In a rapidly evolving technological landscape, the surge in artificial intelligence (AI) is transforming the mobile network sector at an unprecedented rate.
On the sidelines of MBBF 2024, Telecom Review conducted an exclusive interview with Fang Xiang, Vice President of Huawei Wireless Network Product Line, and a leading expert in this domain, to further explore the implications of AI-based terminals and applications, particularly in relation to the evolution of mobile networks towards 5G and 5.5G.
Related: From Vision to Reality: Innovations Shaping 5.5G Connectivity
The Surge of AI in Mobile Technology
As we stand on the cusp of a new technological era, Fang observed, “The AI era is surging, and the explosion of AI-based terminals and applications is transforming the landscape of mobile networks.” At this year’s Mobile Broadband Forum (MBBF), a plethora of AI applications were showcased, all of which promise to reshape our interactions with technology. “Just as we moved from feature phones to smartphones, we are now entering a realm defined by AI-enhanced devices, which industry refers to as ‘AI phones,’” pointed out Fang.
The advancement of AI extends far beyond phones and devices; an entire ecosystem of AI applications is beginning to take shape. Fang added that AI phones are becoming capable of perceiving and responding like humans, helping users complete routines and tasks in the same way a real assistant would, creating a more intuitive and seamless experience. He emphasized the role of sensors—such as cameras and microphones—that allow AI systems to perceive the world around them. “These devices gather real-world data, creating a digital representation that is uploaded to the cloud for processing, establishing a vital link between the digital and physical realms,” he explained.
Fang highlighted the emergence of lightweight AI terminals, such as those evident within Ray-Ban’s smart glasses and Unitree Robotics’ robotic companions. “These innovations are set to enrich our lives by making AI services accessible anytime, anywhere,” he asserted. The integration of AI is not merely an enhancement but a transformation, signaling a significant shift in how people will interact with technology.
“This year has marked a pivotal shift in the application of AI technology,” noted Fang, allowing operators to enhance their service offerings while exploring new avenues for monetization. “The integration of AI into mobile networks is poised to drive substantial changes,” he said, adding that the integration could improve user experiences and operational efficiencies.
Network Evolution for AI Demands
Fang emphasized that, to accommodate the growing demands of AI applications, networks must evolve rapidly. “For an AI phone to deliver a human-like experience, it requires an end-to-end response time of about 400 milliseconds,” he explained. This low latency is crucial because, beyond the phone’s own response time and cloud platform processing, the combined delay across the wireless and core networks needs to fall within the 50 to 70 milliseconds range to maintain seamless and responsive interactions. This necessitates limiting air interface latency to just 20 ms. “Facilitating this low air interface latency means supporting the streaming and uploading of substantial data to the cloud swiftly, which, in turn, requires robust uplink speed” he explained. Fang emphasized the critical importance of the “double 20” network requirements—20 Mbps uplink speeds and 20 ms latency—to create a seamless link between the digital and physical realms. These capabilities are essential for enabling the digital world to understand and respond to human needs in real time, laying the foundation for more intuitive, human-centered digital interactions.
He introduced a second scenario involving smart and autonomous driving technology, which is gaining traction, particularly in the Chinese market. “As this technology evolves, we’re seeing uplink traffic consumption surpass downlink requirements,” Fang noted, stressing the urgent need for high uplink speeds and low wireless latency in this context. He envisions that future mobile networks will be characterized by the convergence of multiple frequency bands. “To meet substantial downlink needs, we’ll continue utilizing TDD bands for larger bandwidths, to meet the low latency and large uplink needs, we utilize the FDD network’s natural advantages of different uplink and downlink,” he outlined, adding that Huawei will also collaborate with device partners to leverage both FDD and TDD technologies to achieve multi-dimensional capabilities.
