On June 22, 1941, Nazi Germany was still unable to gain the air supremacy of the English Channel despite the loss of 1977 planes and 2585 pilots, let alone to disrupt the British ground and naval forces by air raids. It had to give up the "Sea Lion Plan" to invade Britain. The British air war that lasted more than a year ended in the defeat of Nazi Germany. The reason why Britain was able to withstand the attack of the German chariot and win the British air war (mainly because the chariot could not cross the sea) was that, in addition to the great role played by star fighters such as "Spitfire" and "Hurricane", there was also a great hero behind the scenes - the anti air radar called "Chain Home".
The first practical radar in the world was developed by Sir Robert Watson Watt, a British scientist and descendant of James Watt, who invented the steam engine. In order to give early warning to Nazi aircraft as early as possible, the British Air Force decided to deploy such radar in China on a large scale in May 1936, which is the prototype of the "local chain" radar. By the beginning of 1939, a total of 20 radar stations had been put into use. Before the implementation of the "Sea Lion Plan" of Nazi Germany, Britain had built two radar detection networks, with a total of 51 radar stations. These radars have made an important contribution to resisting the air attack of the German Air Force, and since then, the following 80 years of wide application of millimeter wave radars in various fields have begun.
In terms of vehicle mounted millimeter wave radar research, European and American countries have also been in the forefront of the world. Bosch, mainland, Haila and other companies have monopolized the global market. The application of millimeter wave radar in the automotive field can be traced back to the early 1980s. Universities and research institutions in some European and American countries have gradually begun to research the vehicle mounted millimeter wave radar technology. In the mid-1980s, Europe formulated the "PROME THE US", which triggered the research and development of radar technology in Europe, Japan and other automobile countries. In 1995, Mitsubishi Motors first used the "PreviewDistance Control" system on Diamante based on the millimeter wave radar. However, this system can only be regarded as an early version of adaptive cruise, because it only reduces the vehicle speed by controlling the accelerator and downshifting, and does not interfere with the brake itself. Until 1999, Mercedes Benz first applied the real adaptive cruise system on the S-class, opening the era of assisted driving. Its name must be familiar to everyone, which is called Distronic. Early vehicle mounted millimeter wave radar chips mainly used the GaAs process. A millimeter wave radar needs to be equipped with at least 7 to 8 RF chips working in the 24GHz frequency band. The radar wavelength is long, resulting in the millimeter wave radar being too large and bulky, about the size of a laptop. Therefore, the cost is also very expensive. Similar to today's lidar, it can only be applied to a small number of high-end vehicles. At the beginning of 2000, the development of SiGe technology greatly improved the integration of millimeter wave radar chips. A millimeter wave radar only needs 2 to 5 MMICs and 1 to 2 BBICs. The cost has reached the level of 1000 yuan, but the penetration rate is still very low. At present, the mass production 77GHz millimeter wave radar commonly used in high-end models uses this technology. In 2017, ti launched a highly integrated 77GHz millimeter wave radar chip based on CMOS technology. Its AWR1642 series, which is suitable for short and medium range scenarios, integrates the front-end MMIC RF, DSP and MCU modules on a 77GHz millimeter wave radar SOC chip, significantly reducing the cost of millimeter wave radar and greatly reducing the difficulty of hardware development of vehicle mounted millimeter wave radar. What's more exciting is that TI has built a more integrated antenna on chip (AoP) chip for close range scenarios, integrating the antenna into the chip, breaking the job of antenna engineers and lowering the price of millimeter wave radars to 100 yuan. As we all know, compared with other radar chip manufacturers, TI chip development has the characteristics of fools. The development of underlying software is complete, and the tool chain is easy to use, which greatly reduces the threshold for entry. Based on this chip, China has embarked on a vigorous journey of localization of vehicle mounted millimeter wave. In just a few years, there have been about 3 or 40 Ago and Amare companies that have carried out research and development of vehicle mounted millimeter wave radar products in China, which is not a bustle, It can be said that it is not spectacular.
