Since the Chinese Ministry of Industry and Information Technology granted the first 5G licenses for commercial use in June 2019, 5G has accelerated its proliferation in real life applications in China. Even though 5G has not yet spread to all parts of China, the research on 6G has already begun. Tech giants such as Huawei, LG, and Cisco have announced and launched their 6G research plans. What are the challenges related to the large-scale commercial use of 5G? What problems need to be solved for 6G? On August 16, at the sub-forum for intelligent networks during the Second Zhejiang Lab International Youth Talent Forum, Zhao Zhifeng, Head of the Department of Research and Development and Deputy Chief of the Research Center for Intelligent Networks at Zhejiang Lab, made a presentation entitled “From 5G to 6G”, providing an in-depth explanation of the ten major challenges in the development of 5G and 6G technologies.
Limited Spectrum Resource
In the field of mobile communications, the frequency spectrum is a fundamental and critical resource that carries wireless services and facilitates the development of the industry. Due to limited frequency bands, spectrum resources are licensed and pre-allocated for fixed use in China. However, in the 5G/6G era, we need more spectrum bands to support greater bandwidths, promote lower latency, and achieve higher speeds. In this way, we can achieve the interconnection of millions of devices, and meet the drastic increase of data traffic. Consequently, we will see a serious shortage of spectrum resources.
According to Zhao, although the bandwidths required for 5G and 6G can be extended through spectrum shift, which includes high-frequency and small-millimeter base stations, this method will cause problems such as insufficient signal coverage and substantial increase of construction costs for base stations. He believes that “intelligent spectrum sharing and reuse is a necessary precondition for 6G”. Focusing on improving the utilization of spectrum resources, this frugal strategy can effectively alleviate the shortage of spectrum resources.
Inefficient Cellular Networks
The mobile communication networks from 1G to 4G have all been operated on cellular architectures, using hexagonal cells to form cellular networks. These architectures can reduce terminal power and increase system capacity by reusing frequencies, but they are inadequate in terms of signal coverage. Though current heterogeneous cellular networks have improved the coverage, they make spectrum management and user access more complicated, undermine the spectral efficiency, and cause low power efficiency, which will hinder the development of 5G/6G.
Zhao believes that for future innovation, our efforts should be focuses on three parallel aspects, including finding alternatives to current network architectures, improving system standards, and making technical breakthroughs. In addition, he believes that future 6G networks will evolve into hyper-cellular and user-centered networks. In particular, the hyper-cellular technology will separate user access control and data transmission in the frequency band, so as to reduce the complexity of spectrum management. User-centered networks will use beamforming to realize signal coverage based on users’ needs and thus reduce power waste.
Inefficient Mobile Communications
The development of 5G is not only impeded by the low efficiency of cellular networks, but also by the inadequate energy efficiency of communication networks. “This dilemma is not only caused by a tension between high data flow and low information, but also the tension between technological progress and green development,” said Zhao.
Behind the low energy efficiency of mobile communications, there are a series of complex problems, such as energy waste caused by photoelectric conversion and excessive electromagnetic radiation. Currently, though technologies such as optical domain processing, on-demand coverage, and network protocol optimization can improve communication efficiency to a certain degree, we still lack revolutionary theoretical and technological breakthroughs.
Surging Demand for Bandwidth
With the rise of application scenarios such as three-dimensional MR, holographic communications, intelligent cars, and holographic digital humans, network traffic has been increasing dramatically, which leads to a greater demand for bandwidth. For example, in the field of medical care, when creating a holographic digital human for a patient, with wearable devices and some built-in sensors, the patient can generate 1 TB of static data and 10 Gbps of dynamic data per second. For emergencies, the data need to be transmitted for real-time interaction with hospitals when the patient is being moved at a high speed. Even 5G is unable to meet such a high demand for bandwidth.
Facing the surging demand for greater bandwidths, some methods, such as Multiple-Input and Multiple-Output (MIMO) and spectrum shift, can be used to extend spectrum bands. Zhao pointed out that, current technologies such as optical switching and all-optical networks, though facing difficulties in networking, can solve the problems that hinder the end-to-end Tbps data transmission in ground networks. However, the demand for greater bandwidths is still rigid, and it is expected that this demand will trigger massive tech reform.
