Just last week, 3GPP reached another major milestone: REL-18, the first version of 5G Advanced (5G-A), came to an end at the 3GPP Plenary on December 11-15, which also approved the main scope of REL-19 and defined the future direction of 5G-A. This heralds the official arrival of the 5G-A era.
In the face of the continuous evolution of 5G networks, Ericsson has proposed the "Next W**e of 5G" development plan, which is supported by four pillars - excellent performance, green and low-carbon, enabling growth, and high self-intelligence, aiming to build a value-driven excellent network.
2023 is a stage of vigorous development of the communications industry, on the one hand, China continues to lead the development of the entire 5G industry and technology, and on the other hand, we see that the global market is showing a pattern of a hundred flowers. Wu Riping, General Manager of Ericsson China's Network Product Solutions, pointed out in his speech at the "2023 Communications Industry Conference" that the sustainable development of communication networks has good potential, and at the same time, with the occurrence of more innovative technologies, telecom operators are putting forward higher requirements, especially in terms of sustainable development and network value monetization, and the "5G new wave" is Ericsson's answer to these challenges.
The potential for 5G development is still huge
According to the latest data from the recently released Ericsson Mobile Market Report, the number of 5G subscriptions will reach 1.6 billion by the end of 2023, accounting for nearly one-fifth of the total number of mobile subscriptions worldwide. The number of new 5G subscriptions in 2023 is expected to reach 6100 million, an increase of 63% over 2022. Specifically, in the third quarter of 2023, China reached 13 million new signings, making it the country with the largest growth.
What's more, we see great potential for continued development of the entire network, with traffic generated by mobile phone users expected to continue to triple today between the end of 2023 and the end of 2029, and mobile data traffic carried by 5G networks will increase from 25% at the end of 2023 to 76% by 2029. It can be said that 5G networks still have great potential for coverage and capacity development. Wu Riping said.
At the same time, he cited the GSMA report and said that according to the impact of the introduction of 5G on the average revenue per user (ARPU) of operators in eight markets around the world, six of them have changed the previous trend of declining ARPU value and showed positive growth. Even in the European market, where the development of 5G is relatively slow, there are signs of significant improvement in ARPU values.
The combination of communication networks and AI is inevitable
Among the four pillars mentioned above, a high degree of self-intelligence means Ericsson's successful practice in combining artificial intelligence (AI) with communication network technology. Wu Riping said that the industry has sufficient motivation to introduce AI, because the network has become more and more complex, whether it is the network itself, or the services and terminals carried on it. And AI itself is unquestionable in terms of efficiency. "The combination of communication networks and AI is not an option, but an inevitability. He stressed.
Based on this, Ericsson began to invest in the integration of AI and network at an early stage, and planned to develop the technology into four phases: the first stage is rule-based automation, and all scenarios need to be considered when designing automation functions;The second stage is an autonomous network based on AI ML, which uses AI ML** and adapts to some unknown situationsThe third phase is an intent-based cognitive network, which supports a more abstract expression of business requirements rather than an implementation scenario;The fourth stage is the zero-touch network, or AI-native network, which is self-configuring and self-optimizing with minimal human intervention.
Ericsson has already gone through the second phase of commercial deployment and technology accumulation, and is moving towards the third phase of intent-based cognitive networking. Wu Riping especially emphasized, "We are not building AI communication networks for AI's sake, but improving the performance and experience of communication networks through AI." Ericsson believes that AI capabilities can only be better leveraged by selecting the most appropriate network architecture.
Deploy AI with the most appropriate architecture
In September 2023, GTI, together with China Mobile, Ericsson and other industry partners, jointly released the "GTI 5G Wireless Network Intelligent Technology Requirements", which proposes a three-layer architecture: cross-domain intelligence, network intelligence, and network element intelligence, which is consistent with the architecture of centralized AI and distributed AI in Ericsson's concept.
Wu Riping explained, "The AI of communication networks should have a hierarchical architecture, that is, a combination of centralized AI and distributed AI. Depending on the characteristics, including whether the latency involves network-wide collaboration or single-node optimization, select the best node to implement AI, rather than forcing all AI to be done at a certain node." Among them, centralized AI involves network-level collaboration, and the decisions made are complex, but the frequency is low, and the real-time requirements are low, requiring manual access. Distributed AI, on the other hand, involves node-level intelligence, which makes decisions that are simple but frequent, require high real-time performance, and need to be automated.
In response to this, Ericsson provides a centralized AI platform architecture based on EIAP RAPP, which can provide very good AI support capabilities and has a completely open interface, which can be well integrated with communication networks. At the same time, Ericsson's base station product series continues to upgrade the capabilities of the hardware platform (silicon core technology) to continuously improve the support capabilities for distributed AI.
Ericsson's AI network practice results
As a global company with roots in China, Ericsson has always been committed to bringing the best and most cutting-edge technologies to the Chinese market first, and this is also the case in AI solutions for communication networks.
In his speech, Wu Riping introduced Ericsson's innovative practice in the Chinese market and China Mobile - intent-based AI multi-dimensional base station energy-saving optimization, through the introduction of multi-dimensional joint optimization, high-speed closed-loop control (based on KPI intent: such as energy-saving or performance-first control), and data and ML-based judgment and ** of wireless scenarios. Using the same power-saving functions, Ericsson achieves an average power-saving gain of 13 in performance-first scenarios by intent-driven energy-saving optimization of AI base stations1%, and the average energy saving gain is 19 in the energy-saving priority scenario3%。
At the same time, Wu pointed out that generative AI also has very good application scenarios in communication networks. "Looking to the future, we see that generative AI, including semantic communication such as effectively extracting feature words, and how to effectively simplify the investment in digital twins, instead of a lot of ** development, has very good results. However, in Wu's view, the key point of generative AI and other AI applications in communication networks is the same, that is, AI must be combined with communication expertise to produce the best results, and this is the direction that Ericsson has been working towards.
In terms of global AI practice, Ericsson is making comprehensive efforts in three aspects: data, algorithms and computing power. In terms of data, at the beginning of 2023, the Ericsson AI platform has processed more than 175+ TB of network data per dayIn terms of algorithms, Ericsson focuses on selecting the right AI algorithm and combining its communication expertise, and has tailored more than 200 AI applications for operatorsIn terms of computing power, Ericsson has built three ICT centers around the world to provide sufficient R&D computing power for model training.
At the end of his speech, Wu Riping emphasized that Ericsson will continue to be committed to bringing new technologies to China's networks in the new wave of 5G, helping operators build better 5G networks and share a better digital future.
C114 Communication Network Ace.