Dr. Xu Wei, Qualcomm’s head of R&D in China, delivered a speech
In the process of technological evolution from 5G to 6G, there can be two approaches. One approach is to incorporate existing technologies into the 5G or 6G development plan from a technical point of view; the other approach is to start from a demand perspective. We need to observe what market needs cannot be solved under the 5G framework during the commercialization of 5G, and at the same time capture some major development trends as the driving force for our next 5G or 6G development. As a sharing from the industry, I would like to exchange more about some of the needs and trends seen in the 5G business process, so as to explore from a slightly different perspective: 5G commercial, 5G products or 5G standard technologies have encountered What are the problems, and how do we solve these problems? From the perspective of technological evolution, we don’t have to care too much about whether a technology is 5G Advanced or 6G. 5G has just started commercial use in 2019, and there are many problems that 5G technology can solve. It is only for the continuous development of 5G to consider; on the other hand, we are also very concerned about what revolutionary changes must be solved by 6G. I hope to bring different perspectives to teachers in academia.
First of all, look ahead to the global 5G commercial situation. Since 2019, in just two years, more than 175 operators around the world have launched 5G commercial services. Among them, China’s 5G business process is very significant. As of August this year, China has built more than 1 million 5G base stations, accounting for more than 70% of the world’s 5G base stations; at the same time, the number of 5G terminal connections in China has also approached 420 million.
In this context, with a large amount of investment in 5G, we need to think about what applications and services can be achieved through 5G, and when we apply 5G, whether 5G has reached our definition and requirements. The standard you want to achieve. Only after clarifying these issues can we study 6G more instructively. If 5G has not yet reached these standards, has not yet penetrated into the vertical field, and has not yet reached the level of changing lives and society, then we are likely to face similar difficulties when talking about 6G. Therefore, today I will focus on discussing with you the status quo of 5G and the next technological evolution direction.
According to market reports, global mobile data traffic will increase more than 6 times from 2019 to 2026. Although 5G will increase its peak rate by more than 10 times compared with 4G, in the face of increasing social demands, when data traffic continues to rise to more than 6 times the current level, can 5G technology meet these demands? In 2026, the commercialization of 6G has not yet begun. When 5G technology is needed to meet these needs, how can the existing 5G technology solve this problem?
The answer to these questions is actually very simple, that is, to open up higher spectrum applications, such as millimeter wave applications. The reason why there is research on terahertz now is a very important point, because large bandwidth is extremely critical when dealing with large-scale or large-scale traffic growth. I once said that from 5G to 6G, millimeter wave is the only way for technological evolution. It is not only 5G, but also the core technology of 6G. Why do you say that? For example, in the Sub-6GHz frequency band, we can only get 100MHz spectrum resources, but now 5G millimeter wave spectrum can already get 800MHz spectrum resources. Simply speaking, from the point of peak rate alone, the current commercial millimeter wave can achieve 8 times the data enhancement on this point of the spectrum. This is a very important reason why we need to consider the commercial use of 5G millimeter waves. This is also the reason why we introduced millimeter waves in the first version of 5G. Therefore, the commercialization of millimeter waves does not have to wait for 6G. When we discuss the integration of communication and perception, or the integration of radar and other communications, we are actually talking about millimeter waves.
Of course, in the early days of millimeter wave applications, we also heard many different voices. This is because there are indeed a large number of technical challenges in millimeter wave applications, and similar technical challenges will double in the applications of terahertz and optical communications. So how does Qualcomm solve the challenges that arise in millimeter wave applications? When I was doing my doctoral thesis 25 years ago, the challenges of millimeter wave already existed, such as the limited coverage of millimeter wave and the high cost, etc., and everyone’s understanding of millimeter wave at that time still remained almost at millimeter wave mobile. Impossible at this stage. At that time, I was probably one of the first people to do millimeter wave communication channel measurement in the United States, and I needed to build a millimeter wave system entirely by myself. At that time, I studied millimeter waves at 28GHz, 38GHz and even 60GHz, and studied the influence of trees, houses, climate and many other factors on millimeter wave communications, as well as millimeter wave AOA and AOD measurements.