Strict Latency Requirements
In the 5G era, the requirement for ultra-low latency is further enhanced. The latency with air interfaces (the communication links between terminals and base stations) and the end-to-end latency has been lowered to 1 millisecond and 10 milliseconds, respectively. However, the industrial internet has more stringent requirements for latency, with end-to-end latency dropping to 0.1 milliseconds and the latency with terminals to 0.02 milliseconds. Zhao pointed out that, though solutions such as minimal protocols, minimal store-and-forward, and minimal conversion (photoelectric conversion) have partially solved the latency issues to a certain degree, our overall technical system still faces challenges that require us to redesign the hardware and software of terminals, as well as the access mode. Otherwise, we will be unable to meet the demands for communication networks in certain application scenarios such as the industrial internet and automated VR driving.
Insufficient Coverage of the Ubiquitous Network
Although current mobile communication networks have been available to 70% of the world’s population and have covered 20% of the global land, it cannot be ignored that more than 80% of the land and 95% of the oceans in China are still out of signal coverage.
To address this problem, the research on integrated space-ground networks and satellite internets is now in full swing. However, as Zhao pointed out, these two technologies require high energy consumption and it is difficult to develop adaptable terminals. Moreover, based on existing technologies, we are still unable to extend signal coverage to underwater areas, to space, and to the inside of human bodies. It can be said that the Internet of Everything and the ubiquitous network are still facing enormous technical challenges.
Endogenous Security Issues
In the 5G era, we will see “more diversified terminals, more extensive application scenarios, and more open systems”. In fact, 5G has already become part of the internet. However, existing mobile communication networks are relatively closed, lacking systematic endogenous security design. In addition, as current communication security strategies are only designed for user authentication, billing, and communication encryption, we lack countermeasures against the security risks of the 5G era. When the 6G era arrives, it will be necessary to integrate the security system of computer networks with that of mobile communications, not only for user authentication, but also for security purposes.
Contradiction between Complexity and Reliability
In the 5G era, although the integration of communication technology (CT), information technology (IT), and data technology (DT) has made the communication system less complicated and more flexible, it has also caused problems that cannot be ignored, such as reduced reliability and greatly increased difficulties in system maintenance. Network Function Virtualization (NFV) is an area of technological development, but it also faces reliability issues. Precisely speaking, virtualization is to implement traditional functions through software. However, when executing the same functions, software usually performs worse than purposely designed equipment. How can we ensure 99.999% high availability of carrier grades? This question is worth thinking about and it also reminds us to be wary of the trap of NFV.
Contradiction between Slicing and Splitting
The 5G era brings demands for greater bandwidths and massive Internet of Things, as well as highly reliable connections with low latency. However, application scenarios related to these three major demands usually take place in different frequency bands. Therefore, traditional network slicing faces the contradicted requirements of scenario splitting.
Disputes about network slicing and scenario splitting are mainly focused on the following topics: as these two technologies have contradicted optimization goals, can these goals co-exist? Is end-to-end slicing applicable universally or not? If the splitting can be unlimited, is there still a network for 5G/6G?
Artificial Intelligence and 5G/6G development
As 5G/6G and AI are two hot topics of concern, it is impossible to skip the huge impact brought by 5G on AI when we talk about the development of 5G.
Zhao emphasized that, to analyze the impact of 5G on AI, we must first understand the relationship between the two. It is for sure that AI theories and technologies can meet the development needs of 5G/6G. However, the application of AI in 5G/6G cannot be a simple transplantation. The key is to fully leverage the efficacy of AI technologies. Meanwhile, as 5G/6G will create new technological areas such as distributed learning and group intelligence, they will further expand the territory of AI.
Zhejiang Lab’s Actions
Zhejiang Lab is accelerating its pace to implement its 5G/6G research plans through four major research centers targeting different areas including aerospace networks, intelligent networks, information security, and the industrial internet. In particular, the Research Center for Intelligent Networks aims to create a new technological system for B5G/6G. Focusing on key scenarios such as the Internet of Things, the Internet of Vehicles, mobile internet, and efficient data-centered networks, the research teams will study critical technologies such as terahertz and orbital angular momentum, all-optical switching, and full-dimensional definable networks. They will develop Tbit photoelectric terahertz communication devices and systems, build vehicle networking systems for intelligent transportation, and achieve breakthroughs for all-optical switching chips and network protocols. With world-class technical specifications and application goals, the Lab will take the lead in 6G research.