The research and development history of domestic vehicle mounted millimeter wave radar products is not long, which is just 6-7 years, and it is in the initial stage. Up to now, Huayu Automobile, a component company of SAIC Group, has really achieved mass delivery, with an annual shipment of more than 200000 pieces. As the Huangpu Military Academy in the domestic vehicle mounted millimeter wave radar industry, Huayu Automobile (hereinafter referred to as Huayu) is the first in China to carry out the research and development of vehicle mounted millimeter wave radar products in 2014. But that time is not like today. Six years ago, 77GHz millimeter wave radar chip was a high-end novelty. Foreigners imposed an embargo on us and did not take us to play. Therefore, Huayu could only use 24GHz millimeter wave radar as a starting point. Although he is the son of the local local tycoon SAIC Group and the standard "rich second generation", the background and strength of the car borne radar start-up team in Huayu at that time were no better than those of the radar start-up companies now. At first, he was a group of cross-border engineers with PEPS product background, and had nothing to do with millimeter wave radar. Except for money and nothing, he was ignorant and dared to be the first, He plunged into the big wave of millimeter wave radar research and development. For the vehicle mounted millimeter wave radar, a product with high technical threshold, high market threshold and great difficulty in productization, it is really a big problem that there is no professional R&D team. However, as long as there is money, everything is not a problem. It is said that in order to cheat the cattle, the interview place was arranged at No. 489 Weihai Road. SAIC's Gaopuge blessing, luxurious office environment, automatic heating and flushing toilets, attractive canteens, irresistible salaries, and touching feelings. In order to recruit the team, Huayu HR made every effort to recruit the team, which was simple, brutal, and extremely cruel, with four words everywhere: "I have money". Of course, some headquarters chiefs who are used to doing foreign compradors, are used to what they can buy with money, and are not willing to work hard to make their own style. They are still skeptical of the independent research and development ability of the local turtle team. So at the beginning, in order to ensure the success of research and development, Huayu adopted a dual insurance strategy: the parallel way of foreign technology introduction and independent research and development. The domestic team independently researched and developed the solution, and purchased the solution from the famous SMS, smartmicro. It is said that in order to run fast forward, Homogeneous Electronics and Desai Xiwei also bought the solution of SMS home to do radar. Of course, Mr. Rohling is very clever in business. He only sells solutions and will not give any core technology. Mr. Rohling is knowledgeable and has profound attainments in various radar research, but he is still lacking in commercialization. For example, the cost of the radar scheme originally designed for Huayu is not competitive. Of course, it is gratifying that the local team of Huayu overcame difficulties and overcame difficulties, and finally achieved the large-scale production of the first 24GHz radar in China. After a breakthrough of 0 to 1, the subsequent 77GHz angle radar and forward radar also achieved mass production soon. In addition to Huayu, Senstec holds the thigh of Haikang. The shipment of radar in terms of intelligent security and traffic is also very good, but in terms of vehicle mounted radar, I have not heard that their family has a large shipment. Most other vehicle mounted millimeter wave radar R&D companies are still in the PPT and public account stages. They claim to have achieved mass production, and they also take small batches of samples for customers to test. After all, the threshold of vehicle mounted millimeter wave radar is still very high, and the core technology has not been popularized in the industry at present. Including Bosch, Continental and Haila, whose radar has been mass produced for nearly 20 years, there are only a few experts in the team who master the core key Knohow. In addition, for forward radars, involving braking, steering and power systems, these core actuators are in the hands of Bosch, mainland and other giants. It is easy for people to control their own actuators with their own millimeter wave radars and provide a complete set of solutions. Domestic manufacturers hardly have their own actuators, so it is difficult to do forward radars. Making a Demo to output a point cloud is relatively easy for each company, but it is really going to mass production, and BOM cost and performance can be compared with that of ABCD. It also needs to deepen the understanding of radar, including a large number of system, software, and hardware core technologies that need to be broken through, especially the algorithm problems that need to be overcome. Therefore, many domestic millimeter wave radar start-ups have begun to turn to the fields of intelligent security and intelligent transportation. The millimeter wave radar in these fields is relatively low in technical difficulty, the market threshold is not as high as that of the automobile industry, and the profits are somewhat violent. They can see returns in a short period of time. For start-ups, survival is the first goal, and integrity and original intention are like Jobs, which make people suffocate. However, building a vehicle mounted millimeter wave radar is a thankless job. It requires not only thighs to hold it, but also high technical and market barriers, low profit margins, and low volume. It is basically a loss business, unless you can challenge "I am the best in the world" like Mobileye.