One of the possible challenges for millimeter waves is limited coverage. When we studied 5G, the coverage problem has been solved by introducing ultra-large-scale antenna technology. And in many cities or indoor coverage scenarios, we are limited by system capacity rather than coverage. In fact, the density of many 3G/4G base stations is not every two kilometers. In densely populated areas in cities, the distance between most base stations is only a few hundred meters. We have seen in experiments in many cities that with the same base station options, millimeter waves can also have good coverage. This problem is not a limitation of coverage, but a problem of system capacity. This is a completely different networking method from when I was doing my PhD research. Second, millimeter waves have relatively large attenuation during non-line-of-sight transmission, which can actually be an advantage to improve spectrum reuse. In fact, most of the current systems, including 3G and 4G systems, require dense base station deployment, a lot of multiplexing, and the need to deal with interference between base stations. The millimeter wave has a small propagation range and cannot penetrate walls, which can reduce interference and achieve better spectrum reuse. Whether it is indoor multiplexing or multiplexing between satellite communications, or multiplexing between aircraft and aircraft, the principles are similar. Third, now millimeter waves can not only be applied to fixed use cases, such as a large number of use cases for fixed wireless access. With the support of beamforming and beam tracking technology, it can also ensure robust mobility, so the mobility problem is also very good. The solution. Finally, with the maturity of RF technology, there are currently more than 140 types of 5G terminals in the world that can support millimeter wave communications, and the technical challenges faced by millimeter waves in commercialization have also been resolved.
The challenges faced by millimeter waves and our series of solutions also provide us with valuable experience in the expansion of higher frequency spectrum. The challenges we face in terahertz or optical communications are also expected to be solved gradually. Of course, millimeter waves already have abundant spectrum resources. Compared with millimeter waves, terahertz and optical communications will face greater channel attenuation and other challenges.
As for the application scenario of millimeter wave, in simple terms, it is hot spot coverage. This is because the propagation range of millimeter waves is limited, but hot spots can be well covered, and it can provide enough capacity for a large number of users to use the high-frequency network at the same time. In fact, China is a country very suitable for millimeter wave applications, because China is densely populated and the use of smart phones is very common, while the population of large cities is concentrated-in densely populated places with very high user traffic, it is precisely millimeter wave that achieves hot spot coverage. Very good application scenario.
Let’s take a look at the commercial situation of millimeter waves in the United States, Europe, South Korea, Japan and other countries. The result we saw in February this year is that the millimeter wave rate can reach 16 times that of 5G Sub-6GHz and 38 times that of 4G LTE. This is a network that has been technically realized and has been commercialized. The venue of the “Super Bowl”, the Raymond James Stadium in the United States can generally accommodate more than 100,000 people at the same time. The total network traffic of the venue can reach 4.5TB, and the peak download speed of some scenes can reach 3Gbps, which is the peak of 4G LTE. 20 times the download speed. In addition to high speed, millimeter wave also has the characteristics of low delay, its delay is only a quarter of 5G Sub-6GHz. Therefore, we can develop many applications that support low latency, and one of the good use cases is to empower XR.
At present, we have supported many VR products, and at the same time, we have seen that millimeter wave can support low-latency applications. For example, millimeter wave-based 5G XR cloud gaming application scenarios have been or can be commercialized at any time. So in the application scenario of millimeter wave, what do we think is the most important? In fact, most of the technical bottlenecks of millimeter wave have been overcome. What we are talking about is based on millimeter wave support for Release 15 and the applications it can bring to everyone: including multi-view 4K video transmission and sharing For example, in a concert scene, the audience can see everyone in the band through 4K video; in a stadium with high-density connections, the audience can have an overall grasp of the game, and they can also see a specific How the stars dribble the ball; in industrial applications, millimeter waves can also provide precision positioning; in addition, millimeter waves can also support 5G enterprise applications, and multiplayer online games.
