5G Advanced: On the road to 6G

While the first phase of 5G standardisation was completed with Release 15, the focus has shifted to 5G-Advanced, which is being developed through Releases 18, 19 and 20. As has been the case with previous generations, the advanced version of the technology serves as the bridge to the next generation. In that sense, Release 20 is particularly important because it is laying the foundation for 6G.

Much of the discussion around 5G in recent years has centred on monetisation. Yet, one of the industry’s fundamental challenges has been that many networks continue to operate in non-standalone mode. The reality is that a non-standalone network offers limited opportunities to unlock the full capabilities of 5G. Therefore, a first critical step is the migration from non-standalone architecture (NSA) to standalone architecture (SA).

The pace of this migration has remained slow globally. While SA adoption remains limited globally, markets such as China, India, Singapore and the US have emerged as leaders.

India is one of the leading markets in the transition to standalone 5G. The country has crossed 400 million 5G users, deployed more than 500,000 sites and achieved healthy device penetration. Data consumption is increasingly shifting towards 5G, creating a strong foundation for the adoption of more advanced services. SA is what enables operators to tap the full range of capabilities being introduced through 5G-Advanced. Without standalone deployment, many of the capabilities associated with slicing, reduced capability (RedCap) devices and advanced service differentiation cannot be fully realised.

Beyond connectivity

5G-Advanced introduces a range of capabilities spanning performance improvements, new device categories, differentiated services, network exposure, automation and satellite integration.

Mobility, coverage and radio performance continue to improve. Faster handovers, enhanced uplink performance and improvements in multiple input multiple output are addressing some of the long-standing limitations associated with mobility, coverage and capacity.

Moreover, entirely new device categories are beginning to emerge. RedCap devices enable lower-cost, lower-power implementations for wearables, sensors and industrial applications, but require standalone networks to realise their full benefits.

Further, one of the important developments is network slicing. Network slicing is one of the clearest monetisation opportunities in 5G, enabling differentiated services for consumer, enterprise and mission-critical applications.

Satellite integration is becoming increasingly important, supporting seamless connectivity between terrestrial and non-terrestrial networks. The long-term objective is a unified architecture that allows users to move seamlessly between terrestrial and satellite networks, ensuring seamless connectivity.

Another notable shift is the growing emphasis on network exposure and intelligent programmability. Through standardised APIs, exposure frameworks and initiatives such as open gateway, network capabilities can be made available to enterprises and application developers in a uniform manner. Combined with edge computing, this creates opportunities for new services across healthcare, transportation, digital public infrastructure and other sectors while opening additional revenue streams for operators.

The AI challenge

While 5G-Advanced is expanding what networks can do, artificial intelligence (AI) is beginning to redefine what networks need to become.

For decades, network growth was largely driven by human communications, followed by video consumption and machine connectivity. AI introduces a different traffic profile altogether. The challenge is not that AI tokens are large; in fact, they are very small. The challenge lies in the sheer frequency of conversational exchanges and the massive amount of traffic generated between inference systems and AI clusters.

This creates a new kind of pressure on networks. Internet and video traffic primarily consumed downlink capacity. AI will increasingly consume the latency budget of the network.

AI may not be showing serious impacts on networks today because it still represents a relatively small share of overall traffic volumes. However, as AI adoption accelerates, operators will have to prepare for a future where responsiveness, low latency and distributed processing become far more important than they are today.

Cost optimisation and automation will remain important, but they can only go so far. However, future AI workloads will require architectural changes, including distributed compute, edge inferencing and enhanced programmability.

The transition from 5G to 5G-Advanced is therefore about preparing networks for a future in which networks evolve from connecting data to understanding and processing intelligence. That is why Release 20 is attracting so much attention. It is the stepping stone that will shape the architecture, capabilities and expectations of the 6G era.