Scalable platform meets demand for high computing performance. eeNews Automotive. 2020. December 17.
31. Versal AI Edge Series – samaya masshtabiruemaya i adaptiruemaya platforma dlya okonechnykh i vstroennykh sistem ot Xilinx. Makro Grupp. 2021. 21 iyunya. URL: https://www.macrogroup.ru/news/2021/versal-ai-edge-series-samaya-masshtabiruemuyu-iadaptiruemaya-platforma-dlya-okonechnyh-i. [in Russian]
32. Joosting J.-P. Xilinx AI Edge SoC platform offers highest AI performance–per–Watt. eeNews Embedded. 2021. June 10.
33. Xilinx Extends Edge Compute Leadership with World’s Highest AI Performance–per–Watt. Businesswire. 2021. June 09.
34. Joosting J.-P. Adaptive compute acceleration platform targets big data. MWee RF – Microwave. 2021. July 15.
35. Hammerschmidt С. Nvidia announces new superprocessor for autonomous cars. eeNews Automotive. 2021. April 12.
36. Hammerschmidt С. Continental develops auto–driving computing platform with Nvidia. eeNews Autromotive. 2018. February 05.
37. Flaherty N. NIO launches commercial selfdriving car with 1000TOPS. eeNews Europe. 2021. January 11.
38. Pell R. Tesla unveils custom chip for AI–training supercomputer. Smart2zero. 2021. August 23.
39. Detinich G. Samyy bol'shoy v mire protsessor Cerebras okazalsya v sotni raz bystree superkomp'yutera na baze NVIDIA GPU. 3Dnews. 2020. 19 noyabrya. [in Russian]
40. Kobal'chinskiy V. II-chip Cerebras obognal superkomp'yutery v modelirovanii fizicheskikh protsessov. KO. 2020. 1 dekabrya. URL: https://ko.com.ua/ii-chip_cerebras_obognal_superkompyutery_v_modelirovanii_fizicheskih_ processov_135488. [in Russian]
41. Moore S.K. Supersize AI. IEEE Spectrum. 2021. July.
42. Pell R. Industry's largest AI processor has 2.6 trillion transistors. Smart2zero. 2021. April 21.
43. Cerebras Systems Announces World's First Brain–Scale Artificial Intelligence Solution. Design & Reuse. 2021. August 30.
44. Flaherty N. Analog Bits moves IP to 5 nm, targets 3 nm. eeNews Europe. 2021. May 31.
45. Detinich G. Analog Bits predstavila nabor IP dlya proektirovaniya analogovykh i smeshannykh chipov dlya 12–nm tekhprotsessa GlobalFoundries. 3Dnews. 2019. 4 iyunya. [in Russian]
46. Panel Level Packaging Consortium 2.0 – The First Year!. 2021. August 18. URL: https://www.izm.fraunhofer.de/ en/news_events/tech_news/panel-level-packaging-consortium-2-0-the-first-year.html
47. Hammerschmidt С. Fisker Ocean EV will come with advanced digital radar. eeNews Automotive. 2021. July 28.
48. Magna to Bring Driver Assistance Into the Digital Age With Industry–First Capabilities in 2022. Microwave Journal. 2021. July 27.
49. Uhnder and Yunshan Deliver Digital Radar to Accelerate Smart Port Automation. Microwave Journal. 2021. April 21.
50. Hammerschmidt C. Uhnder, dSpace to jointly develop radar technology. MWee RF – Microwave. 2020. April 20.
51. Hammerschmidt С. TDK, Uhnder partner for high–precision imaging radar platform. eeNews Automotive. 2021. January 11.
52. FUJITSU TEN develops automotive compact 77 GHz 3D radar. Microwave Journal. 2021. October 22.
53. Kawakubo A., Tokoro S., Yamada Y. et. al. Electronically–scanning millimeter–wave RADAR for forward objects detection. SAE Congress, 2004. P. 127–134.
54. Jeong S.-H., Yu H.-Y., Lee J.-E. et. al. A Multi–beam and Multi–range Radar with FMCW and Digital Beam–forming for Automotive Applications. Progress in Electromagnetics Research. 2012. V. 124. P. 285–299.
55. Metawave Demonstrates First 77 GHz Analog Beam Steering 3D Radar. Microwave Journal. 2021. July 27.
56. 4D Automotive Radar Overlays AI on Existing Hardware. Microwave Journal. 2021. August 13.
57. Oculii Unveils the Industry’s Highest–Resolution Commercial Imaging 4D Radars. Microwave Journal. 2021. March 9.
58. Advances in Radar and Components. Microwave Journal. eBook. 2021. September.
59. Hindle P. New 4D Automotive Radars Hitting the Market. Microwave Journal. 2021. August 13.
60. Browne J. Automotive Radar Paves the Way to Safer Roads. Microwaves & RF. 2021. April. V. 60. № 3. P. 18–22.
61. Glover K., Peterson B. UWB: Enhancing Positioning, Safety and Security for Connected Vehicles. Microwave Journal. 2021. September. V. 64. № 9. Р. 22–34.
62. Antenna Processor Enables All–Digital Radar, EW and Communications. Microwave Journal. 2021. September. V. 64. № 9. Р. 46. September Supplement 2021. Military & Aerospace. V. 64. Edition: 09 Sup.