At present, we already have millimeter wave technology to support high-speed and low-latency applications. Only by getting first-hand feedback from these commercial applications can we know what else millimeter wave has in the process of leading to 5G advanced and 6G. The bottleneck needs to be resolved. After the millimeter wave was applied in the first version of the 5G standard, we also saw some 5G features that need to be enhanced in the second and third versions. The first is the technical enhancement of millimeter wave networking, that is, how each small millimeter wave base station is connected to the next base station. This technology is called integrated access and backhaul (IAB). In addition, 5G mobile phones or millimeter wave mobile phones will consume more power, and we will enhance its power-saving features to reduce power consumption. In addition, we have also done many enhancements including positioning, as well as projects for Release 17 and future versions.
The first level I wish to introduce is that 5G millimeter wave technology is a relatively mature commercial technology. If we want to lead 6G in the future, we also need to promote millimeter wave applications in China. I once heard the opinion of a domestic scholar that China’s research in the millimeter wave field started almost at the same time as other countries in North America. However, if there is still no millimeter wave application in China, it will be difficult to get relevant application feedback, and it may be possible in the future. The industry has more delays in millimeter wave applications. Qualcomm has completed a lot of interoperability tests with its Chinese partners, and has also conducted a lot of millimeter wave applications and promotion under the guidance of the Institute of Information and Communications Technology. We look forward to China’s commercialization of millimeter waves as soon as possible.
On the second level, I want to talk in depth about the rapid development of 5G in some vertical fields that we have seen. In addition to mobile phones, 5G technology has penetrated into a variety of different fields, one of which is a very important field is automobiles. We know that many mobile phone manufacturers have announced that they will enter the automotive field. From a certain perspective, a car is a very large mobile phone with four wheels. The car can be equipped with 5G technology, cellular car networking (C-V2X) technology, digital cockpit, advanced driver assistance system (ADAS) and autonomous driving. Technology, as well as car-to-cloud services, each of which is related to 5G or future 6G. We hope that 5G can be better applied in the automotive field, and it can be foreseen that the application in the automotive field will have a direct impact on the development of 6G.
First, a simple example is that the driver can receive real-time communications while driving, and can obtain vehicle information through technologies such as the Internet of Vehicles. For example, when other vehicles enter the monitoring range, these technologies can provide the driver with coverage of the perspective and provide better security; at the same time, they can also see the video data of other vehicles. These are very good that 5G can support. Application scenarios. In addition, in terms of vehicle electrification and automation, and autonomous driving, 5G can also provide strong technical support. One is the Internet of Vehicles technology and the other is high-speed 5G connectivity. A fully automated driving solution based on a pure vision system has very high requirements for computing power. If computing power cannot be solved in the car but needs to be solved in the cloud, ultra-high-speed, low-latency, and high-reliability data transmission is required. , This is a problem we think we need to solve with 5G connections as much as possible. If 5G still cannot solve it, 6G may be able to provide more choices for good application scenarios and solutions.
Automakers are adopting 5G technology faster than 4G. From 2013 to 2014, there were only two automakers that launched 4G cars; and between 2021 and 2023, they have released or plan to release auto manufacturers that support 5G cars. There are more than 18 merchants, which is a very good trend. After 5 or 10 years, when 6G arrives, all vehicles will need strong wireless connections. This will not only have a great impact on the connection between cars and cars, and between cars and roads, but also make smart cities and Intelligent networking becomes possible. Of course, these functions also have high requirements on the network and wireless communication technology. Looking at the needs of our technology from the perspective of application, it will have certain guiding significance for the next step of 5G and 6G research and development.
Another example is the digital chassis. Here, the chassis of the vehicle is actually the same as the mobile phone platform. It needs to be connected to artificial intelligence, needs to be connected to the vehicle, and needs the support of various wireless technologies such as the Internet of Vehicles. . At the same time, on our Snapdragon Ride platform, supporting autonomous driving will require a lot of sensors, including ultrasound, mid-range radar, long-range camera, mid-range camera, etc., to ensure the realization of the automatic driving function. A very important direction for us to do 5G or 6G research is to make this system more effective. Integrated Sensing and Communication is one of the directions of efforts. From another perspective, we also need such a platform, so that technologies such as wireless technology, artificial intelligence and sensor fusion can be very well applied.
