Designing 6G - Enabling Scalable Mobile Connectivity for an AI-Driven Future

While 5G Advanced is slowly being rolled out globally, early work on 6G is already underway within 3GPP, and this Qualcomm and Light Reading webinar provided a timely look at how foundational design choices are being considered. A consistent message throughout the session was that 6G should not be viewed as a narrow performance upgrade, but as a platform designed to scale across radically different use cases, deployment models and device types in an AI-dominated era.

The starting point for this discussion is the observation that mobile networks are facing a convergence of pressures. Data traffic continues to grow, uplink traffic is becoming more significant due to sensing and AI-driven applications, and networks are expected to support far more than human-centric communications. AI agents, robots, vehicles, extended reality devices and ambient IoT all place very different demands on coverage, latency, reliability and energy efficiency. Designing 6G around a single dominant service model would therefore be fundamentally limiting.

From a radio perspective, scalability was framed as a first-order design requirement. Rather than pushing peak data rates in isolation, the focus is on how a 6G radio access network can efficiently scale coverage and capacity while remaining economically viable. This includes extending the useful life of existing spectrum assets, improving spectral efficiency in lower and mid-band frequencies, and selectively introducing new spectrum where wide bandwidths are needed. Importantly, this is not presented as a wholesale replacement of existing infrastructure, but as a layered evolution that builds on 5G investments.

A significant technical theme was the evolution of the air interface itself. The webinar highlighted ongoing work in 3GPP Release 20 studies, particularly around 6G radio design. Concepts such as evolved waveforms, enhanced MIMO schemes, improved reference signal design and tighter coordination between duplexing modes were discussed in the context of delivering meaningful gains in efficiency rather than just headline throughput. In some bands, the ambition is to achieve substantial performance improvements without requiring a complete replacement of radio hardware, which has obvious implications for deployment cost and sustainability.

AI-native operation featured heavily throughout the session, but in a very specific sense. The argument was not that AI will simply automate existing network functions, but that AI should be embedded into the protocol stack and system architecture from the outset. This includes AI-assisted radio resource management, adaptive beamforming, mobility optimisation and dynamic service delivery. By treating AI as a native capability rather than an overlay, 6G networks could adapt in real time to user context, application requirements and environmental conditions.

This AI-native approach also extends beyond the network to devices and cloud infrastructure. The speakers described a tiered inferencing model, where AI processing can occur on the device, at the network edge or in the central cloud, depending on latency, energy and privacy constraints. Supporting this efficiently requires tight integration between connectivity and compute, something that current architectures only partially enable. In 6G, this integration is expected to be much more explicit, with the network actively participating in AI workflows rather than simply transporting data.

Architecture was another area where the webinar moved beyond incremental change. A strong case was made for transitioning from legacy reference-point-based RAN designs towards a service-based architecture with open APIs. The motivation here is twofold. First, it allows network functions to be deployed, scaled and upgraded using cloud-native principles, aligning telecom platforms more closely with hyperscale computing environments. Second, it enables new services, such as location, sensing and AI-driven analytics, to be exposed directly by the RAN rather than being tightly bound to specific interfaces.

This architectural shift also supports more flexible deployment models. Concepts such as dual-stack operation, where 6G can coexist with 5G or WLAN, were discussed as practical ways to introduce new capabilities without forcing disruptive network upgrades. Similarly, mechanisms for aggregating 5G and 6G carriers could allow devices to benefit from new spectrum and features even before 6G achieves full standalone coverage.

Integrated sensing emerged as one of the clearer examples of services beyond connectivity. By leveraging wider bandwidths and advanced antenna systems, 6G networks could support sensing functions such as object detection, environmental mapping and localisation alongside traditional communications. When combined with AI and digital twin concepts, this opens the door to applications in transportation, smart cities, industrial automation and public safety. The key point here is that sensing is treated as a native network capability, not an external add-on.

Security was addressed as an area where 6G is expected to build on, rather than replace, the foundations laid by 5G. However, the webinar highlighted the need to anticipate future threats, particularly from quantum computing. Concepts such as quantum-safe cryptography, stronger symmetric key algorithms and zero trust architectures were discussed as part of a more robust and adaptable security framework. The emphasis was on protecting both user privacy and network integrity across increasingly dynamic and distributed deployments.

Energy efficiency ran as a cross-cutting concern throughout the presentation. Supporting massive AI-driven traffic growth without a proportional increase in energy consumption is a core challenge for 6G. Many of the proposed techniques, including AI-assisted optimisation, dynamic activation of network functions and more efficient use of spectrum and antennas, are explicitly aimed at improving energy efficiency at scale. This positions sustainability not as a separate goal, but as an inherent design constraint.

Taken together, the webinar painted a picture of 6G as a system defined more by architectural choices and design philosophy than by a single breakthrough technology. The emphasis on scalability, AI-native operation, unified services and pragmatic evolution from 5G reflects a recognition that future networks must serve a much broader role in the digital ecosystem.

As 3GPP studies progress and early assumptions begin to harden into specifications, these foundational ideas will shape what 6G ultimately becomes. For now, this session provided a valuable insight into how some of the key technical questions are being framed, and why decisions made in the next few years will have long-lasting implications for mobile connectivity in an AI-driven world.

The video of the webinar is embedded below and the slides are available here.

Related Posts: 

Posted by on

Comments

Post a Comment