For the evolution of Mobile AI networks, it’s crucial to emphasize how a multi-dimensional target network design effectively addresses the wide range of experience requirements that mobile AI services demand. This design approach ensures that networks are not only optimized for high-speed data transfer and low latency but are also adaptable to various use cases, such as real-time processing, autonomous systems, and smart applications. By incorporating multiple frequency bands, advanced resource management, and intelligent scheduling, the network can seamlessly support the diverse needs of AI-driven services with low latency, large uplink and wide coverage. Ultimately, this multi-dimensional network strategy enables a more robust, scalable, and responsive infrastructure capable of delivering the high-performance experiences expected from next-generation mobile AI applications.
Addressing Product Challenges in Network Development
When discussing the challenges operators face in evolving their networks, Fang said, “We must consider device diversity, as current networks support a range of devices—from 4G-only to 5G NSA and 5G SA. It's crucial to effectively balance these different user requirements.” He noted that the variety of services is also expanding, which presents its own challenges. “In addition to traditional enhanced mobile broadband (eMBB), we’re seeing an increase in large uplink services, low-latency applications, and IoT services. We need to provide an adaptive user experience across this wide array of services,” he stressed.
To meet the "20-20" network requirements, Huawei outlined two key directions for technological evolution: vertical and horizontal. Vertically, Huawei’s GigaBand solution, which tailors networks to services and devices through Intelligent AIR and Intelligent Gear, has significantly enhanced both uplink and downlink user experiences. “This approach allows us to efficiently utilize available spectrum resources to meet diverse requirements, enhancing both downlink and uplink experiences,” he explained. He highlighted that the Gigaband solution addresses the fragmentation issues often seen in FDD frequency bands, which can lead to low spectrum utilization. “By pooling these resources, we can flexibly allocate RF resources to different devices and services, ensuring optimal performance,” he stated.
Horizontally, Huawei’s advanced multi-antenna technology offers greater simplicity, energy efficiency, and system capacity compared to other solutions. Uplink performance, in particular, has seen significant improvements. Huawei’s M-Receiver technology boosts uplink coverage and optimizes the pairing efficiency of single-user MIMO (SU-MIMO), enabling a consistent uplink speed of 20 Mbps across the network. He added, “Our new tri-band M-MIMO solutions provide larger bandwidth for both uplink and downlink, improving coverage by 10 dB while simultaneously delivering a better user experience and lower latency.”
Fang also elaborated on the advancements in Huawei’s multi-antenna solution, which also forms part of the innovative GigaGreen platform. “Our new mid-band 8T8R solution is 30% smaller, 30% lighter, and consumes 30% less energy compared to previous generations,” he noted. “This not only reduces operational expenditures for operators but also enhances coverage and improves user experience.”
Huawei’s next-generation GigaGreen RF module series incorporates both vertical and horizontal advancements. By leveraging these modules, operators can build robust networks that meet the multi-dimensional requirements of mobile AI applications, including large uplink capacity, low latency, and wide coverage. This enables them to improve the real-time service experience for users and seize new opportunities created by mobile AI.
Read More: MBBF 2023: Connecting the World with 5G And Beyond
A Future Shaped by AI Connectivity
In the context of Mobile AI, Fang asserted that low-band frequencies are critical for ensuring coverage. “This year, we are launching our low band 4T4R GigaGreen RRU solution, showcasing a 20% improvement in size, weight, and power consumption compared to previous models.” Through the GigaGreen solution, he affirmed Huawei’s commitment to supporting operators in meeting the demands of Mobile AI while simultaneously reducing operational expenditure (OPEX).
“By addressing vertical challenges with the Gigaband solution and tackling horizontal requirements with our multi-antenna GigaGreen solution,” Fang concluded, “We enable telecom operators to build a network with multi-dimensional capabilities. This allows them to offer a broader range of services and better monetize their network capabilities, thus, seizing the opportunities presented by the Mobile AI era.”
Fang’s vision reflects a future where technology not only connects but enriches lives, driven by the limitless potential of AI.
Huawei at MBBF:
Li Peng: Maximizing New Growth Opportunities in the Mobile AI Era
Huawei's Yang Chaobin: 5.5G Unleashes the Potential of Mobile AI