The third use case is enterprise private network and industrial Internet of Things. This is a very suitable use case for applying 5G. At the beginning of 5G design, it is considered to be one of the representatives of the 5G transformation industry, and it is also a direction actively promoted by the country. If we see a modern factory with industrial robots, handheld terminals, automated guided vehicles (AGV), computer vision, sensors, the three major scenarios of 5G can easily find its applications. In addition, we are also promoting a variety of technologies to meet industrial IoT requirements, such as ultra-reliable low-latency communications (URLLC), licensed, shared and license-free spectrum, time-sensitive networking (TSN) and positioning. I think that on this issue, there is still room for 5G to be strengthened.
For example, the first one is positioning. 5G positioning was introduced in Release 16 and Release 17. Now Release 16 can achieve positioning accuracy of 3 meters indoors and 10 meters outdoors, and Release 17 hopes to achieve sub-meter accuracy. In industrial scenarios, the required accuracy is often higher than the meter-level accuracy. This is a very good technical requirement that requires us to continue to promote and try to make 5G better.
Another example is Time Sensitive Network (TSN). It is often questioned that 5G can achieve millisecond-level latency on the air interface, but in the network, the latency will reach tens of milliseconds. Now, with Time Sensitive Networking (TSN), we have taken the first step. However, Time Sensitive Networks (TSN) can only be time-stamped at present to inform the network of the delay, but cannot minimize the network delay. How to make fundamental changes and minimize network latency in 6G is a question worthy of further consideration.
When talking to a medical expert once, he said that I don’t actually need very cool application scenarios. What I need is a promise that the delay will never exceed 10 milliseconds or X milliseconds. It is like telemedicine or telesurgery. It’s not enough to demonstrate success alone. It is necessary to ensure that every time is foolproof. This is actually very difficult and requires us to achieve the ultimate in technology. Can 6G solve this problem and achieve this goal? This should be a very good indicator for us to consider from a commercial or practical point of view. 5G has made a lot of technical standards, and we also hope that these technologies can be truly commercialized.
Next, I want to continue to share the general trends of 6G, one of which is that we see the integration of the digital world, the physical world, and the virtual world. There has been a lot involved here before, so I won’t go into it in depth.
Another trend that we are more exposed to is the combination of 5G or 6G and artificial intelligence. Such a combination will greatly change the network architecture. In addition, many artificial intelligence algorithms are in the cloud, and only the cloud can meet the highest algorithm requirements. However, if this demand cannot be met, some technologies need to be placed on the edge or locally, which will have a far-reaching impact on the network architecture; it will also rationally allocate resources on the terminal side, the cloud, and the edge to provide reliable wireless communication. Put forward higher requirements.
At the same time, this is also a typical case of realizing XR requirements. When we talk about Metaverse, what we can provide in the communications field is high-speed, low-latency, reliable and effective connections. One of our concerns is the problem of network architecture, and the other is the problem of quality control in transmission. The problems of Metaverse in social and gaming levels are not solved by communication technology.
Another good intersection between AI and communication is how to use artificial intelligence to solve certain problems in communication. The communication field is a relatively mature field. We can find theoretically optimal solutions in many algorithms. The wireless communication problems that need to be solved by machine learning are those that we are not good at; in other words, they are due to computing. The amount is too large, leading to problems that we know how to solve but have no way to solve. This is a direction in which I think artificial intelligence has guiding significance for the development of communications.
I have talked a lot about the direction of 6G. Qualcomm hopes to use our technology to maximize the potential of 5G. At present, the technical issues involved, such as the issues in artificial intelligence, enhanced positioning, green energy-saving networks, sensing and other fields that I mentioned just now, we will try our best to solve them under the 5G framework. Of course, there will definitely be parts that are difficult to solve under the 5G framework, and we can try to solve them in 6G. And our perspective on 6G is more based on the continuously evolving technology in the development process of 5G, as well as everyone’s needs. In the process of technology research and development, our research does not have to label a certain technology as 5G or 6G. Our ultimate goal is to apply and promote 5G technology to everyone’s daily life, so as to make it truly affect people’s lives and society. We hope to communicate and cooperate with all partners to jointly promote not only the evolution of the technology itself, but also the development and application of technology in the country and the world.
The Links: LM190E08-TLJ2 LM80C03P