In addition to the traditional ADAS millimeter wave radar, the 4D millimeter wave radar with high resolution point cloud imaging capability has become the industry hotspot in recent two years for the high-precision perception needs of high-level automatic driving. Some giants and start-ups at home and abroad are focusing on this new concept product, with the goal of replacing or making up for the laser radar in some scenarios. After all, the cost and reliability of laser radar are still difficult to land in recent stages, In severe weather such as rain and snow, millimeter wave radar is also required to perform the task of accurate and stable perception. The so-called 4D refers to 3D contour high-resolution point cloud contour and high-precision speed information. It is like concentrating the capabilities of the phased array radar on an aircraft carrier onto a radar the size of an Apple phone. The technical difficulty is still greatly improved compared with the traditional ADAS millimeter wave radar. At present, 4D millimeter wave radar mainly has two mainstream routes. One is to use traditional MMIC RF chips for multi chip cascade to form a multi send and multi receive antenna array, and obtain 4D point cloud imaging through antenna array layout and algorithm optimization. For example, Bosch's LRR5 Premium, mainland's ARS540, and our products all use this route; The other way is to use self-developed large-scale multi-channel on-chip integrated chips, such as Israel's Vayyar, Arbe, etc., which integrate 48 reception channels+48 transmission channels and antennas on one chip. At present, the mainland ARS540 has a preliminary prototype, but no public test results have been seen. It is said that it is customized for BMW electric vehicle iX. The range can be detected to 300m, the horizontal angle resolution can reach 1.2 °~1.68 °, and the pitch angle resolution can reach 2.3 °; The horizontal angle resolution of Bosch LRR5 is about 2 °, and the pitch angle resolution is about 2.2 °; Arbe official website claims that the horizontal angle resolution can reach 1 ° and the pitch angle resolution can reach 2 °. No matter which technical route, the radar volume of each company is about the same, about 12cm * 13cm * 3.5cm, indicating that no one has broken through the theoretical constraint between antenna aperture and angular resolution. To achieve high resolution, the physical size of antenna array must be increased. Oculii introduced that his virtual imaging technology can expand the number of antenna channels by 10~100 times. We also have such technology, but it is not very meaningful in practical applications. At present, the industry usually adopts four or two cascaded chips to increase the number of receiving and transmitting channels, and combines MIMO technology to form a large-scale virtual array to improve the angle resolution in the horizontal and pitch directions. Although this method can relatively save costs, it will also bring about the corresponding problem, that is, the speed range is greatly reduced without blurring. How to solve this problem is also a bottleneck for multi chip cascaded MIMO millimeter wave radars, So we see that many domestic radar manufacturers only use two transmission channels in the multi chip cascade, wasting the remaining transmission channels to avoid the problem of speed ambiguity resolution. By cascading four Ti AWR2243, our 4D millimeter wave radar can achieve a horizontal resolution of 1 °, a pitch angle resolution of 1.4 °, and a maximum unambiguous speed measurement range of - 250km/h~+250km/h. We can achieve high-resolution imaging of the surrounding environment and targets. Related products have been applied to L3 and L4 automatic driving in several car factories. In the process of product development, we have broken through the following core key technologies:
1. Large scale virtual antenna design technology, greatly improving the level and angle resolution
2. Multilayer lamination design and mixed compression technology of high-frequency RF board to ensure low cost and yield of products
3. Fast technology for horizontal and pitch correction of large-scale antenna
4. Radar complex waveform design technology to meet the requirements of automatic driving
5. The maximum unambiguous speed measurement range expansion technology meets the requirements for accurate speed measurement in high-speed scenes
6. Expand the target clustering and tracking technology to obtain accurate 3D BoundingBox information of the target
Of course, the above technologies are just simple examples. In order to realize the commercialization of millimeter wave radar, there are still a lot of tool chain problems, engineering technology problems and algorithm problems to be solved. We have communicated with many technical teams before and found that although the team strength is strong, the actual level is uneven. Even the BSD millimeter wave radar is in the stage of crossing the river by feeling the stones, so we try to talk about the forward radar and 4D millimeter wave radar with higher technical difficulty. Due to the current situation of high entry threshold, great technical difficulty and weak domestic related technical foundation of vehicle mounted millimeter wave radar, there is also a lack of reference books and materials to systematically introduce vehicle mounted millimeter wave radar technology in the industry. Therefore, the author combines the research and development experience and accumulation of millimeter wave radar products at home and abroad for more than ten years, and is the first in opening business to share the millimeter wave radar technology with everyone, and to share with peers.