SKT and DoCoMo see 'room for improvement' in vRAN

"While various vendors are developing vRAN solutions, there is still room for improvement to fully realize anticipated benefits of vRAN," the operators said in the introduction to their latest joint white paper, snappily titled 'Key Considerations for vRAN: Insights from SK Telecom and NTT DOCOMO.'

Essentially, major vendors are mainly focusing on stabilising the performance of virtualised RAN and bringing it up to speed with existing RAN technology. As a result, the development of the specific features that will be unique to vRAN are on the back burner.

The white paper actually contains more questions than answers. It's basically a way for the Japanese and South Korean mobile market leaders to positioning themselves as thought leaders, or perhaps more to the point, technology leaders, when it comes to crucial aspects of 6G technology. It's their third such offering in the past 15 months, following a missive on the metaverse in November 2022 and another paper on green networks and 6G a couple of months later.

"The white paper reviews the future prospects of virtualized RAN (vRAN) base station equipment for use by telecommunications operators, drawing on the two companies' combined experience in network construction and operation," the pair declared.

But while this might not be a full-on operator roadmap for vRAN, it does provide valuable food for thought for the operator community.

The paper looks at some key future requirements of vRAN, including increased processing capacity and reduced power consumption; capabilities such as resource pooling, scaling, and auto-healing; and integration technology and tools for base station equipment. It can be downloaded here.

However, the main thrust of the document is around accelerator technologies and the choices mobile network operators will need to make in this area.

As DoCoMo and SK Telecom point out, mobile network operators need to be able to match the cell capacity and energy efficiency of traditional RAN if they are to move to vRAN. And key to that is the Layer 1 accelerator, which ease the workload on the CPU, and the ongoing debate over look-aside versus inline L1 accelerators.

In an over-simplified nutshell, vRAN uses off-the-shelf hardware, but the workload it places on a general purpose CPU could prove too great. L1 accelerators help to address that. But there are two types: look-aside, which are essentially integrated into the CPU, and inline, which require a separate card.

Inline accelerators can reduce the CPU workload more than an integrated look-aside accelerator and offer greater flexibility, but can increase complexity, to name a couple of key pros and cons.

As DoCoMo and SK Telecom point out, there's no one answer to which solution a telecoms operator should pick.

"The requirements of vRAN are not one-dimensional. They can vary based on network architecture and each MNO's specific requirements such as operational policies. Therefore, it is important to understand which L1 accelerator type is more suitable for each MNO's network deployment," they advise.

Or to put it another way, it's horses for courses.

Broadly speaking, operators with a centralised RAN architecture, or C-RAN, would lean towards inline L1 accelerators, due to their ability to support a large number of cells. But players with a D-RAN architecture – the D is for distributed – will find that cell capacity and scalability are a less significant factor, but could be attracted by the low power consumption per-site of look-aside.

There's more – of course – but those are some of the key points. The strongest message here is that mobile operators need to look carefully at their own network architectures and policies – per-cell traffic load, number of MIMO layers per cell and so forth – before making an accelerator decision.

"vRAN is gradually spreading in MNOs’ networks with its potential to provide various benefits, such as flexibility in software and hardware combinations and is also expected to be applied in the 6G era," SK Telecom and NTT DoCoMo's paper concludes. "However, existing challenges to further enhance performance and to fully achieve anticipated benefits necessitate ongoing research and development in vRAN technologies."

The telcos might not have provided all the answers here, but they are helping their peers to ask the right questions. And that's a large part of the battle.